mbfinotti/partnerships-skills已通过检查
SKILL DETAIL
partner-ecosystem-expansion
mbfinotti/partnerships-skills/partner-ecosystem-expansion
Decide which partner categories to add next to a partner ecosystem and in what sequence - a staged partner-category expansion roadmap with readiness gates, capacity limits, and kill criteria. Covers B2B categories (tech/ISV, reseller/VAR, MSP, SI/GSI, agency, referral/affiliate, marketplace) and consumer-side equivalents (retail/wholesale, creator, licensing, co-branding). Use whenever the user mentions partner ecosystem expansion, partner category planning, a partner mix roadmap, or asks which partner type to launch next, even if they never say expansion. Category-level sequencing only. Do NOT use for ranking named candidate companies - use mbfinotti/partnerships-skills@alliance-prioritization instead.
安装量 · 509查看来源
Installation
npx skills add https://github.com/mbfinotti/partnerships-skills --skill partner-ecosystem-expansion
技能文件
SKILL.md
最近同步 · 2026年10月23日
evals/evals.json›
{
"skill_name": "partner-ecosystem-expansion",
"evals": [
{
"id": 1,
"prompt": "I run partnerships at Hadleigh Forms (B2B form-automation SaaS, about $6M ARR, self-serve to $99/mo then sales-assisted). We launched a referral partner program 14 months ago: 62 signed partners, 5 of them have ever sent us a deal, and the only two deals that closed both came from the same person. My VP wants to add an affiliate program next quarter because it's basically the same plumbing and costs us almost nothing. Give me a 12-month partner category roadmap. First result has to land by end of Q2 and I have no new headcount coming.",
"expected_output": "A roadmap whose stage 1 is finishing the unproven referral motion rather than adding a category, with the 8% activation read as a stop-recruiting signal, the affiliate ask refused or gated, capacity and deadline treated as caps, and evidence-based gates plus a Not now list with re-entry triggers.",
"files": [],
"expectations": [
"States the referral motion is not yet a minimum viable ecosystem (MVE) because it has no repeatable joint wins, only one partner producing.",
"Computes or states the referral cohort's activation rate as roughly 8% (5 of 62).",
"States that activation below roughly 20% means stop recruiting and fix enablement, not recruit harder.",
"Makes stage 1 of the roadmap finishing or fixing the existing referral motion rather than adding a new category.",
"Declines to open an affiliate category in stage 1, or gates it explicitly behind the referral motion's gate.",
"Rejects 'it is cheap and reuses the same plumbing' as a sufficient reason to add a category, stating that cheapest and most efficient are different orderings.",
"Treats the end-of-Q2 date as a re-ranking input that promotes fast-acting categories and rules out any category whose first revenue arrives after that date.",
"Names the no-new-headcount constraint as a capacity cap on the roadmap, not as a detail to work around.",
"Attaches kill criteria that include an activation floor and a time-to-first-deal window.",
"States that every stage gate opens on evidence, never on a calendar date alone.",
"Places at least one category in a 'Not now' list with an explicit re-entry trigger.",
"Carries a provenance flag on any external benchmark quoted, for example flagging the sub-20% unmanaged activation floor as vendor-sourced."
]
},
{
"id": 2,
"prompt": "Norberg Ledger, B2B accounting-automation platform, $14M ARR, mid-market. Live today: a referral track (35 partners, 44% activation, joint wins repeating across at least 9 different partners) and nothing else. We have a documented public API, multi-tenant architecture, a sandbox, and 4 integrations customers built themselves without any help from us. Capacity: 2 partner managers, one enablement writer, $400k program budget for the year. Leadership's instinct is to just widen referral into a bigger affiliate network because it's the cheapest thing we could do. What should we actually add next, and in what order over 24 months?",
"expected_output": "A 24-month sequence opening with a formalized technology/ISV category, justified by the efficiency order and by the integration surface making a later marketplace stage cheap, with capacity numbers fixed before sequencing and at least one category deleted with a re-entry trigger.",
"files": [],
"expectations": [
"Recommends formalizing the technology/ISV category as the first added stage.",
"States the B2B efficiency order explicitly, with tech/ISV ahead of referral/affiliate (tech/ISV > referral/affiliate > agency > marketplace > MSP == SI/GSI > reseller/VAR).",
"Explains that referral/affiliate is the cheapest category but not the most efficient one, because nothing it builds makes the next category cheaper.",
"States that the ISV integration surface is reused by a later marketplace stage, and by agencies or MSPs, which is what makes those stages cheaper.",
"Names the 4 customer-built integrations as existing uncompensated third-party activity and the cheapest thing to formalize.",
"Confirms the referral motion already qualifies as a proven MVE (44% activation, repeatable joint wins) so stage 1 need not be finishing it.",
"Sequences any marketplace stage after the ISV stage, not before it.",
"Fixes capacity numbers (partner-manager-to-partner ratios, enablement bandwidth, budget) before presenting a sequence.",
"Opens one stage at a time, allowing at most two in parallel and only when they share enablement infrastructure, stated explicitly.",
"Sets each stage's gate threshold as repeatable joint wins across multiple partners, not a single first win.",
"Deletes or defers at least one category from the pool with the constraint that ruled it out and a re-entry trigger.",
"Does not recommend reseller/VAR as an early stage, and if it is discussed, names its structurally collapsed margin."
]
},
{
"id": 3,
"prompt": "Board slide says 'launch MSP tier, sign two global SIs, and stand up a reseller channel' within the next 12 months. Reality at Kestrel Vault (B2B backup and disaster-recovery software, $9M ARR, sales-led, average ACV $38k): I am the entire partnerships team. One person. We do have an ISV motion that's real, 11 integration partners and 6 of them actively sourcing. Nothing else exists. Product has no multi-tenant admin console and it is not on the roadmap. Sequence this honestly for me.",
"expected_output": "A sequence that deletes all three board categories from the candidate pool on capacity and readiness grounds, each recorded in Not now with a re-entry trigger, and proposes something the single-manager capacity can actually carry off the proven ISV base.",
"files": [],
"expectations": [
"Deletes MSP, SI/GSI and reseller/VAR from the candidate pool rather than scheduling them later in the same year.",
"States that a partner team of one cannot carry a standing-job category and the proven motion at the same time.",
"Names the missing multi-tenant administration as a readiness prerequisite failure that independently rules out MSP.",
"Records each deleted category in a 'Not now' section with the specific constraint that ruled it out.",
"Gives each deleted category a named re-entry trigger, such as multi-tenant administration landing on the product roadmap or a second partner manager being hired.",
"Refuses to park the ruled-out categories at the bottom of a ranked list, stating that a demoted category reappears later as scope.",
"Identifies the ISV motion as the proven base and builds adjacency from it.",
"Proposes at least one category the current capacity can actually carry instead of the board's three.",
"States explicit capacity numbers (manager-to-partner ratio, enablement bandwidth) as the cap on the roadmap.",
"Notes that 'sign two global SIs' names companies rather than a category, so it is not a category decision.",
"States that the efficiency order starves SI/GSI and MSP despite their high value, and names the condition that would promote them.",
"Frames the roadmap as a reasoned, gated bet rather than a validated formula."
]
},
{
"id": 4,
"prompt": "Reviewing this before it goes to our exec team Thursday. Company is Loomis Metrics, B2B analytics, $22M ARR.\n\nPARTNER STRATEGY 2027\nQ1: launch reseller program - industry data shows partners drive 28% of revenue\nQ2: sign 3 GSIs, launch MSP tier\nQ3: two hyperscaler marketplace listings + affiliate program\nQ4: 220 total partners, partnerships = 20% of pipeline\nFraming: Horizon 1 core resellers, Horizon 2 services, Horizon 3 ecosystem platform\n\nTighten this up for me, mainly the dates. I think Q2 is too aggressive.",
"expected_output": "A refusal to tune the dates, followed by a line-by-line teardown naming the missing capacity numbers, the calendar gates, the unflagged analyst figure, the logo-count target, the three-horizons force-fit, the reseller-first prerequisite failure and the missing conflict check, then a rebuild from the MVE and capacity up.",
"files": [],
"expectations": [
"Refuses to fix the plan by adjusting quarter dates, and rebuilds it from the MVE and the capacity numbers instead.",
"Names the complete absence of capacity numbers (who manages, enables and supports each cohort) as the plan's central defect.",
"Names recruiting ahead of enablement as the failure this plan schedules in advance.",
"Rejects McKinsey's three horizons as a force-fit borrowed from innovation-portfolio planning that supplies vocabulary but no sequencing logic.",
"States that the three-horizons framing is standing where a gate structure should be, so no stage in the plan can fail.",
"Rejects the 220-partner target as logo counting and replaces it with activation, time-to-first-deal and partner-sourced pipeline.",
"Flags '28% of revenue' as an analyst figure describing mature programs, irrelevant to a company launching resellers next quarter.",
"Attaches a provenance flag to that 28% figure if it is kept at all (Forrester, ANALYST).",
"Identifies every stage as calendar-gated and replaces the quarter triggers with evidence thresholds.",
"Identifies reseller-first as placing the highest-prerequisite transact category first, and names the prerequisites the plan never mentions: deal registration, margin structure, certification, support model.",
"Notes that the efficiency order puts reseller/VAR last and that the plan states no buyer-route reason that would promote it.",
"Points out that 'sign 3 GSIs' is a named-company goal rather than a category decision, and that SI/GSI is the slowest, highest-load category.",
"Names the missing channel conflict check for the reseller, MSP and marketplace categories that all overlap the direct motion."
]
},
{
"id": 5,
"prompt": "Building the justification slide for our partner expansion ask at Brightmoor Signal (B2B security operations platform, $19M ARR). Numbers I want to use: buyers now complete 68% of their journey digitally so we need influence partners; 95% of Microsoft's revenue flows through partners; mature programs get 28% of company revenue through partnerships; and around 25% of the channel struggles to break even so we should stay away from resellers. We'll target 35% activation on the new category in year one. Sanity-check these for me and tell me what to put in the deck.",
"expected_output": "A provenance audit of each quoted number, identifying the 68% figure as a revised 2013 statistic Forrester itself called a myth, the 95% as Microsoft's own 2019 self-report, the break-even figure as superseded by the Omdia survey, with flags on everything surviving into the deck and the activation target rebuilt from the company's own base rates.",
"files": [],
"expectations": [
"Flags the 68% digital-journey figure as derived from a 2013 SiriusDecisions statistic that Forrester itself later publicly called a myth and revised to '70-80% self-directed'.",
"States that the direction of that figure may hold but its precision does not, and refuses to present it as settled fact.",
"Traces the 95% Microsoft figure to Microsoft's own 2019 statement (Q2 FY2019 earnings commentary or corporate blog) and presents it as one company's self-reported commercial-revenue framing, not a current industry constant.",
"Flags the 28% mature-program revenue share as ANALYST (Forrester) and notes it describes mature programs.",
"States the '~25% of the channel struggles to break even' figure is superseded, and gives the current Omdia 378-partner survey figures: 58% expecting profit decline, 51% forecasting double-digit drops.",
"Attaches an explicit provenance flag to every benchmark that survives into the deck.",
"Uses or names the flag taxonomy consistently (ACADEMIC, ANALYST, VENDOR, VENDOR-COMMISSIONED, CONSULTANT, MIXED).",
"States that any number appearing in the roadmap without a provenance flag is a defect to fix before shipping.",
"Warns that nearly every published benchmark in this field comes from a party that profits when the number looks good, so vendor figures are directional context and never proof.",
"Tells the user to set the 35% activation target from their own existing categories' base rates, using published figures only to sanity-check.",
"Identifies academic alliance-portfolio research as the only independent, non-self-interested evidence base in this field.",
"Warns against comparing a vendor benchmark to a cohort measuring something different, naming signed partners versus active partners as the classic mismatch."
]
},
{
"id": 6,
"prompt": "Verrill & Co, DTC skincare, $18M revenue, 92% of it through our own site. We have an affiliate program (coupon and cashback sites mostly, roughly $1.1M attributed last year) and nothing else. Growth has been flat at about 4% year over year for three quarters. Everyone here wants to go after Target and Ulta next year because that's the biggest step available to us. Our founder has zero appetite for a standing job: she wants the first new channel producing within five months and won't fund permanent headcount. What's the category roadmap?",
"expected_output": "A B2C sequence opening with creator/influencer, deleting retail/wholesale and distributor on the stated effort ceiling while acknowledging flattened DTC growth as a genuine but overridden promotion condition, with incrementality as the sequencing driver and a cannibalization check on any overlapping route.",
"files": [],
"expectations": [
"States the B2C efficiency order with creator/influencer first (creator/influencer > affiliate > marketplace > retail/wholesale > licensing == co-branding > distributor).",
"Recommends creator/influencer as the first added category.",
"Deletes retail/wholesale and distributor from the candidate pool because the stated effort ceiling excludes a standing job and permanent headcount.",
"Records retail/wholesale in 'Not now' with a named re-entry trigger, such as headcount approved or a retail buyer arriving inbound.",
"Acknowledges that flattened DTC growth is a genuine condition that would otherwise promote retail/wholesale, and names which constraint overrides it here.",
"Names the five-month deadline as the input promoting fast-acting categories and ruling out any whose first revenue arrives later.",
"States that the existing affiliate program converts demand that mostly already existed and is rarely incremental.",
"Uses incrementality, reaching buyers before intent exists, as the B2C sequencing driver.",
"Notes that creator/influencer disclosure rules are enforcement-backed and that the liability is the brand's as well as the creator's.",
"Flags retail/wholesale and distributor as the categories this efficiency order starves, rather than silently omitting them.",
"Names DTC versus retail versus marketplace cannibalization as the dominant governance risk, and requires the conflict check before any overlapping category's launch stage.",
"Attaches kill criteria to the creator stage including a performance or activation floor and a time-to-first-revenue window."
]
},
{
"id": 7,
"prompt": "Ostergaard Industrial Cloud, B2B, $31M ARR, enterprise-only: average deal $220k, nine-month cycles, and every single customer needs three to six months of configuration before go-live. We already run a 28-person professional services org that does all implementations today and sits at 90% utilization. On the partner side we have a referral track (proven, 52% activation) and 9 ISV integrations. Two partner managers. Which category next?",
"expected_output": "A sequence promoting SI/GSI to the front against the default efficiency order, justified by the already-staffed services arm and the implementer-led buyer route, with the services-org conflict sized and rules required before launch.",
"files": [],
"expectations": [
"Promotes SI/GSI to the front of the sequence despite the default efficiency order ranking it near the back.",
"Names the existing 28-person services org as the reason: the standing job that makes SI/GSI expensive is already staffed and already carries the methodology.",
"States that the buyer's route to market outranks the effort ranking when the buyer purchases through implementers.",
"Notes that the default efficiency order starves SI/GSI and MSP, and names this user's specific promoting condition.",
"Treats the 90% services utilization as a capacity constraint rather than as spare capacity.",
"If the partner-services multiple is quoted ($7.13 of services per $1 of platform for services-heavy partners versus $1.26 for resell-focused), flags it as vendor-commissioned (Canalys for AWS) and directional only.",
"Names the conflict between an SI category and the in-house services org, and requires rules of engagement before that stage launches.",
"Lists SI/GSI readiness prerequisites that must exist: implementation methodology, certified-practitioner program, executive sponsorship, MSAs and sub-contracting terms.",
"Attaches an evidence-based gate to the SI stage requiring repeatable joint wins across multiple partners.",
"Opens one stage at a time and gates every later stage on the previous stage's evidence.",
"Keeps the output at category level and does not rank named consultancies."
]
},
{
"id": 8,
"prompt": "Piedmont Dispatch, B2B logistics SaaS, product-led, $99 to $899/mo self-serve, median time from signup to first value is 11 minutes, no onboarding call, no configuration. $7M ARR across about 4,000 paying accounts. Our new CRO came from an enterprise software company and wants a systems-integrator partner program plus an agency channel, because that's how they scaled at his last place. We have a public API and a growing set of third-party integrations. Two partner managers. Talk me through the sequence.",
"expected_output": "A sequence demoting SI/GSI and agency below marketplace because the product has no services gap, recommending ISV and marketplace instead, with the ego-system check applied and the CRO's precedent named as not being evidence.",
"files": [],
"expectations": [
"States that a product needing no implementation demotes SI/GSI and agency below marketplace in the ranking.",
"Names the absence of a services gap as the reason neither SI/GSI nor agency has a job to do for this product.",
"Runs the ego-system check, asking what an SI or agency partner gains by joining now, and notes a category with no answer will sign logos and never activate.",
"Recommends the technology/ISV and/or marketplace categories ahead of the CRO's request.",
"Uses the existing public API and third-party integrations as the adjacency that makes the recommended stage cheap.",
"Deletes or defers SI/GSI with the constraint that ruled it out and a re-entry trigger, such as an enterprise tier shipping that requires configuration.",
"States plainly that the CRO's precedent from another company is not evidence for this product's sequence.",
"Presents 2-3 candidate sequences with trade-offs and a recommendation rather than a single unargued answer.",
"Argues the strongest case against its own recommended sequence before presenting it.",
"Attaches per-stage gates that open on evidence, with kill criteria including an activation floor and a rule for two consecutive missed quarters.",
"Costs each stage as a total loaded cost (headcount, enablement build, incentives, tooling, program spend), not as a headcount ask alone."
]
},
{
"id": 9,
"prompt": "We need a partner roadmap for Ashgrove Analytics (B2B data platform, $12M ARR, mid-market and up). Candidates on the table: Accenture, Slalom, a regional firm called Northpoint Data, Snowflake's partner network, and two boutique agencies we've worked with informally for a year. Rank these and tell me who to sign first and in what order, plus what the gates should be.",
"expected_output": "A refusal to rank the named companies with a handoff to the named-candidate prioritization sibling, followed by the same request answered at category level with readiness prerequisites, gates and a Not now list.",
"files": [],
"expectations": [
"Declines to rank the named companies, stating this work is category-level sequencing only.",
"Routes the named-candidate ranking to the sibling skill for prioritizing named alliances (mbfinotti/partnerships-skills@alliance-prioritization).",
"Re-expresses the named candidates as the categories they belong to, such as SI/GSI, regional SI, technology/ISV or marketplace, and agency.",
"Produces a category sequence rather than a company sequence.",
"Asks for the current ecosystem picture (categories live today, partners per category, share active) before sequencing, or routes to the ecosystem-mapping sibling when none exists.",
"Names the efficiency ranking it used and states what that ranking starves.",
"Attaches readiness prerequisites per candidate category rather than per named firm.",
"Attaches per-stage gates with a pilot cohort shape, a partner-market-fit threshold of repeatable joint wins, and kill criteria.",
"Places at least one category in 'Not now' with a re-entry trigger.",
"States that the pilot cohort must receive full enablement, because a pilot starved of enablement tests nothing."
]
},
{
"id": 10,
"prompt": "Getting our MSP stage funded at Calderwood Telemetry (B2B observability, $16M ARR). Finance told me MDF is the easiest bucket to pull from: we already hold about 4% of channel revenue in MDF and it's underspent. So my ask is one partner manager's worth of time carved out of the existing team, plus $180k of MDF for partner marketing over four quarters. CFO wants the request expressed in headcount terms. Help me write the business case and the stage plan.",
"expected_output": "A business case that promotes headcount for this standing-job stage against the default funding-efficiency order, refuses the MDF-funded and headcount-framed asks, presents a total loaded cost, and writes MSP prerequisites, pilot shape and kill criteria before launch.",
"files": [],
"expectations": [
"States the funding-mechanism efficiency order (program budget > MDF > headcount) and then promotes headcount for this stage because it opens a standing-job category.",
"States that MDF and program budget buy activity, never the standing capacity an MSP category runs on.",
"Warns that a standing job funded from a one-period budget fails at the second review.",
"Presents the ask as a total loaded cost covering salaries, enablement build, incentives, tooling and program spend.",
"Rejects the CFO's headcount framing, stating the request is for capital rather than enthusiasm.",
"Notes typical MDF utilization is only 40-60%, and flags the MDF sizing figures (2-6% of channel revenue, 25-50% cost-share per activity) as mixed vendor/analyst.",
"Names MDF's compliance cost: per-activity claims, proof-of-performance administration and audit rights.",
"Lists MSP readiness prerequisites and marks each met or missing: multi-tenant administration, monitoring and billing hooks, tiered support, support model, contracting.",
"States that MSP is a standing job both to stand up and to run, with slow time to first revenue.",
"Defines a pilot cohort with full enablement and an evidence gate before any scale investment.",
"Writes kill criteria before launch, including an activation floor near 20%, a time-to-first-deal window, and the two-consecutive-missed-quarters rule.",
"Names cross-functional sign-off owners (product, sales, finance, legal) rather than finance alone."
]
},
{
"id": 11,
"prompt": "We finished a full partner ecosystem map six weeks ago: every partner type we have, account overlap data pulled from our overlap tool, coverage gaps by segment and geography, the whole thing. Now my SVP wants to book a two-day offsite with sales, marketing, CS and product leadership to brainstorm which partner types we add next. Tarrant Systems, B2B infrastructure software, $27M ARR. Can you design the offsite agenda?",
"expected_output": "A recommendation to start with the near-zero-cost map re-read and value blueprint rather than the two-day workshop, with the format efficiency order stated, the stopping rule given, and the condition that would genuinely promote the business-design workshop named.",
"files": [],
"expectations": [
"States the facilitation-format efficiency order with the ecosystem-map re-read first and the business-design workshop last (map re-read > value blueprint > account mapping > partner persona > business-design workshop).",
"Recommends starting with a re-read of the existing map rather than rebuilding it or opening with the two-day workshop.",
"States the stopping rule: run from the top of the order and stop when the candidate pool stops changing.",
"Names the condition that would promote the two-day cross-functional business-design workshop: the answer would change how a GTM function operates day to day, such as comp plans, CS ownership or data sharing.",
"Notes that the existing account-overlap data makes the account-mapping session cheap, because the week of data work is already paid for.",
"Includes the value blueprint session and its ego-system check, drawn from the customer's position rather than the company's.",
"Names co-innovation risk and adoption-chain risk as the two risks the value blueprint surfaces.",
"Applies one question per message and does not batch questions.",
"Offers 2-3 candidate approaches with trade-offs and a recommendation at each real choice, then waits for the user.",
"Names assumptions aloud, including which category the room already wants and which capacity number is a guess.",
"Presents the deliverable section by section, validated one at a time, and gates finalization on the user's explicit approval."
]
},
{
"id": 12,
"prompt": "Setting up reporting for the agency partner category we launched 10 weeks ago at Wrenfield Commerce (B2B ecommerce ops platform, $21M ARR). 14 agencies signed so far. I want the dashboard to lead with category share of new revenue, since that's the number our board asks about, then partner NPS from a quarterly survey. Stage gate review is in 8 weeks. What should I instrument and what should the gate read?",
"expected_output": "An instrumentation plan led by activation rate and time-to-first-deal, refusing category share of new revenue as the stage-gate read and naming the funding-decision condition that would promote it, with NPS demoted and the gate threshold set on repeatable joint wins.",
"files": [],
"expectations": [
"States the KPI instrumentation efficiency order with partner activation rate first (activation rate > time-to-first-deal > partner-sourced pipeline > enablement completion == deal-registration volume > category share of new revenue > partner NPS).",
"Declines to lead the stage-gate read with category share of new revenue, stating it needs roughly a year of cohorts before it says anything.",
"Names the condition that does promote category share of new revenue: a funding or board decision rather than a stage gate.",
"States partner NPS is opinion, useful for diagnosing a failing cohort but never sufficient to open a gate.",
"Defines partner activation rate as signed partners closing at least one deal, and makes it the gate's first read.",
"Adds time-to-first-deal measured against the expected window for the agency category.",
"Adds enablement or certification completion inside the onboarding window, noting the 90-120 day window with its ANALYST (Forrester) flag.",
"Sets the gate threshold as repeatable joint wins across multiple partners rather than a single first win.",
"Names the gate outcome vocabulary: pass-and-scale, extend-pilot once with a named fix, re-scope, or kill.",
"Warns that 100% gate survival suggests soft gates while 0% suggests a broken candidate pool.",
"Sets gate thresholds from the company's own base rates where any exist, using published figures only to sanity-check."
]
},
{
"id": 13,
"prompt": "Hollenbeck Data, B2B data-quality platform, $48M ARR. We run six partner categories now: referral, affiliate, ISV, agency, two hyperscaler marketplaces, and a reseller tier we added 18 months ago. Partner-sourced revenue has been flat at roughly $9.4M for five straight quarters while my team went from 3 to 7 partner managers, and we bought a PRM platform last year that nobody really uses. Exec team wants an MSP category added for FY28 and is also talking about cutting the agency tier to free up headcount. What's the right sequence?",
"expected_output": "A recommendation to add nothing and deepen instead, reading the flat revenue against rising coordination cost as the inverted-U signal with academic provenance, refusing the MSP addition, and treating the proposed agency cut as an equally high-risk subtraction.",
"files": [],
"expectations": [
"Recommends adding no new category now, and deepening the existing ones instead.",
"Identifies the inverted-U signal: coordination cost rising (3 to 7 partner managers) while total partner-sourced revenue flattens.",
"Cites the alliance-portfolio-diversity inverted-U research and flags it ACADEMIC, for example Jiang, Tao & Santoro 2010 in Strategic Management Journal or the Lee et al. 2017 meta-analysis.",
"States that this research argues for restraint and sequencing but prescribes no order.",
"States it is the field's only independent, non-self-interested evidence base.",
"Declines to open the MSP category for FY28, or gates it behind the freeze lifting with a named trigger.",
"Treats cutting the agency tier as high-risk as adding a category, requiring the conflict and dependency to be sized before shrinking that route.",
"Names over-rotation on one category while cutting others as a documented failure pattern, not a neutral cost saving.",
"Names the PRM platform nobody uses as tooling bought ahead of the workflow that needed it, sitting as compliance theater rather than as capacity.",
"Reads per-category activation before deciding anything, rather than treating the flat $9.4M as a single aggregate.",
"Names category sprawl followed by forced pruning as the pattern this ecosystem is already in.",
"Sets a review cadence and explicit re-sequencing triggers, such as a gate result, a category's activation collapsing, a product-readiness slip, a conflict incident or a market shift."
]
}
],
"trigger_queries": [
{ "query": "Which partner category should we add next, and in what order?", "should_trigger": true },
{ "query": "Build me a 24-month partner category expansion roadmap with readiness gates.", "should_trigger": true },
{ "query": "We have resellers and referral partners today. What's the next partner type to launch?", "should_trigger": true },
{ "query": "Turn our ecosystem map into a staged category roadmap with kill criteria.", "should_trigger": true },
{ "query": "Leadership wants MSPs, SIs and a marketplace motion all this year. Sequence them against our capacity.", "should_trigger": true },
{ "query": "partner mix roadmap for FY28", "should_trigger": true },
{ "query": "Should we do ISV integrations or an agency channel first?", "should_trigger": true },
{ "query": "what partner types should come next for us", "should_trigger": true },
{ "query": "We only have affiliates today. What do we add after that?", "should_trigger": true },
{ "query": "Sequence our partner program expansion honestly, we can't do all of it.", "should_trigger": true },
{ "query": "How do I decide which channel type to stand up next?", "should_trigger": true },
{ "query": "Our DTC brand does creators and affiliates. What channel comes after those?", "should_trigger": true },
{ "query": "I need a staged plan for growing our partner ecosystem over three years.", "should_trigger": true },
{ "query": "Which of ISV, MSP, SI and reseller should we open first?", "should_trigger": true },
{ "query": "We're adding partner types too fast. Help me put an order on it.", "should_trigger": true },
{ "query": "partner category planning", "should_trigger": true },
{ "query": "What should stage 1 of our partner roadmap actually be?", "should_trigger": true },
{ "query": "Can you help me figure out the right partner mix and the order to build it in?", "should_trigger": true },
{ "query": "Board wants four new partner channels next year. Is that realistic?", "should_trigger": true },
{ "query": "We've proven referral. What's next?", "should_trigger": true },
{ "query": "Draft a partner expansion plan with readiness gates and kill criteria.", "should_trigger": true },
{ "query": "How many partner types can two partner managers actually carry at once?", "should_trigger": true },
{ "query": "Give me the order to launch reseller, MSP and marketplace in.", "should_trigger": true },
{ "query": "Our integration partners work well. Should we go to a cloud marketplace next or add agencies?", "should_trigger": true },
{ "query": "What's the honest sequence for building out our channel over 36 months?", "should_trigger": true },
{ "query": "we want to expand the ecosystem, where do we start", "should_trigger": true },
{ "query": "Which partner type fills the gap in our coverage first?", "should_trigger": true },
{ "query": "Plan our route-to-market partner additions stage by stage.", "should_trigger": true },
{ "query": "Is it too early for us to add a reseller channel?", "should_trigger": true },
{ "query": "Should we launch a creator program or go into wholesale first?", "should_trigger": true },
{ "query": "My exec team wants an SI program. Are we ready, and if not what comes before it?", "should_trigger": true },
{ "query": "How do we avoid adding too many partner types at once?", "should_trigger": true },
{ "query": "We have five partner categories and sourced revenue is flat. Add a sixth or stop?", "should_trigger": true },
{ "query": "Put gates on our partner channel launch plan.", "should_trigger": true },
{ "query": "partner ecosystem expansion roadmap", "should_trigger": true },
{ "query": "Help me build the case for which partner channel to fund next.", "should_trigger": true },
{ "query": "I inherited a partner program with three motions and no plan. What order do we build from here?", "should_trigger": true },
{ "query": "Sequencing question: agencies or marketplaces next?", "should_trigger": true },
{ "query": "write the kill criteria for each stage of our partner expansion", "should_trigger": true },
{ "query": "What partner types should a $10M ARR B2B SaaS company add, and when?", "should_trigger": true },
{ "query": "give me a phased plan for our channel buildout", "should_trigger": true },
{ "query": "We keep signing new partner types and nothing activates. What should we have sequenced differently?", "should_trigger": true },
{ "query": "Which bucket of our partner coverage is empty, and what do we add to fill it?", "should_trigger": true },
{ "query": "Design the stage gates for launching a new partner category.", "should_trigger": true },
{ "query": "when should we open an MSP motion", "should_trigger": true },
{ "query": "Our wholesale team wants shelf space and marketing wants creators. Which one first?", "should_trigger": true },
{ "query": "Roadmap for adding new routes to market over the next two years.", "should_trigger": true },
{ "query": "Do we widen the affiliate program or start technology integrations?", "should_trigger": true },
{ "query": "We need a defensible order for the partner channels we're launching.", "should_trigger": true },
{ "query": "Help me say no to three of the five partner types leadership listed.", "should_trigger": true },
{ "query": "partner channel sequencing plan", "should_trigger": true },
{ "query": "What comes after a proven referral motion?", "should_trigger": true },
{ "query": "Build a category-level expansion plan for our partner org.", "should_trigger": true },
{ "query": "Our product is fully self-serve. Which partner types even make sense for us to add?", "should_trigger": true },
{ "query": "How should we phase adding a reseller category given our capacity?", "should_trigger": true },
{ "query": "Which new partner category gives the most back for the effort we can actually spend?", "should_trigger": true },
{ "query": "I want a roadmap that says what we're NOT adding, with triggers to revisit each one.", "should_trigger": true },
{ "query": "Is our partner motion mature enough to add another category?", "should_trigger": true },
{ "query": "minimum viable ecosystem for our partner program, are we there yet", "should_trigger": true },
{ "query": "Rank partner categories by value per unit of effort for a team of two.", "should_trigger": true },
{ "query": "we sell only through implementers, does that change which partner types we add first?", "should_trigger": true },
{ "query": "Plan the next three partner categories and what each one has to prove before the next opens.", "should_trigger": true },
{ "query": "Our CRO wants the same channel mix he had at his last company. Sanity-check that order for our product.", "should_trigger": true },
{ "query": "Show me a staged partner roadmap where each stage has to earn the next.", "should_trigger": true },
{ "query": "which partner categories should we deliberately not build this year", "should_trigger": true },
{ "query": "Should licensing or co-branding come before wholesale for our brand?", "should_trigger": true },
{ "query": "We did a marketplace listing and it did nothing. What should we have built first?", "should_trigger": true },
{ "query": "Map out partner types by time-to-first-revenue and effort so we can pick one.", "should_trigger": true },
{ "query": "Give me the order and the gates for expanding beyond our current two partner types.", "should_trigger": true },
{ "query": "Our partner-sourced pipeline stalled after we added a third category. Keep going or consolidate?", "should_trigger": true },
{ "query": "Map every partner type we work with today and show me the coverage gaps.", "should_trigger": false },
{ "query": "Build a partner landscape inventory with an account overlap analysis.", "should_trigger": false },
{ "query": "Score our current partners on a weighted scorecard and put them on a 2x2.", "should_trigger": false },
{ "query": "Who is in our ecosystem today, and which journey stages do they already cover?", "should_trigger": false },
{ "query": "Rank Accenture, Deloitte and Slalom by expected value and tell me which to pursue.", "should_trigger": false },
{ "query": "We have nine named alliance candidates. Which three deserve our capacity?", "should_trigger": false },
{ "query": "Write a go/no-go memo on the Snowflake partnership.", "should_trigger": false },
{ "query": "Score these five named technology partners on a weighted alliance scorecard.", "should_trigger": false },
{ "query": "We have no partner program at all. Design one from scratch.", "should_trigger": false },
{ "query": "How many tiers should our new channel program have, and what benefits go in each?", "should_trigger": false },
{ "query": "Are we even ready to start a partner program?", "should_trigger": false },
{ "query": "Write the partner value proposition for our new channel program.", "should_trigger": false },
{ "query": "Redesign our silver/gold/platinum tiers and the promotion rules.", "should_trigger": false },
{ "query": "Our tiers are compressed, everyone is gold. Fix the qualification criteria.", "should_trigger": false },
{ "query": "What should the demotion rules be for our partner tiers?", "should_trigger": false },
{ "query": "Should we list on AWS Marketplace, Azure, or both, and what ROI should we expect?", "should_trigger": false },
{ "query": "Is a Shopify app store listing worth the engineering work?", "should_trigger": false },
{ "query": "When should we delist from a marketplace we're already on?", "should_trigger": false },
{ "query": "Model the P&L of our relationship with Northgate Systems. Is it worth keeping?", "should_trigger": false },
{ "query": "What's our partner CAC versus direct CAC for this one reseller?", "should_trigger": false },
{ "query": "Build the unit economics for a single OEM deal.", "should_trigger": false },
{ "query": "Two partners are claiming the same account. Write the rules of engagement.", "should_trigger": false },
{ "query": "Our direct reps are fighting the channel on renewals. Draft an escalation ladder.", "should_trigger": false },
{ "query": "Write a MAP policy for our retail partners.", "should_trigger": false },
{ "query": "Design our deal registration policy: protection window, approval SLA, rejection grounds.", "should_trigger": false },
{ "query": "How do we split deal credit between partner-sourced and partner-influenced?", "should_trigger": false },
{ "query": "Set up comp neutrality so reps don't lose money on partner-assisted deals.", "should_trigger": false },
{ "query": "Build the certification curriculum for our reseller partners.", "should_trigger": false },
{ "query": "What training does a partner need before their first deal?", "should_trigger": false },
{ "query": "Design the partner onboarding curriculum and its refresh cadence.", "should_trigger": false },
{ "query": "Build a per-partner scorecard and a QBR structure.", "should_trigger": false },
{ "query": "One of our resellers has missed target twice. What's the underperformance ladder?", "should_trigger": false },
{ "query": "Plan a joint webinar and co-marketing campaign with Vantiva.", "should_trigger": false },
{ "query": "Write the better-together story for our integration launch with Datameer.", "should_trigger": false },
{ "query": "What commission rate should we pay affiliates, and should it be recurring?", "should_trigger": false },
{ "query": "Draft our affiliate program terms and conditions.", "should_trigger": false },
{ "query": "Audit this month's affiliate payout run before we disburse it.", "should_trigger": false },
{ "query": "Our affiliate traffic looks fake. Build the fraud detection rules.", "should_trigger": false },
{ "query": "Write the welcome email series for newly approved affiliates.", "should_trigger": false },
{ "query": "Which metrics belong on our affiliate performance dashboard?", "should_trigger": false },
{ "query": "Write outreach emails to recruit coupon sites into our affiliate program.", "should_trigger": false },
{ "query": "Does this sponsored TikTok caption meet FTC disclosure rules?", "should_trigger": false },
{ "query": "Define the criteria to shortlist creators for our spring campaign.", "should_trigger": false },
{ "query": "Write the campaign brief for the creator we just signed.", "should_trigger": false },
{ "query": "Prepare me to negotiate usage rights and exclusivity with a creator's manager.", "should_trigger": false },
{ "query": "Draft a DM pitching a partnership to one creator we've already vetted.", "should_trigger": false },
{ "query": "How do we measure one influencer campaign before it launches?", "should_trigger": false },
{ "query": "How big should our refer-a-friend reward be against LTV and CAC?", "should_trigger": false },
{ "query": "People are creating fake accounts to farm referral rewards. What guardrails do we need?", "should_trigger": false },
{ "query": "How do I break into partnerships from a sales role?", "should_trigger": false },
{ "query": "Write the job posting and interview loop for a channel partner manager.", "should_trigger": false },
{ "query": "I'm starting partner work at a new company. Which skill should I use?", "should_trigger": false },
{ "query": "How many partner managers should we hire, and at what levels?", "should_trigger": false },
{ "query": "Which PRM vendor should we buy?", "should_trigger": false },
{ "query": "Design our reseller margin structure and discount bands.", "should_trigger": false },
{ "query": "Build our international distributor contract template.", "should_trigger": false },
{ "query": "Plan our expansion into the German and Nordic markets next year.", "should_trigger": false },
{ "query": "Which product lines should we launch next, and in what order?", "should_trigger": false },
{ "query": "Build a developer platform strategy for our public API ecosystem.", "should_trigger": false },
{ "query": "Which SDK languages should we support next?", "should_trigger": false },
{ "query": "Design an app marketplace so third-party developers can publish on our platform.", "should_trigger": false },
{ "query": "Which integrations should our product team build next?", "should_trigger": false },
{ "query": "Plan the expansion of our customer success org into EMEA.", "should_trigger": false },
{ "query": "What's the sequence for rolling out our new pricing across segments?", "should_trigger": false },
{ "query": "Which customer segments should we expand into next, and in what order?", "should_trigger": false },
{ "query": "Which conferences should we sponsor next year?", "should_trigger": false },
{ "query": "Expand our content marketing into three new topic clusters. What order?", "should_trigger": false },
{ "query": "Plan our go-to-market expansion into the public sector vertical.", "should_trigger": false },
{ "query": "Sequence the rollout of our new partner portal features.", "should_trigger": false },
{ "query": "Which open-source ecosystem projects should we contribute to next?", "should_trigger": false }
]
}
references/expansion-workshops.md›
# Expansion Workshops
The profession brainstorms this decision in facilitated formats, not free ideation. The five below are ordered by efficiency - candidates surfaced and assumptions killed per hour of the room's time - highest ratio first.
Every format follows the same facilitation discipline:
- One question per message.
- 2-3 options with trade-offs and a recommendation at every real choice.
- Assumptions named aloud.
- Sections validated one by one.
- Explicit approval before finalizing.
- cost (heaviest first): `business-design workshop > partner persona == account mapping > value blueprint > map re-read`
- value (biggest payoff first): `business-design workshop > value blueprint > account mapping > partner persona > map re-read`
- efficiency (run first): `map re-read > value blueprint > account mapping > partner persona > business-design workshop`
`partner persona == account mapping` on cost: both need a day in the room plus a week of evidence gathered outside it - partner-principal interviews, or account-overlap data - and neither can be run from what the ecosystem map already holds.
Run the top of the order first and stop when the candidate pool stops changing; the lower formats earn their cost only when the cheaper ones leave a real question open. The order is a default: it shifts with what data already exists (no ecosystem map drops the first format to last, no overlap data does the same to account mapping) and with who is in the room.
**What this order starves: the business-design workshop.** Highest value, highest cost, last every round. Promote it when the answer would change how a GTM function operates day to day - comp plans, CS ownership, data sharing between teams. No cheaper format can get that decision made, and a category the organization cannot operate will sign partners and never activate them.
## 1. Ecosystem-map re-read
Input: an existing ecosystem map (from the mapping sibling). Near-zero cost when the map exists - do not rebuild it, interrogate it for expansion candidates.
- Which journey buckets (influence / transact / retain-and-deliver), segments, and geographies show no partner coverage at all?
- Which existing categories are saturated or redundant - where would more of the same type add noise, not reach?
- Which gaps did the map size largest, in accounts or revenue at stake?
- What did the map's closed-won review show third parties already doing uncompensated - the cheapest categories to formalize?
## 2. Value blueprint session (Adner)
Input: the value proposition and the actors it needs; no data gathering required. Half a day, and the only format that kills the ego-system trap before it costs a stage.
- Who must do what, in what order, for the customer to receive the promised value?
- Where is co-innovation risk - which candidate category only creates value if someone else's piece also ships?
- Where is adoption-chain risk - which actor in the chain gains nothing by participating, and what would change that?
- Redrawn from the customer's seat, are we central at all? (the ego-system check)
- What is the smallest configuration - the MVE - that proves value, and is it already proven with repeatable joint wins?
## 3. Account-mapping session
Input: account-overlap data with existing partners of any category - the mechanic popularized by the ecosystem-led-growth community and tools like Crossbeam (named here as sourced practice, not a requirement). Real evidence rather than room opinion, which is why it outranks the two formats below despite costing a week of data work when the overlap data is missing.
- Which existing partner category generates the most ecosystem-qualified leads per partner - evidence for deepening it before diversifying?
- Do overlap patterns show a category you lack repeatedly present in your best accounts - for example, an implementation partner visible in every large win you did not source?
- Where is overlap with existing categories already crowded - a warning that a new overlapping category buys conflict, not coverage?
If no overlap data exists, approximate from closed-won interviews and customer tech-stack evidence - never from assumption.
## 4. Partner persona / jobs-to-be-done session
Input: buyer knowledge and any partner interviews. A day in the room, plus the interviews.
- What job is the buyer hiring done - and which partner category does that job today (advises, implements, operates, bundles, resells)?
- For each candidate category, build the persona of the partner principal: their KPIs, margins, constraints - what would make your program worth their shelf space?
- Which candidate category's business model is structurally strained right now (the resell-margin collapse is the live case), and which bucket is where its economics moved?
- Arena scan (boundaryless-style): in which adjacent arenas is the customer's job being done by ecosystems you are absent from?
## 5. Ecosystem business-design workshop (GoToEcosystem-style, per Allan Adler / Digital Bridge Partners)
Input: the GTM organization itself - multi-day, cross-functional, and a standing job to organize. A maturity lens applied per team, asking whether the company can operate the category it wants, not just sign it.
- For each GTM function (sales, marketing, customer success, product), how would this candidate category change its daily motion - and is that function ready?
- Which cluster of partner types would serve one common ICP together ("selling in clusters"), rather than each category running as a silo?
- What must be reorganized - process, compensation, data sharing - before the category can activate, not just sign?
references/kpis-and-benchmarks.md›
# KPIs and Benchmarks
## The vendor-bias warning - read first
Nearly every published number in this field comes from a party that profits when it looks good:
- PRM and ecosystem-platform vendors.
- Analyst studies commissioned by an interested party.
- Consultancies quoting their own engagement models.
The only independent evidence base is academic alliance-portfolio research, which measures firm performance rather than program vanity metrics. Therefore, **every benchmark quoted in a roadmap carries a provenance flag**, and vendor figures are directional context, never proof.
Flag taxonomy:
- **ACADEMIC**: peer-reviewed, independent.
- **ANALYST**: research firm, often paywalled.
- **VENDOR**: published by a tool/platform vendor.
- **VENDOR-COMMISSIONED**: analyst research paid for by an interested party.
- **CONSULTANT**: a consultancy's own models.
- **MIXED**: aggregated across the above.
## Leading indicators per added category
Judge a new category by cohort, early. Instrument these in efficiency order - gate decisions settled per unit of measurement effort and waiting time, highest ratio first - and stop when the gate can be read:
1. **Partner activation rate** - signed partners closing at least one deal in year one. One query against data you already hold, and it settles the recruit-ahead-of-enablement question by itself.
2. **Time-to-first-deal per partner**, against the expected window for that category. Same data, and it separates a slow category from a failing one.
3. **Partner-sourced and partner-influenced pipeline** (B2B); incremental attributed revenue (B2C). Needs attribution built and a quarter of accumulation, and it is what the funding conversation runs on.
4. **Enablement/certification completion** inside the onboarding window - whether the cohort did the motion.
5. **Deal-registration volume** - whether deals are surfacing before they collide.
6. **Category share of new revenue**, once the category matures.
7. **Partner NPS** for the pilot cohort - opinion, useful for diagnosing a failing cohort, never sufficient to open a gate.
- cost (heaviest first): `category share of new revenue > partner-sourced pipeline > partner NPS > activation rate > time-to-first-deal > enablement completion == deal-registration volume`
- value (biggest payoff first): `category share of new revenue > partner-sourced pipeline > activation rate > time-to-first-deal > enablement completion > deal-registration volume > partner NPS`
- efficiency (instrument first): `activation rate > time-to-first-deal > partner-sourced pipeline > enablement completion == deal-registration volume > category share of new revenue > partner NPS`
`enablement completion == deal-registration volume` on both lines: each falls out of a system the category's prerequisites forced you to build anyway (the certification track, deal registration), and each is read as a report rather than measured. Neither proves revenue - both only report whether the cohort is running the motion.
**What this order starves: category share of new revenue.** The number leadership actually judges the program by, ranked second-to-last because it needs a year of cohorts before it says anything. Promote it whenever the review is a funding or board decision rather than a stage gate - there, no cheaper indicator is accepted as the answer.
This order is a default; it shifts with the motion and with who reads it. A category with no deal registration has no line 5, a B2C affiliate cohort reads incremental attributed revenue where a B2B one reads sourced pipeline, and a user whose CRM already reports partner attribution should promote line 3 to the top because its cost is already paid.
## Published numbers, with flags
| Metric | Figure | Source + flag |
| --------------------------------------- | --------------------------------------------------------------------------------------------- | ------------------------------------------------------------ |
| Healthy partner activation | 30-50%; below 20% when unmanaged | Introw - VENDOR |
| Onboarding window before likely failure | 90-120 days | Forrester Partner Onboarding Framework - ANALYST (paywalled) |
| Ramp analog for a partner seller | SMB ~3 mo; mid-market 5-7 mo; enterprise 9-12 mo | RevOps sources incl. The Bridge Group - MIXED |
| Orchestration-investment threshold | Partner-sourced 5-10% of new ARR | Forecastable - CONSULTANT |
| Ecosystem revenue multiplier | 5x-10x on annual partner spend over 3-4 yrs; GSI 6x-12x | Alexander Group's own models - CONSULTANT |
| Partner-services multiplier | $7.13 services per $1 platform sold (services-heavy partners) vs $1.26 (resell-focused) | Canalys for AWS - VENDOR-COMMISSIONED |
| Co-sell effects | +51% average revenue growth, 65% higher close rates, 54% larger deals for frequent co-sellers | Canalys for AWS, 2024 - VENDOR-COMMISSIONED |
| Mature programs' revenue share | 28% of overall company revenue through partnerships | Forrester - ANALYST |
| MDF sizing | 2-6% of channel revenue; 25-50% cost-share per activity; 40-60% utilization | MIXED |
## Known-unstable statistics - never quote as settled fact
- **"Buyers spend 68% of their journey digitally / 71% reach vendor selection digital-only."** Derives from a 2013 SiriusDecisions statistic that Forrester itself later publicly flagged as a myth and revised to "70-80% self-directed". Vendor-analyst assertion; often used to justify adding influence-category partners early - the direction may hold, the precision does not.
- **"95% of Microsoft revenue flows through partners."** Traceable to a real primary source: Microsoft's own corporate blog (February 2019) quotes this figure directly from Satya Nadella and Amy Hood's Q2 FY2019 earnings-call commentary ("95 percent of Microsoft's commercial revenue flows directly through our partner ecosystem"), and then-CVP Gavriella Schuster repeated it in her own remarks. Genuinely sourced to Microsoft rather than an untraceable round number - but still one company's self-reported commercial-revenue framing from 2019, not an audited or updated figure, so present it as Microsoft's own claim rather than a current industry constant.
- **Reseller-profitability figures drift.** The older "~25% of the channel struggles to break even" has been superseded by Omdia's 378-partner survey: 58% expecting profit decline, 51% forecasting double-digit drops. Use the current figures and note the drift.
## Using the numbers
- Set gate thresholds from the user's own base rates wherever any exist (current categories' activation, time-to-first-deal); use published figures only to sanity-check, never to substitute.
- When a plan's viability depends on a flagged number, list it in the roadmap's Open Questions with what would verify it.
- Refuse the comparison when a vendor benchmark and the user's cohort measure different things - signed partners vs active partners is the classic mismatch.
- Any number appearing in the roadmap without a flag fails the Pass Threshold; fix it before shipping.
references/partner-category-economics.md›
# Partner Category Economics
The per-category facts sequencing runs on, ordered by efficiency: value returned per unit of effort, highest ratio first. All ranges are directional - the underlying figures mix vendor and consultant sources and are not independently benchmarked; carry that flag when quoting them.
Effort is what the team spends - build time, standing coordination, headcount held, reversibility - never a currency amount. Readiness prerequisites per category are scored at a separate workflow step and are not repeated here.
## B2B categories
Efficiency order (open first): **tech/ISV > referral/affiliate > agency > marketplace > MSP == SI/GSI > reseller/VAR**
| Category | Commercial model | What it buys | Time to first revenue | Effort to stand up | Standing load | Compliance exposure |
| ------------------------------------- | ----------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- | --------------------- | --------------------------------------------------------------------------------------------------------- | ---------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------- |
| Technology / ISV | Rev-share on integrations, marketplace referral fees, co-sell credits | Closes whole-product gaps named in lost deals, and compounds - the integration surface makes a later marketplace stage cheap | Medium | A quarter - integration surface, docs, sandbox, built once and reused by every partner | An hour per partner once the docs hold | Low - standard agreement; a security review only if a marketplace stage follows |
| Referral / affiliate | Fixed fee or referral commission; simple contract | Top-of-funnel pipeline from people already advising your buyer; no delivery capacity, no coverage | Fastest | A week - tracking, payout, simple terms | Near-zero per partner | Low - disclosure plus a standard agreement; ends when the agreement ends |
| Agency | Referral fee or margin, plus their own services around your product | Coverage of buyers who arrive with an advisor, and delivery capacity you never staff | Medium | A quarter - certification content, co-marketing assets, lead sharing | A week per cohort | Low - disclosure plus a standard agreement |
| Marketplace (hyperscaler / app store) | Platform take rate on transactions, or listing plus co-sell | The procurement path itself: committed spend, an existing vendor relationship, a buying process you skip | Medium | A quarter - listing readiness, billing integration, security review | A week per quarter of listing ops and policy changes | High - platform policy, security review, price parity; delisting and take-rate changes are the platform's call, not yours |
| MSP | Recurring, bundled managed-service revenue | Retained revenue plus an operating layer you never staff - the bucket the surviving channel margin moved into | Slow | A standing job - multi-tenant administration, monitoring/billing hooks, tiered support | A standing job | Medium - contracting, support SLAs, data handling inside a partner-operated tenant |
| SI / GSI, consultancy | Services-first; co-sell credits, SOWs - partner services revenue often 2-5x license | The enterprise route to market itself, in segments that buy only through implementers | Slowest | A standing job - implementation methodology, certified-practitioner program, executive sponsorship | A standing job, named-account by named-account | Medium - MSAs, sub-contracting terms, named-account alignment |
| Reseller / VAR | Margin/discount on booked revenue | Booked revenue through a procurement relationship you already lost - at a margin that has structurally collapsed | Medium-slow | A quarter or more - deal registration, margin structure, sales and technical certification, support model | A standing job | Medium-high - channel pricing discipline and margin committed contractually before you learn the volume |
The axes disagree, so read them separately:
- cost (heaviest first): `SI/GSI > MSP > reseller/VAR > tech/ISV == marketplace == agency > referral/affiliate`
- value (biggest payoff first): `SI/GSI > MSP > tech/ISV > marketplace > reseller/VAR > agency > referral/affiliate`
- compliance cost (most exposure first): `marketplace > reseller/VAR > MSP > SI/GSI > tech/ISV > agency == referral/affiliate`
- efficiency (open first): `tech/ISV > referral/affiliate > agency > marketplace > MSP == SI/GSI > reseller/VAR`
The ties, and why they are genuine:
- `MSP == SI/GSI` on efficiency: both buy the durable retain-and-deliver economics, both cost a standing job to run, and neither converts ahead of the other unless the buyer route already goes through an implementer - the same condition promotes both.
- `tech/ISV == marketplace == agency` on cost: each is roughly a quarter of build reused across the whole category (integration surface, listing plus security review, certification track), and none of them adds a per-partner headcount.
- `agency == referral/affiliate` on compliance cost: both carry only disclosure and a standard agreement, and both unwind by ending that agreement - no platform policy, no committed margin, nothing to claw back.
**Why tech/ISV leads and referral does not.** Referral is the cheapest category, not the most efficient one: it buys pipeline only, and nothing it builds makes the next category cheaper. The integration surface an ISV motion forces you to build is reused by the marketplace stage, by agency implementers, and by every MSP that operates the product later. Cheapest-first is the practitioner-asserted prior - it is the cost line above, not the recommendation.
**What this order starves: SI/GSI and MSP.** Highest value in the table, highest effort, so a ratio ranking pushes both to the back every single round - the documented pattern where programs over-serve cheap influence categories and starve technical and services enablement.
Promote them anyway when any of these holds:
- The buyer purchases only through implementers (enterprise, regulated, public sector).
- Implementation capacity is named in three or more lost deals.
- The roadmap needs retained rather than new revenue.
The services multiple is the argument to eventually get there - Canalys' AWS-commissioned studies report $7.13 of partner-services revenue per $1 of platform sold for services-heavy "Expert" partners vs $1.26 for resell-focused partners (VENDOR-COMMISSIONED) - sourced from a party with an interest in that conclusion.
**Reseller/VAR sits last for a structural reason.** McBain/Omdia document point-of-sale reseller margins falling from ~20% to ~3% as vendors shifted to subscription; in a 378-partner survey, 58% expected profit decline, 51% forecasting double-digit drops (ANALYST). The category still buys real booked revenue, but at an effort and margin commitment the current economics no longer repay for most vendors.
## B2C categories
Efficiency order (open first): **creator/influencer > affiliate > marketplace > retail/wholesale > licensing == co-branding > distributor**
| Category | Commercial model | What it buys | Time to first revenue | Effort to stand up | Standing load | Compliance exposure |
| -------------------- | ------------------------------------------------------- | ------------------------------------------------------------------------------------------------ | --------------------- | ------------------------------------------------------------------------------------ | ------------------------------------------------- | --------------------------------------------------------------------------------------------------- |
| Creator / influencer | Flat fee, commission, or hybrid; gifting at the low end | Demand created before intent exists, plus content assets reusable in paid and owned channels | Fast | A week - brief and asset kit, usage-rights terms | A week per campaign cohort | Medium - disclosure rules are enforcement-backed, and the liability is yours as well as theirs |
| Affiliate | Commission on tracked sales (often 10-15% for creators) | Conversion of demand that mostly already existed - cheap, and rarely incremental | Fastest | A week - link/code attribution, payout process | Near-zero per partner | Low - disclosure plus a standard agreement |
| Marketplace | Platform take rate | Shelf reach with no trade negotiation, and a buying flow customers already trust | Medium | A quarter - listing operations, fulfillment integration, price-parity policy | A week per quarter of listing ops | High - platform policy and price parity; the platform can change the take rate or delist you |
| Retail / wholesale | Wholesale margin | Physical shelf reach and the volume that comes with it - the biggest distribution step available | Medium-slow | A standing job - trade pricing, fulfillment at retail cadence, merchandising support | A standing job | High - trade terms, promotional commitments, and margin fixed before the sell-through is known |
| Licensing | Royalty | Royalty revenue with no COGS and no fulfillment, on a brand you already own | Slow | A quarter of legal and brand work, then near-zero ongoing | An hour per quarter, plus quality-control reviews | High - brand guidelines, quality control, and long terms that are the hardest thing here to reverse |
| Co-branding | Shared campaign economics | Borrowed brand equity and a reach spike for one campaign; nothing compounds after it ends | Medium | A quarter - brand-fit vetting, joint approval process | A week per campaign | Medium - trademark and joint-approval terms, ending with the campaign |
| Distributor | Distributor margin layered above retail | Reach into retail you cannot serve directly, at a second margin layer stacked on the first | Slow | A standing job - volume production, channel pricing discipline | A standing job | High - channel pricing discipline and terms that bind what you can do direct |
- cost (heaviest first): `distributor > retail/wholesale > licensing > marketplace == co-branding > creator/influencer > affiliate`
- value (biggest payoff first): `retail/wholesale > distributor > licensing > marketplace > creator/influencer > co-branding > affiliate`
- compliance cost (most exposure first): `licensing > retail/wholesale > distributor > marketplace > creator/influencer > co-branding == affiliate`
- efficiency (open first): `creator/influencer > affiliate > marketplace > retail/wholesale > licensing == co-branding > distributor`
The ties:
- `licensing == co-branding` on efficiency: both buy borrowed brand equity rather than distribution, both need an approval and quality-control machine you do not have unless legal is already in-house, and neither makes the next category cheaper.
- `marketplace == co-branding` on cost: both are a quarter of setup work (listing and fulfillment integration; brand-fit vetting and a joint approval process) with no standing headcount attached.
- `co-branding == affiliate` on compliance cost: both carry a disclosure obligation and a trademark clause, and both end when the campaign or the agreement does.
**What this order starves: retail/wholesale and distributor.** The largest distribution steps a consumer brand can take, ranked last because both are a standing job with margin committed up front. Promote them when the category the buyer shops in is physical-first, when a retail buyer has already asked, or when DTC growth has flattened and the roadmap needs reach rather than efficiency.
## Trifurcation buckets (candidate-pool axis)
Every category above sits in one of McBain/Canalys' three journey buckets - use them to spot the empty bucket, never as a sequence:
- **Influence**: referral, affiliate, creators/influencers, agencies (advisory side), analysts.
- **Transact**: resellers/VARs, distributors, marketplaces, retail/wholesale.
- **Retain-and-deliver**: MSPs, SIs/GSIs, agencies (delivery side), licensing operations.
## Re-ranking against this user
Both orders above are defaults, not laws. They shift with context and with who executes them, so re-rank before presenting anything:
- **Delete, never demote.** A category the user's constraints rule out leaves the candidate pool entirely, with the reason recorded in the roadmap's `Not now` line and a re-entry trigger. Examples: no channel margin room deletes reseller/VAR, no multi-tenant administration on the roadmap deletes MSP, no in-house legal capacity deletes licensing. A ruled-out category parked at the bottom of a ranked list reappears later as scope.
- **An existing services arm promotes SI/GSI to the top**: the standing job that makes it expensive is already staffed and already carries the methodology.
- **A product that needs no implementation demotes SI/GSI and agency below marketplace**: there is no services gap for either to fill.
- **A partner team of one deletes every standing-job category** - MSP, SI/GSI, reseller, retail, distributor - until headcount changes; a single manager cannot carry one of them and the proven motion at the same time.
- **The buyer's route outranks the effort ranking.** If the buyer only purchases through implementers, an SI category comes early despite its load - the sequence serves the route to market.
- **The current base changes the adjacency.** The cheapest next category depends on what is already proven: an ISV motion reaches a marketplace stage cheaply, while the same stage is premature for a referral-only base.
- **A whole-product gap named in lost deals** pulls its filling category forward regardless of where the efficiency line puts it.
- **A category's structural economics can shift.** The reseller margin collapse is the live example: an order that made sense a decade ago can point at a bucket whose economics have moved.
references/readiness-and-capacity-gates.md›
# Readiness and Capacity Gates
A category enters the sequence only when its prerequisites exist and capacity can carry it. The rule practitioners repeat: gate on capacity, not enthusiasm.
## Readiness prerequisites
Named consistently across independent practitioner sources:
| Prerequisite | Needed for | Pass test |
| ------------------------------------- | --------------------------------------- | ------------------------------------------------------------------------------------------ |
| Tracking/attribution + payout process | Referral, affiliate, creator | A referred deal can be traced and paid without manual archaeology |
| API maturity + multi-tenancy | ISV/tech, MSP | A third party can build and operate against the product without your engineers in the loop |
| Certification / training content | Reseller, SI, MSP, agency | A partner seller or implementer can pass a real test on the product |
| Deal registration | Reseller, SI, any co-sell motion | Conflict-relevant deals are visible before they collide |
| Margin structure | Reseller, distributor, retail/wholesale | The category's margin leaves unit economics intact at plan volume |
| Support model | Reseller, MSP | Second-line support exists for partner-delivered customers |
| Contracting | All formal categories | A standard agreement exists; legal is not a per-deal bottleneck |
The recurring practitioner rule behind the prerequisite list: a resell-type category is ready only "when your product is mature enough that a third party can sell and support it without heavy involvement". Readiness gates a category; it does not rank one - ranking is a separate step, run on per-category economics.
## Capacity math
- Fix partner-manager-to-partner ratios per category before recruiting - one manager carries far fewer SIs than affiliates.
- Count enablement build (content, certification, portal workflows) as project capacity with an owner, not as spare time.
- Defer PRM/portal tooling until a motion produces the workflow that needs it; bought early, "the platform sits as compliance theater" (Forecastable - CONSULTANT).
- The documented cost of skipping this: activation runs 30-50% when a category is managed and drops below 20% unmanaged (Introw - VENDOR). "High-intensity recruitment followed by low-intensity enablement" is reported as the single most common SaaS partner-program failure mode (attributed to Forrester's 2025 Channel Benchmark via a vendor blog - treat the exact citation cautiously).
## The business case
Fund each stage as a capital request, not a headcount ask:
- **Total loaded cost**: salaries, enablement build, incentives, tooling, program spend. Forecastable's phrasing: present it "in total cost, not headcount" - "you are asking for capital, not enthusiasm."
- **Funding mechanisms**, ranked by capacity bought per unit of approval effort and commitment:
- efficiency (fund from first): `program budget > MDF > headcount`
- cost (heaviest first): `headcount > MDF > program budget` - headcount is a standing commitment through the slowest approval path and the hardest to reverse; MDF costs per-activity claim and proof-of-performance administration; program budget is one approval and unwinds at the end of the period.
- value (biggest payoff first): `headcount > program budget > MDF` - only headcount buys the standing capacity a gate reads, while MDF buys activity of which 40-60% is typically used at all.
- compliance cost (most exposure first): `MDF > headcount > program budget` - MDF claims carry contractual proof-of-performance and audit rights; program budget carries none.
- Typical MDF sizing: 2-6% of channel revenue, 25-50% cost-share per activity, industry utilization only 40-60% and over half sometimes unused (MIXED vendor/analyst figures - flag when quoted).
- **What this order starves: headcount** - the highest-value mechanism, ranked last because it is the slowest and least reversible ask. Promote it whenever the stage opens a standing-job category (MSP, SI/GSI, reseller): program budget and MDF buy activity there, never the capacity the category runs on, and a standing job funded from a one-period budget fails at the second review.
- **A defensible return model**: vendor/consultant multipliers (5x-10x on annual partner spend over 3-4 years; GSI 6x-12x, per Alexander Group's own models - CONSULTANT) are framing, never proof.
- **Cross-functional sign-off**: product (readiness items on the roadmap), sales (conflict rules), finance (margin), legal (contracting). Shared KPIs buy the buy-in.
## Stage-gate design
Per stage:
- **Pilot**: a deliberately small cohort of the new category - a handful of partners, named target accounts - with full enablement given to that cohort. A pilot starved of enablement tests nothing.
- **Gate threshold - partner-market fit**: repeatable joint wins, not a first win. Multiple partners in the cohort independently closing, a joint value pitch that repeats, a documented repeatable process. When momentum shows, it typically shows within two to three quarters (CONSULTANT).
- **Evidence read at the gate**: cohort activation rate, time-to-first-deal against the category's expected window, sourced pipeline, enablement completion inside the onboarding window (Forrester's onboarding framework: failure to enable within 90-120 days makes the partnership "likely to fail" - ANALYST, paywalled).
- **Gate outcomes vocabulary**: pass-and-scale, extend-pilot (once, with a named fix), re-scope, kill.
## Kill criteria
Less formalized in the published literature than product stage-gates - which is exactly why the roadmap must write them down before launch:
- Activation below ~20% in the pilot cohort → stop recruiting; fix enablement or kill.
- No partner-sourced pipeline within the category's expected time-to-first-deal window → re-scope or kill.
- Thresholds missed two consecutive quarters → kill or re-scope; never stack more enablement spend onto a failing motion.
- Roadmap-level inverted-U trigger: coordination cost rising while total sourced revenue flattens → freeze new stages and deepen existing categories instead.
Every kill leaves a record: what was tested, what the evidence showed, and the re-entry trigger that would reopen the category.
references/sequencing-frameworks.md›
# Sequencing Frameworks
What genuinely fits this decision, what partially fits, and what to reject. No peer-reviewed optimal sequence exists; the honest method assembles the pieces below.
Deliberately unranked: these five are not competing options for one goal but complements used together in a single pass.
- The trifurcation builds the pool.
- The value blueprint decides order.
- Capacity math caps it.
- The inverted-U says when to stop.
- Moore says which gap a segment demands.
Ranking complements by value per unit of effort would be false precision, since dropping the "low-efficiency" one leaves a question none of the others answers. The rankings that do apply are built at their own workflow steps, each against its own axes: candidate categories, facilitation formats, and leading indicators.
The force-fits in the last section are not options in that pool at all - they are rejects kept only so a user who arrives with one gets a straight answer.
## The anchor: Adner - MVE, staged expansion, value blueprint
- **Minimum viable ecosystem (MVE)**: "the smallest configuration of activities (and partners) that can create enough evidence of value creation to attract new partners" (Adner). Every expansion conversation starts here: is the current configuration an MVE with repeatable joint wins? If not, stage 1 of the roadmap is finishing it, not adding.
- **Staged expansion**: "the order in which an additional partner or activity should be brought on board." Additions are deliberate and sequential; each stage builds on the evidence the previous one created.
- **Value blueprint**: map every actor and link required for the value proposition to materialize, drawn from the end customer backward. It surfaces the two risks that decide category order: **co-innovation risk** (a category only creates value if someone else's piece also ships) and **adoption-chain risk** (an actor in the chain has no incentive to participate). A candidate category whose blueprint shows unresolved adoption-chain risk gets sequenced later or fixed first, however attractive its economics.
- **The ego-system trap**: presuming your own centrality - drawing the blueprint with yourself at the center and assuming partners will orbit. The check: redraw from the customer's position and ask, for each candidate category, what that partner type gains by joining _now_. A category with no answer will sign logos and never activate.
- Cross-adoption evidence: BCG's ecosystem-design work explicitly builds on Adner's MVE - the framework has left academia and entered consulting practice.
Sources: _The Wide Lens_ (2012); _Winning the Right Game_ (2021) - ACADEMIC/PRACTITIONER.
## The restraint evidence: the alliance-portfolio inverted-U
Alliance-portfolio-diversity research (Jiang, Tao & Santoro 2010, _Strategic Management Journal_; de Leeuw, Lokshin & Duysters 2014; Lee, Kirkpatrick-Husk & Madhavan 2017 meta-analysis) finds that partner-type diversity has an **inverted-U** relationship with firm performance: adding categories helps up to a point, then coordination and governance costs dominate. This is the field's only independent, non-self-interested evidence.
It justifies restraint and sequencing; it prescribes no order. Practical use: the standing argument against "add them all", and the roadmap-level signal - coordination cost rising while sourced revenue flattens - that says stop adding and deepen. ACADEMIC.
## The taxonomy: McBain / Canalys channel trifurcation
- **Influence**: referral, affiliate, advocacy, creators, analysts.
- **Transact**: resellers, VARs, distributors, marketplaces.
- **Retain-and-deliver**: MSPs, SIs, agencies delivering outcomes.
A classification for finding the empty bucket - which journey stages carry no partner coverage - never a sequence. Use it to build the candidate pool, then sequence with the other tools. ANALYST.
## The partial fit: Moore - whole product and bowling alley
- **Whole product**: pragmatist buyers purchase the complete solution, and partner categories exist to close the gap between what you ship and what the segment needs. Genuinely useful for deciding _which_ category a segment demands: a gap named in three or more lost deals pulls its filling category forward.
- **Bowling-alley adjacency**: expansion works pin by pin, each conquest making the next cheaper. The adjacency logic transfers to categories - the cheapest next category is the one that reuses the most of what the current base already built (an ISV motion makes a marketplace stage cheap; a referral motion makes an agency-referral track cheap). The segment-marketing machinery around it does not transfer; take the adjacency logic, not the whole book.
## The working method: coverage/capacity math and stage-gate pilots
Unbranded but real - this is what practitioners actually do:
- Coverage-gap analysis (segment × geography × journey stage) to justify a category.
- Partner-manager-to-partner ratios and enablement bandwidth as the hard cap.
- Pilot cohorts of one to three partner types, gated on partner-market fit (repeatable joint wins) before any scale investment.
Alexander Group's published SaaS guidance follows this shape - activate new categories one at a time, "prove the model with a handful of named accounts", and extend only once the flywheel shows momentum, typically within two to three quarters (CONSULTANT - the firm's own modeled engagement figures, not independent evidence).
## Named force-fits - reject these
- **Ansoff matrix applied to channels.** Built for product/market growth options; "new channel = new quadrant" supplies a label, no sequencing logic, and there is no evidence of systematic partner-category use.
- **McKinsey three horizons.** Innovation-portfolio timing vocabulary ("core/adjacent/transformational partner bets") borrowed loosely; nothing in it says which category or when.
- **BCG growth-share matrix.** Portfolio-balance language retrofitted to partner mix; intuitive, but no documented channel methodology behind it.
- **Iansiti & Levien (keystone/niche/dominator).** Describes your _role_ in an ecosystem, not which partner types to add next; academically influential, rarely used by channel practitioners for this decision.
When a user arrives with one of these, keep its vocabulary if it helps the room communicate - then rebuild the actual decision on MVE + gates + capacity.
## A genuine open tension - say it, don't resolve it silently
Whether partnering is a day-one motion or a later-stage accelerant is genuinely contested across sources, and the resolution is category-dependent. Influence-bucket categories (referral, affiliate, creators) are cheap enough to run early, while transact- and deliver-bucket categories (reseller, MSP, SI) presuppose a proven product motion and real enablement infrastructure.
That is a statement about which categories are _available_ early, not a sequence - the per-category efficiency order decides what opens first among the available ones. Tell users this is a tension in the evidence rather than picking a side for them.
references/worked-examples.md›
# Worked Examples
Both examples are fictional companies; the mechanics are the point.
## Positive example - B2B SaaS, referral motion proven, 24-month roadmap
Scenario: a workflow-automation vendor with a referral/affiliate motion live for 18 months (40 partners, 45% activation, repeatable joint wins - the MVE is proven) and six ad-hoc, unmanaged ISV integrations. Capacity: 1.5 partner managers, an enablement writer at half-time, $275k program budget.
```
PARTNER-CATEGORY EXPANSION ROADMAP - WorkflowCo, FY planning
Current base : referral/affiliate proven (40 partners, 45% activation, repeatable joint
wins); 6 ad-hoc ISV integrations, unmanaged
Evidence posture: gate thresholds set from own referral-cohort base rates; every external
benchmark flagged (see kill criteria)
Candidate pool : tech/ISV (closes the integration gap named in 5 lost deals); agency
(buyer's implementers appeared in 3 of 10 closed-wons); marketplace
(would ride the ISV work); reseller/VAR (rejected below); MSP (deferred)
Capacity : 1.5 partner managers - one manager carries ~8 managed ISVs or ~4 agencies;
enablement budget covers one certification track per year
Sequence : Stage 1 (now): formalize tech/ISV - API, docs, sandbox exist; prerequisite
check passed
Stage 2 (gate-dependent): agency referral track - reuses the referral
tracking and payout infrastructure already built
Stage 3 (gate-dependent): marketplace listing - rides Stage 1 integrations
Gates : Stage 1 → 2: 8 managed ISV partners, ≥40% cohort activation (own base
rate), 3 partner-sourced deals within 2 quarters; owner: VP Partnerships
Stage 2 → 3: 5 agencies certified, 2 independently sourcing, joint pitch
repeatable across partners
Kill criteria : any stage: cohort activation <20% (unmanaged-floor figure, VENDOR flag)
→ stop recruiting, fix enablement or kill; no partner-sourced pipeline
within the stage's expected window → re-scope; 2 consecutive missed
quarters → kill
Conflict check : agency category overlaps direct SMB sales - deal registration and rules
of engagement required BEFORE Stage 2 launch (handed to conflict sibling)
Cost and ask : Stage 1 $95k loaded (0.5 manager, docs/sandbox hardening, pilot MDF);
Stage 2 $120k; Stage 3 $60k; program budget + one FY headcount ask
Not now : reseller/VAR - rejected: required margin breaks unit economics at current
price point, and the category is under structural margin pressure;
re-entry trigger: enterprise tier ships with room for channel margin.
MSP - deferred: multi-tenant administration not on the product roadmap
for 12 months; re-entry trigger: roadmap commitment lands.
Open questions : agency-principal willingness unverified (2 interviews scheduled);
marketplace take rate vs effective CAC pending real listing terms
```
Why this works:
- The sequence is the efficiency order re-ranked for this base: tech/ISV first because its integration surface is what makes Stage 3 cheap, referral already proven and so not re-opened, reseller and MSP deleted from the pool rather than parked below marketplace.
- One stage opens at a time, each gated on evidence the previous stage produced.
- The capacity numbers cap the sequence.
- Two categories were explicitly not added, each with a re-entry trigger.
- The conflict check fires before the overlapping stage, not after.
- Gate thresholds come from the vendor's own base rates with external figures flagged.
- Kill criteria were written before launch, when nobody was attached to the outcome.
## Negative example - the everything-now expansion plan, annotated
The way this goes wrong in practice - a plan a partnerships lead brings for review:
```
PARTNER STRATEGY - "Ecosystem Everywhere"
Q1: launch reseller program (industry data: partners drive 28% of revenue)
Q2: sign 3 GSIs + launch MSP tier
Q3: two hyperscaler marketplace listings + affiliate program
Q4: 200 total partners; partnerships = 20% of pipeline
Framing: Horizon 1 core resellers, Horizon 2 services, Horizon 3 ecosystem platform.
```
What is wrong, line by line:
- **Four category launches in four quarters, no capacity numbers anywhere.** Nothing says who manages, enables, or supports any cohort - the recruit-ahead-of-enablement failure, scheduled in advance.
- **"Partners drive 28% of revenue" as justification.** An analyst figure about _mature_ programs, quoted with no provenance flag and no relevance test; it describes companies a decade into the discipline, not one launching resellers next quarter.
- **Reseller first.** The highest-prerequisite transact category, launched before any motion is proven - no mention of deal registration, margin structure, certification, or a support model existing. The efficiency order puts it last - a standing job, at a collapsed margin, for booked revenue - and nothing here states the buyer-route reason that would promote it.
- **"Sign 3 GSIs" is a named-company goal, not a category decision** - and GSI is the slowest, highest-load category there is, placed in quarter two.
- **Calendar gates only.** Every stage triggers on a quarter; none on evidence. Q2 opens whether or not Q1 produced a single activated partner.
- **A partner-count target (200) as the success metric.** Logo counting; activation, time-to-first-deal, and sourced pipeline appear nowhere.
- **Three-horizons framing doing the work of a gate structure.** Vocabulary from an innovation-portfolio grid stands where kill criteria should be; no stage in the plan can fail.
- **No conflict check.** Resellers, MSPs, and marketplaces all overlap the direct motion; not one rule of engagement is named.
- **Nothing is deferred.** Every category considered was added, so the plan decided nothing - the loudest sponsor will decide later, off the record.
- **The ego-system trap runs underneath it all.** The plan assumes GSIs, MSPs, and marketplaces will show up because they are invited; no line explains what any of those partner types gains by joining now.
When a user brings a plan shaped like this, rebuild from the MVE and the capacity numbers up rather than tuning its dates.
SKILL.md›
---
name: partner-ecosystem-expansion
description: Decide which partner categories to add next to a partner ecosystem and in what sequence - a staged partner-category expansion roadmap with readiness gates, capacity limits, and kill criteria. Covers B2B categories (tech/ISV, reseller/VAR, MSP, SI/GSI, agency, referral/affiliate, marketplace) and consumer-side equivalents (retail/wholesale, creator, licensing, co-branding). Use whenever the user mentions partner ecosystem expansion, partner category planning, a partner mix roadmap, or asks which partner type to launch next, even if they never say expansion. Category-level sequencing only. Do NOT use for ranking named candidate companies - use mbfinotti/partnerships-skills@alliance-prioritization instead.
license: MIT
metadata:
author: Maya-Beth Finotti
version: "1.2.1"
---
# Partner Ecosystem Expansion
Plan which categories of partner to add next and in what order, as a staged roadmap with gates.
Be honest about the evidence before anything else:
- No peer-reviewed optimal partner-category sequence exists.
- Most published "frameworks" for this decision are borrowed strategy grids force-fit onto channels.
- Nearly every quoted benchmark comes from a vendor or consultant that profits when the number looks good.
What the evidence genuinely supports:
- **Adner's minimum viable ecosystem (MVE) + staged expansion + value blueprint** (_The Wide Lens_; _Winning the Right Game_) - the one framework designed for exactly this problem: prove the smallest partner configuration that evidences joint value, then add categories in a deliberate order, gated on evidence.
- **The alliance-portfolio-diversity inverted-U** (Jiang, Tao & Santoro 2010, _SMJ_; Lee et al. 2017 meta-analysis) - the field's only independent evidence: partner-type diversity helps firm performance up to a point, then coordination and governance costs dominate. It argues for restraint and sequencing; it prescribes no order.
- **Operational coverage/capacity math and stage-gate pilots** - unbranded, but the actual working method: gate every category on capacity, not enthusiasm.
McBain/Canalys' channel trifurcation supplies the category taxonomy (never a sequence), and Moore's whole product / bowling alley partially transfers. Name and reject the force-fits when a user brings one: Ansoff applied to channels, McKinsey three horizons, BCG growth-share, Iansiti & Levien for category selection. See [references/sequencing-frameworks.md](references/sequencing-frameworks.md).
This skill sequences categories, never named companies. Related decisions live in sibling skills:
- Ranking named candidates: `mbfinotti/partnerships-skills@alliance-prioritization`
- Mapping the current ecosystem: `mbfinotti/partnerships-skills@partner-ecosystem`
- Program structure (tiers, benefits, requirements): `mbfinotti/partnerships-skills@partner-channel-program`
## Interview
Ask before sequencing anything. One question per message; offer multiple-choice options; skip whatever the user already answered.
- B2B, B2C, or hybrid? What do you sell, to whom, at what price band, through which motion?
- Current ecosystem shape: which partner categories are live today, how many partners per category, and what share are actually active? (An ecosystem map from the mapping sibling is the ideal input.)
- Which existing category is proven - repeatable joint wins, not just signings? Any category still a half-built stub?
- Product readiness: API maturity and multi-tenancy, certification/training content, deal registration, margin structure, support model, contracting - which exist today, which are still on the roadmap?
- Internal capacity: partner managers and their current load, enablement content bandwidth, tooling, program budget?
- How does your buyer actually buy - self-serve, sales-led, through implementers, through retail or marketplaces?
- Goal of the expansion: coverage of a segment/geography/journey stage, a whole-product gap, new-logo pipeline, retention/expansion, margin capture?
- Geography and segment coverage today vs target?
- Budget and sign-off: who funds this, who approves, and is there a planning cycle it must land in?
- Time horizon for the roadmap - 12, 24, or 36 months?
Three answers re-rank the category order before any sequencing starts - ask them here, never beside the ranking:
- **By what date must the first result land?** A hard near-term date promotes the fast-acting categories (referral/affiliate, creator, agency) and deletes the standing-job ones from stage 1 - a category whose first revenue arrives after the date cannot be stage 1 whatever its ratio.
- **A one-off win, or a compounding asset?** Compounding promotes tech/ISV and the categories it makes cheap later (marketplace, MSP); one-off promotes referral/affiliate, creator and co-branding, and demotes everything that only pays back across years.
- **What is the effort ceiling** - hours, headcount, political capital, reversibility? A ceiling below one standing job deletes MSP, SI/GSI, reseller, retail and distributor outright; a ceiling that includes an existing services arm promotes SI/GSI to the front.
## Workflow
Wherever a real choice exists - candidate pool, sequence order, gate thresholds, where the capacity line falls - present 2-3 candidate approaches with trade-offs and your recommendation, then wait. Never let an assumption harden into the roadmap unstated.
1. Run the Interview. Confirm the input is a current-ecosystem picture (a map, or an honest mix description). No picture at all → route to the ecosystem-mapping sibling first.
2. State the evidence posture up front: the roadmap is a reasoned, gated bet, not a validated formula - and every benchmark used will carry a provenance flag.
3. Define the MVE: the smallest partner configuration that creates enough evidence of value to attract the next category. If the current motion is not yet an MVE with repeatable joint wins, the honest roadmap's stage 1 is finishing it, not adding.
4. Run the ego-system check: draft the value blueprint from the customer's position, not yours. A sequence that presumes your own centrality (Adner's ego-system trap) recruits categories that have no reason to show up.
5. Build the candidate pool from the trifurcation buckets - influence, transact, retain-and-deliver - plus the B2C vocabulary where relevant, and from the whole-product gaps each category would close. See [references/partner-category-economics.md](references/partner-category-economics.md).
6. Fix the capacity numbers before sequencing: partner-manager-to-partner ratios, enablement bandwidth, budget. These cap the roadmap; a sequence without them is a wish list.
7. Score each candidate category's readiness and economics: commercial model, relative time-to-first-revenue, effort to stand up, standing load, compliance exposure, prerequisites met vs missing. See [references/readiness-and-capacity-gates.md](references/readiness-and-capacity-gates.md).
8. Rank the pool by efficiency: value returned per unit of effort, never cheapest-first. Use the ordering and the per-axis lines in [references/partner-category-economics.md](references/partner-category-economics.md), then re-rank it against this user: their three interview answers above, an existing services arm, a product needing no implementation, a partner team of one. Delete every category their constraints rule out, with the reason, instead of parking it at the bottom; it lands in `Not now` with a re-entry trigger. Say the ordering out loud, and say what it starves. Draft 2-3 candidate sequences from the re-ranked pool with trade-offs and your recommendation.
9. Argue the strongest case against your own recommended sequence before presenting it: the inverted-U cost of another category, the conflict it creates, the capacity it steals from the proven motion.
10. Attach a stage gate to every stage: pilot cohort shape, a partner-market-fit threshold (repeatable joint wins), and explicit kill criteria. Gates trigger on evidence, never on a calendar date alone.
11. Pre-check channel conflict for any category that overlaps an existing route (direct, or another partner category). Detect and size it here; writing the rules of engagement belongs to `mbfinotti/partnerships-skills@partner-channel-conflict` - but a launch scheduled before rules exist is a roadmap defect.
12. Cost each stage as a capital request: total loaded cost (headcount, enablement build, incentives, tooling, program spend), never headcount alone, with the funding mechanism named.
13. Run the brainstorming discipline below; present the roadmap section by section, validating each with the user before drafting the next. Require explicit approval of the assembled roadmap before finalizing.
14. Check the Pass Threshold; iterate the sequence, gates, or capacity assumptions until every check holds.
15. Set the review cadence and re-sequencing triggers: a gate result, a category's activation collapsing, a product-readiness slip, a conflict incident, a market shift. The roadmap is a standing hypothesis, dated for its next review.
16. If your harness has persistent memory, memorize the approved roadmap - the sequence, each stage's gates and kill criteria, and the capacity numbers - so a later run (a gate review, a re-sequencing) resumes from them instead of re-interviewing.
17. If you can browse the web, verify time-sensitive facts the plan leans on (marketplace policies, a candidate category's structural economics) before finalizing; otherwise flag them as user-supplied and unverified.
## Brainstorming the sequence
Enter brainstorming mode before drafting any roadmap - never open with a finished sequence. Run it in the profession's own formats, not generic ideation.
The five are ranked by efficiency in [references/expansion-workshops.md](references/expansion-workshops.md) - `map re-read > value blueprint > account mapping > partner persona > business-design workshop` - with their question sets and the axes behind that order. Run from the top, confirm the format with the user, and stop when the candidate pool stops changing; the order shifts with what data already exists.
- One question per message, always; never batch.
- Every real choice gets 2-3 candidate approaches with trade-offs and your recommendation - then wait for the user.
- Name assumptions out loud before they harden: which category the room already wants, which readiness claim nobody verified, which capacity number is a guess.
- Argue against your own recommended sequence before presenting it (workflow step 9) - enthusiasm for a category is the documented way this exercise fools itself.
- Present the deliverable section by section - current base, candidate pool, capacity, sequence, gates, costs - each validated before the next is drafted.
- Gate finalization on the user's explicit approval of the assembled roadmap.
## The Staged Expansion Roadmap
Deliver the engagement as this artifact:
```
PARTNER-CATEGORY EXPANSION ROADMAP - <business area>, <date>
Current base : categories live today, activation per category, MVE status
Evidence posture: what is validated vs asserted; every benchmark used, with provenance flag
Candidate pool : categories considered, trifurcation bucket, gap each would close
Capacity : manager ratios, enablement bandwidth, budget - the numbers that cap the plan
Sequence : Stage 1..N - category, why now, why this order, prerequisites met/missing
Gates : per stage - pilot shape, partner-market-fit threshold, evidence required, owner
Kill criteria : per stage - activation floor, time-to-first-deal window, the two-quarter rule
Conflict check : overlap of each new category with existing routes, sized; rules needed pre-launch
Cost and ask : total loaded cost per stage, funding mechanism, sign-off owner
Not now : categories deleted from the ranked pool - deferred or rejected, each with the
constraint that ruled it out and a re-entry trigger
Open questions : what the evidence could not answer, and what would change the sequence
```
A full worked roadmap and a bad expansion plan annotated line by line live in [references/worked-examples.md](references/worked-examples.md).
## Pass Threshold
Ship nothing until all of these hold; iterate until they do:
1. **One stage opens at a time.** Each stage opens only when the previous stage passed its gate; at most two in parallel, and only when they share enablement infrastructure, stated explicitly.
2. **Evidence-gated, never calendar-gated.** Every gate names the evidence that opens it - repeatable joint wins, activation, sourced pipeline - never a date alone.
3. **Capacity-consistent.** Summed stage loads (manager time, enablement build, budget) fit inside the stated capacity numbers.
4. **Kill criteria on every stage** - at minimum an activation floor, a time-to-first-deal window, and the rule for two consecutive missed quarters.
5. **Every benchmark flagged.** Each quoted number carries its provenance flag (vendor, vendor-commissioned, analyst, consultant, academic); an unflagged number is a defect.
6. **At least one category deferred or rejected**, with a re-entry trigger. A roadmap that adds everything sequenced nothing.
7. **Conflict pre-checked** for every category that overlaps an existing route, with the needed rules named before that category's launch stage.
## KPIs
Judge each added category by cohort against its gate - and the roadmap by whether its gates worked. Instrument in efficiency order - gate decisions settled per unit of measurement effort - and stop when the gate can be read:
- Partner activation rate per new category (signed partners with at least one closed deal in year one), against the gate's floor.
- Time-to-first-deal per category, against the expected window set at the gate.
- Partner-sourced and partner-influenced pipeline (B2B) or incremental attributed revenue (B2C) per category, by cohort.
- Enablement/certification completion inside the onboarding window, per cohort.
- Category share of new revenue over time - starved by that order (a year of cohorts before it says anything), and promoted whenever the review is a funding decision rather than a stage gate.
- Gate survival: share of piloted categories that pass their gate rather than get killed or re-scoped - 100% suggests soft gates, 0% suggests a broken candidate pool.
- The inverted-U watch: coordination cost rising while total sourced revenue flattens as categories accumulate - the signal to stop adding and deepen.
Full indicator definitions and published figures with provenance flags: [references/kpis-and-benchmarks.md](references/kpis-and-benchmarks.md).
## B2B and B2C
Genuinely identical on both sides, so use the same method rather than reinventing it:
- Coverage/incrementality analysis to decide which category fills a gap.
- The MVE and pilot-then-scale discipline.
- Channel conflict and cannibalization as the central governance risk when a new category overlaps an existing route.
- Partner economics as a sequencing gate.
- The inverted-U logic that too many overlapping categories create cost and noise.
A DTC brand ranking creators, marketplace and wholesale by value per unit of effort runs the same category-portfolio logic as a SaaS vendor ranking ISV, agency and MSP. Both efficiency orders live in [references/partner-category-economics.md](references/partner-category-economics.md).
What genuinely changes:
| Dimension | B2B tech/SaaS | B2C/consumer |
| ---------------------------- | ----------------------------------------------------------------- | ------------------------------------------------------------------------------------ |
| Category vocabulary | ISV, VAR, MSP, SI/GSI, agency, referral, distributor, marketplace | Retail/wholesale, distributor, affiliate, creator/influencer, licensing, co-branding |
| Head of the efficiency order | Tech/ISV, then referral/affiliate | Creator/influencer, then affiliate |
| Sequencing driver | Product readiness, coverage gaps, buyer route | Incrementality (reach buyers before intent) and margin capture |
| Attribution | Deal registration, account mapping, co-sell | Codes, affiliate links, last-click |
| Dominant risk | Enablement lag; conflict with direct sales | DTC vs retail vs marketplace cannibalization |
## Failure Modes
| Failure | Fix |
| -------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Ego-system trap - the sequence presumes partners will orbit you | Blueprint from the customer's position; verify each category's reason to join now |
| Recruiting ahead of enablement - the field's most-cited failure | Capacity gate before recruiting; activation below ~20% means stop recruiting, not recruit harder |
| Over-rotation on one category, cutting others (the Nike/Under Armour pattern) | Treat category subtraction as high-risk as addition; size the conflict and dependency before shrinking a route |
| Category sprawl, then forced pruning (the HubSpot agency-badge pattern) | Kill criteria plus the deferred list; watch the inverted-U signal |
| Premature diversification - category two before category one is proven | MVE gate: repeatable joint wins first; the honest stage 1 may be "finish the current motion" |
| Cheapest category first, mistaken for the most efficient one | Rank by value per unit of effort: referral is the cheapest rung and buys the least, while an ISV motion costs a quarter and makes three later categories cheaper |
| A ruled-out category parked at the bottom of the ranking | Delete it from the pool with its reason; it belongs in `Not now` with a re-entry trigger, not in a list the loudest sponsor can re-read upward |
| Borrowed strategy grid presented as the method (Ansoff, three horizons, BCG, Iansiti-Levien) | Keep it as vocabulary if it helps the room; rebuild the decision on MVE + gates + capacity |
| Calendar-gated stages - "SI program launches in Q3" regardless of evidence | Gates trigger on evidence; dates are forecasts, not triggers |
| Vendor benchmarks quoted as proof | Provenance flag on every number; vendor figures are directional context only |
| Roadmap drifts into ranking named companies | Categories only; hand named candidates to `mbfinotti/partnerships-skills@alliance-prioritization` |
## Invocation Examples
- "We have resellers and referral partners. Which category should we add next, and when? Two partner managers, no new headcount."
- "Turn our ecosystem map into a three-year category roadmap with readiness gates and kill criteria."
- "Leadership wants MSPs, SIs and a marketplace motion all this year. Sequence them honestly against our capacity."
## Reference
- [references/partner-category-economics.md](references/partner-category-economics.md) - per-category economics ranked by efficiency (commercial model, what it buys, time-to-first-revenue, effort, standing load, compliance exposure), B2B and B2C vocabularies, the per-axis orderings, what the order starves, and how to re-rank it for this user.
- [references/sequencing-frameworks.md](references/sequencing-frameworks.md) - Adner's MVE/staged expansion/value blueprint and the ego-system trap, the inverted-U, the trifurcation, whole product/bowling alley, and the named force-fits with why each fails.
- [references/readiness-and-capacity-gates.md](references/readiness-and-capacity-gates.md) - readiness prerequisites per category, capacity math, the business-case shape, stage-gate design, kill criteria.
- [references/kpis-and-benchmarks.md](references/kpis-and-benchmarks.md) - leading indicators, published numbers with provenance flags, the vendor-bias warning, the known-unstable statistics.
- [references/expansion-workshops.md](references/expansion-workshops.md) - the five facilitation formats ranked by efficiency, with their question sets.
- [references/worked-examples.md](references/worked-examples.md) - a full worked staged roadmap and a bad expansion plan annotated line by line.
Sibling skills (same collection):
- `mbfinotti/partnerships-skills@partner-enablement` - the enablement build each stage's prerequisites demand.
- `mbfinotti/partnerships-skills@referral-incentive-design` - incentive mechanics for a referral/affiliate stage.