sales-discovery-questions

{
  "skill_name": "sales-discovery-questions",
  "evals": [
    {
      "id": 1,
      "prompt": "I run the SDR team at Corvidex, we sell warehouse labour-planning software. Tomorrow I've got a 15-minute call with an IT manager at Tallgrass Accounting, mid-market, about 400 people, who filled in the 'book a demo' form on our pricing page yesterday. My VP wants the BANT fields filled in the CRM after every first call, so please open with the budget and timeline questions and get the decision maker's name early, then whatever else fits in the time. Give me the question list.",
      "expected_output": "A 15-minute SDR call sheet whose question path runs pain, root cause, impact first, with budget, timeline and authority questions pushed behind impact or justifiably skipped, plus an explicit do-not-ask list naming the two logistics-first offenders.",
      "files": [],
      "expectations": [
        "Places budget and timeline questions after the impact stage, never as openers.",
        "Names 'What's the timeline for this project?' and 'Is there budget allocated?' (or near-identical wordings) on an explicit do-not-ask-as-openers list.",
        "Explains that logistics-first questions harvest deal admin before establishing that a problem exists, rather than only refusing the request.",
        "States that budget and timeline become legitimate later in the call, once impact has been quantified.",
        "Moves the decision-maker/authority question into a decision-context stage that follows impact, or justifies skipping that stage on a 15-minute SDR call.",
        "Sizes the set to roughly one primary question plus its follow-ups per 3-4 minutes of questioning time.",
        "Does not cite a target or ideal number of discovery questions as a rule to hit.",
        "Caps situation questions at 1-2 on the grounds that research should answer the rest.",
        "Distinguishes the inbound origin from outbound, treating this buyer as already problem-aware rather than needing awareness built first.",
        "Orders the question path pain, then root cause, then impact, with root cause probed before any quantifying question.",
        "Attaches a 2-3 rung follow-up ladder to at least one primary question.",
        "Delivers a call sheet carrying ranked hypotheses to test, branch triggers, and a dated next-step ask."
      ]
    },
    {
      "id": 2,
      "prompt": "I'm building a standard discovery template for my five AEs at Brightmoor Systems - HR compliance SaaS, about $40k ACV, 45-minute first calls. I read that top reps ask 11-14 questions and keep a 43:57 talk-to-listen ratio, so I want the template to land on exactly 13 questions, and I want to score the reps against that ratio in our Monday review. Draft the template and tell me how to enforce the two numbers.",
      "expected_output": "A 45-minute template sized by call minutes rather than by a question count, with both benchmarks explicitly caveated as correlational vendor claims that the vendor's own later data revised, and conversion-per-persona proposed as the real measurement.",
      "files": [],
      "expectations": [
        "Refuses to set 13, or any fixed count, as a target the template must hit.",
        "States that the published question-count figure is a vendor claim, correlational, and drawn from a non-auditable corpus.",
        "Notes that the same vendor's later refresh reported different numbers - won deals around 15-16 questions and lost deals around 20 - so the benchmark is internally inconsistent across its own refreshes.",
        "States that the 43:57 talk-to-listen ratio was revised by the same vendor's own later data to a much smaller gap, roughly 57% rep talk on won calls against roughly 62% on lost.",
        "Recommends using talk-listen only as a directional coaching signal, never as a scored target.",
        "Sizes the 45-minute call by minutes, roughly one primary question plus its ladder per 3-4 minutes of questioning time, instead of by a question count.",
        "Coaches the behaviour - fewer, deeper, evenly spread questions with follow-ups - rather than the number.",
        "Proposes first-call-to-next-meeting or first-call-to-opportunity conversion, tracked per persona and per template version, as the measurement that matters.",
        "Notes that the question-count rule of thumb explicitly breaks down against C-suite buyers.",
        "Includes per-call success signals such as several distinct problems surfaced with at least one quantified in the prospect's own number.",
        "Adds depth through follow-up ladders rather than by expanding the count of standalone primary questions.",
        "Does not present either the ratio or the count as an established or proven law of discovery."
      ]
    },
    {
      "id": 3,
      "prompt": "Q3 ends in five weeks and I need this one to land. Aventrix Fleet, freight telematics - I'm talking to their Director of Safety on Thursday for 30 minutes. It's the second call, the first was mostly a product overview. She hasn't given me any deadline of her own. Write me questions that build urgency and get her to feel like waiting is expensive. I want to walk out with a signed pilot scope.",
      "expected_output": "A question path that quantifies impact before any why-now question, hunts a real trigger event, records the honest absence of a compelling event if none exists, and keeps the rep's quarter-end out of the wording.",
      "files": [],
      "expectations": [
        "Places every urgency or why-now question after at least one quantifying impact question.",
        "States that urgency asked before impact is quantified reads as pressure rather than diagnosis.",
        "Cites published buyer-indecision research finding that added pressure deepens indecision rather than resolving it.",
        "Instructs the rep to record the honest absence of a compelling event rather than inventing or implying a deadline.",
        "Treats the absence of a compelling event as deal intelligence rather than a failed call.",
        "Contains no leading questions with the desired answer embedded in the wording, such as 'that must be costing you a fortune, right?'.",
        "Frames cost of inaction in the prospect's own number rather than in the rep's ROI math.",
        "Notes that mentioning the seller's ROI math during discovery correlates with worse next-step outcomes.",
        "Includes at least one question hunting a real trigger event - renewal, launch, audit, season - instead of a manufactured one.",
        "Probes root cause before quantifying impact, so the figure attaches to the problem rather than to a symptom.",
        "Keeps the rep's own quarter-end out of the question wording, or states plainly that it is not a reason the buyer can act on.",
        "Closes with a specific, dated next step argued from the impact figure uncovered on the call."
      ]
    },
    {
      "id": 4,
      "prompt": "I sell residential solar installs, average job around $28,000, and I do a 75-minute in-home sit-down with both homeowners. Right now I just walk through the system sizing sheet and ask about their electric bill. Build me a proper question set for the appointment - Dale and Priya Achterberg, they've been in the house eleven years, called us off a yard sign.",
      "expected_output": "A B2C high-ticket consult sheet running two spines, one per homeowner, leading impact with personal dimensions, with the decision-context stage deleted and prior-attempts carrying root cause.",
      "files": [],
      "expectations": [
        "Asks the pain and impact questions of each homeowner separately, treating two decision makers as two spines.",
        "Leads the impact stage with personal impact - household time, money, stress - ahead of the business and technical dimensions.",
        "Deletes the decision-context or buying-process stage outright rather than including a reduced version of it.",
        "Carries the root-cause stage with a prior-attempts question in the form of 'what's kept you from doing this until now?'.",
        "Opens the pain stage with the trigger question 'what finally made you book this conversation?' or an equivalent asking what prompted the appointment now.",
        "Uses the eleven-year tenure as the anchor for a why-now question rather than assuming a deadline exists.",
        "Includes a branch for the two homeowners disagreeing that surfaces the disagreement as a question instead of taking a side.",
        "Excludes hypothetical purchase questions such as 'would you buy, use, or pay for X'.",
        "Caps situation questions at 1-2, keeping the system sizing and electric-bill fact-finding out of the opening.",
        "Keeps the same spine, sequencing rules, follow-up ladders and quality gate used for B2B rather than inventing a separate B2C structure.",
        "Ends with a summary in the prospects' own words followed by a specific dated next step."
      ]
    },
    {
      "id": 5,
      "prompt": "Our whole floor runs MEDDPICC and the CRM has a field per letter. I'm an AE at Kestrel Ledger - treasury automation, enterprise, deals run 5-7 months. I have a 45-minute first discovery with a Director of Treasury Ops at a $2B distributor next Tuesday. Give me the question set and, at the end, score where the deal sits on each MEDDPICC letter so I can paste it straight into the CRM.",
      "expected_output": "A 45-minute question set with every spine stage relabeled into MEDDPICC vocabulary, the other frameworks removed from the sheet, decision-context gathered as questions only, and the scoring request declined and routed elsewhere.",
      "files": [],
      "expectations": [
        "Maps every spine stage to MEDDPICC vocabulary so the answers land in the team's CRM fields.",
        "Removes SPIN, Sandler and GAP from the delivered sheet rather than ranking them below MEDDPICC.",
        "Gathers decision-context elements as questions only, producing no qualification score, verdict, or letter-by-letter rating.",
        "Routes the scoring itself to a separate MEDDPICC scorecard skill instead of performing it here.",
        "Uses the Three Whys - 'Why anything? Why you? Why now?' - for the urgency stage.",
        "Assigns 'why anything' and 'why now' to discovery and defers 'why you' as positioning for later.",
        "Keeps the decision-context stage in the set, given the enterprise deal and multi-month cycle.",
        "Still orders the path pain, root cause, impact, urgency rather than walking the MEDDPICC letters in their acronym order.",
        "Includes a metrics question in the form of 'what would you need to see to know it worked?'.",
        "Includes a prior-evaluation question covering competition, including the do-nothing option.",
        "Notes that MEDDPICC elements are gathered rather than ticked off as a checklist.",
        "Sizes the set to the 45-minute call rather than to the eight letters."
      ]
    },
    {
      "id": 6,
      "prompt": "Got 20 minutes with the COO of Marrowfield Grocers on Monday - she's stepping into a slot her Director of Supply Chain originally booked. We sell demand forecasting. I want all three impact angles covered, technical, business and personal, plus a good walk through their current process so I understand the setup before I ask anything hard.",
      "expected_output": "A 20-minute executive sheet with technical impact deleted rather than compressed, a halved question count, the process walkthrough replaced by a point-of-view statement, and any skipped stage logged as a next-touch gap.",
      "files": [],
      "expectations": [
        "Deletes the technical impact dimension outright rather than compressing all three into 20 minutes.",
        "States that two funded impact dimensions beat three rushed ones, and that the quality bar asks for two.",
        "Keeps business impact ahead of personal impact for this persona.",
        "Halves the question count relative to a standard discovery call because an executive is in the room.",
        "Trades the situation or current-process walkthrough for a point-of-view statement rather than delivering it as questions.",
        "Leads with the single sharpest pain hypothesis instead of testing several in sequence.",
        "Caps situation questions at one at most, declining to run the requested full process walkthrough with the COO.",
        "Notes that published question-count rules of thumb explicitly break down against C-suite buyers.",
        "Prioritizes whichever lens reaches a quantified figure fastest with this executive.",
        "Still probes root cause before the impact question rather than dropping that stage to save minutes.",
        "Marks any stage skipped for time as an explicit gap to close on the next touch."
      ]
    },
    {
      "id": 7,
      "prompt": "Enterprise deal, I'm the AE. Selling clinical scheduling software into Ridgeline Health, a 9-hospital system. First real discovery is a 45-minute call with their VP of Nursing Informatics, and I already know at least eight people will touch this before it's signed - nursing, IT security, legal, supply chain, and a CFO somewhere. Build my question set. I'd also like a couple of questions about her personal comp or bonus targets, since that usually motivates people.",
      "expected_output": "A regulated, multi-stakeholder set that funds decision-context coverage first, adds a compliance and procurement review question, drops the personal-comp request outright, and splits into persona-specific variants.",
      "files": [],
      "expectations": [
        "Prioritizes decision-context coverage first when allocating the call's question minutes, because the buying group runs past a single economic buyer and the buyer is regulated.",
        "Drops the requested questions about the VP's personal compensation or bonus targets.",
        "States that personal-finance-adjacent questions are off limits with a regulated buyer, rather than merely rephrasing them.",
        "Adds an explicit security, compliance, or procurement review question for a purchase of this kind.",
        "Expands decision-process and procurement coverage relative to a non-regulated deal.",
        "Generates persona-specific question variants rather than one set reused across all eight stakeholders.",
        "Aims workflow-pain questions at end users, business-outcome questions at the economic buyer, and architecture or security questions at technical evaluators.",
        "Routes stakeholder strategy itself - naming the champion, economic buyer and blockers - to a separate stakeholder-mapping skill rather than mapping the committee here.",
        "Keeps personal impact in scope as the team's week-to-week experience, distinct from the banned personal-finance questions.",
        "Gathers decision-context elements as questions without producing a qualification verdict.",
        "Notes that discovery recurs per persona across the cycle rather than finishing in this one call.",
        "Keeps the pain, root cause, impact, urgency ordering intact within the call itself."
      ]
    },
    {
      "id": 8,
      "prompt": "I'm certain what's wrong at Quillon Freight - their AP team is drowning in manual invoice matching, I've seen it at every 3PL this size. Thirty-minute call with their Controller on Wednesday, outbound, he replied 'sure, send a time' to my third email. Write me questions that walk him into admitting the matching problem so I can get to a demo. I've already read their annual report and their careers page.",
      "expected_output": "A set built on 2-3 ranked hypotheses to test rather than one to confirm, with no leading questions, past-specific wording, capped situation questions, and a branch for when no pain surfaces.",
      "files": [],
      "expectations": [
        "Restates the manual-invoice-matching belief as a hypothesis to test rather than a fact to confirm.",
        "Produces 2-3 candidate pain hypotheses, ranked, instead of building the whole set around the single assumed one.",
        "Contains no leading questions with the desired answer embedded in the wording.",
        "Requires a recent, specific example before a pain counts as found.",
        "Uses past-specific wording such as 'when did that last happen, and what did it cost?' rather than hypotheticals about imagined situations.",
        "Keeps questions product-agnostic until the close, with the product entering only after the gap is established.",
        "Includes a branch for no pain surfacing after two hypotheses, offering an honest fallback question.",
        "Instructs the rep not to manufacture pain when neither hypothesis lands.",
        "Caps situation questions at 1-2 because the annual report and careers page already answer the rest.",
        "Opens the pain stage for this outbound contact with a one-sentence point-of-view statement to react to, rather than assuming problem awareness.",
        "Earns 2-3 follow-up rungs on each meaningful answer before moving to a new topic."
      ]
    },
    {
      "id": 9,
      "prompt": "Second discovery with Halbrook Dental Group's Operations Director, 30 minutes on Friday. On the first call she volunteered, unprompted, that rework on their insurance claims costs them about $14,000 a month, then immediately asked what we charge and I fumbled it. I don't want to waste her time re-asking about the problem, so I'd like to go straight from that number into urgency and the proposal. My manager also wants this sheet reusable for every practice manager we call.",
      "expected_output": "A sheet that banks the $14,000 figure but returns to root cause before urgency, handles the price question with a brief answer and a trade back into the spine, and ships as a reusable persona template with full ladders.",
      "files": [],
      "expectations": [
        "Banks the $14,000-per-month figure and returns to the root-cause stage rather than skipping it.",
        "States that quantifying a symptom quantifies the wrong thing, so a volunteered figure does not exempt the root-cause stage.",
        "Carries the $14,000 monthly figure through to the close summary and the next-step ask.",
        "Includes a prior-attempts question - what they tried, and whether it worked - in the root-cause stage.",
        "Explains that the spine orders exit conditions rather than a fixed script, which is why an out-of-order answer is banked instead of dropped.",
        "Handles the mid-call price question with a brief answer followed by a trade back into the spine, not a pricing discussion.",
        "Treats the reusable-across-practice-managers requirement as a compounding asset that funds the full follow-up ladder rather than the bare spine.",
        "Delivers a persona-level template with slots to fill per account, not a one-off sheet tied only to this prospect.",
        "Requires at least one impact dimension beyond the business figure already volunteered.",
        "Places urgency questions after root cause has been established, not immediately after the banked figure.",
        "Ends with an after-the-call section listing which stage exit conditions went unmet, as gaps for the next touch."
      ]
    }
  ],
  "trigger_queries": [
    {
      "query": "build my discovery question set for tomorrow's 30-minute call with a VP Ops at a 200-person logistics company",
      "should_trigger": true
    },
    { "query": "what should I ask this prospect on Thursday", "should_trigger": true },
    { "query": "discovery call prep for an inbound demo request", "should_trigger": true },
    { "query": "I need qualification questions for a 15-minute SDR call", "should_trigger": true },
    { "query": "give me SPIN questions for a first sales meeting", "should_trigger": true },
    {
      "query": "pain funnel questions for a prospect who keeps giving me vague answers",
      "should_trigger": true
    },
    { "query": "first meeting agenda for a new sales opportunity", "should_trigger": true },
    {
      "query": "I have a call with a CFO tomorrow and no idea what to ask her",
      "should_trigger": true
    },
    {
      "query": "help me plan the questions for my in-home consultation with a couple buying a kitchen",
      "should_trigger": true
    },
    {
      "query": "what questions get someone to tell me what a problem actually costs them",
      "should_trigger": true
    },
    {
      "query": "sales discovery questions for a security-conscious enterprise buyer",
      "should_trigger": true
    },
    { "query": "my discovery calls feel like interrogations, fix my question list", "should_trigger": true },
    {
      "query": "I keep hearing 'we're just too busy' and can't get past it on calls",
      "should_trigger": true
    },
    { "query": "write me the question path for a 45-minute AE deep discovery", "should_trigger": true },
    { "query": "how do I structure a first call with a prospect", "should_trigger": true },
    { "query": "need a call sheet for Tuesday's prospect call", "should_trigger": true },
    { "query": "questions to uncover root cause on a sales call", "should_trigger": true },
    { "query": "prep me for a discovery call with a director of supply chain", "should_trigger": true },
    {
      "query": "I want a reusable question template my AEs can run on every first call",
      "should_trigger": true
    },
    { "query": "how do I get a prospect to quantify their pain", "should_trigger": true },
    { "query": "what should I ask a technical evaluator during discovery", "should_trigger": true },
    { "query": "questions for a solar consultation with homeowners", "should_trigger": true },
    {
      "query": "my reps jump straight to the demo, give me something that makes them diagnose first",
      "should_trigger": true
    },
    {
      "query": "give me follow-up questions to go deeper when someone answers in one line",
      "should_trigger": true
    },
    { "query": "MEDDPICC discovery questions for a treasury ops director", "should_trigger": true },
    {
      "query": "help me build a question set to test my hypothesis about their problem",
      "should_trigger": true
    },
    {
      "query": "I don't know what to ask on a first sales call with a hospital system",
      "should_trigger": true
    },
    {
      "query": "need questions that find the compelling event without being pushy",
      "should_trigger": true
    },
    {
      "query": "discovery questions for an outbound prospect who has no idea they have a problem",
      "should_trigger": true
    },
    { "query": "what do I ask to figure out how they'd actually buy this", "should_trigger": true },
    {
      "query": "I have 20 minutes with a COO, what are the highest value questions",
      "should_trigger": true
    },
    { "query": "build a qualification call script for our new SDRs", "should_trigger": true },
    { "query": "questions to ask before I scope a pilot", "should_trigger": true },
    {
      "query": "how many questions should I ask on a discovery call and which ones",
      "should_trigger": true
    },
    {
      "query": "prep questions for a second call where the first was all product overview",
      "should_trigger": true
    },
    {
      "query": "give me a list of things to ask a prospect and what to listen for in their answers",
      "should_trigger": true
    },
    {
      "query": "our first calls never turn into second meetings, help me with the questions",
      "should_trigger": true
    },
    {
      "query": "design the questioning sequence for a complex multi-stakeholder deal",
      "should_trigger": true
    },
    {
      "query": "I want to stop asking about budget and timeline first, what should I ask instead",
      "should_trigger": true
    },
    { "query": "what do I say after 'tell me about your current process'", "should_trigger": true },
    { "query": "questions for a high-ticket coaching consultation", "should_trigger": true },
    { "query": "sales call questions that aren't leading", "should_trigger": true },
    { "query": "need a do-not-ask list for my team's discovery calls", "should_trigger": true },
    { "query": "build discovery for a regulated finance buyer", "should_trigger": true },
    { "query": "how should I sequence pain, impact and urgency questions", "should_trigger": true },
    { "query": "help me prepare for tomorrow's meeting with a new lead", "should_trigger": true },
    {
      "query": "what do I ask a practice manager to find out where their money leaks",
      "should_trigger": true
    },
    {
      "query": "write the first 10 seconds of my cold call so they don't hang up",
      "should_trigger": false
    },
    { "query": "what do I say when they pick up the phone", "should_trigger": false },
    { "query": "they said we're too expensive, what do I say back", "should_trigger": false },
    {
      "query": "build me an objection library for price and timing pushback",
      "should_trigger": false
    },
    { "query": "score this call transcript and tell me how my rep did", "should_trigger": false },
    { "query": "grade my discovery call recording against a rubric", "should_trigger": false },
    { "query": "score this opportunity on MEDDPICC", "should_trigger": false },
    { "query": "is this deal real, score it against BANT", "should_trigger": false },
    { "query": "who's the actual champion in this deal", "should_trigger": false },
    { "query": "build a power map for this account", "should_trigger": false },
    { "query": "scan these call notes for deal risk", "should_trigger": false },
    { "query": "this deal has gone quiet, is it slipping", "should_trigger": false },
    { "query": "write the follow-up email after my discovery call", "should_trigger": false },
    { "query": "turn my call notes into a recap with next steps", "should_trigger": false },
    { "query": "find me a reason to reach out to this prospect", "should_trigger": false },
    { "query": "they just raised a Series B, what's my angle", "should_trigger": false },
    { "query": "how many touches should my outbound sequence have", "should_trigger": false },
    { "query": "design a 14-day cadence with calls and emails", "should_trigger": false },
    { "query": "test these three subject lines for me", "should_trigger": false },
    { "query": "why are my cold emails landing in spam", "should_trigger": false },
    { "query": "build the ROI business case for this deal", "should_trigger": false },
    { "query": "what payback period should I show the CFO", "should_trigger": false },
    { "query": "procurement wants 20% off, plan my concessions", "should_trigger": false },
    { "query": "what should I trade for a multi-year commitment", "should_trigger": false },
    { "query": "define our ideal customer profile", "should_trigger": false },
    { "query": "build a fit score for our accounts", "should_trigger": false },
    { "query": "where should the cutoffs between our account tiers sit", "should_trigger": false },
    { "query": "how much quota should an AE carry next year", "should_trigger": false },
    { "query": "design the commission plan for our SDRs", "should_trigger": false },
    {
      "query": "what pipeline coverage ratio do we need to hit the number",
      "should_trigger": false
    },
    { "query": "write the interview questions for hiring an AE", "should_trigger": false },
    { "query": "build a scorecard and interview loop for a new SDR hire", "should_trigger": false },
    { "query": "how do I break into tech sales from support", "should_trigger": false },
    { "query": "prep me for my SDR interview tomorrow", "should_trigger": false },
    { "query": "which sales podcasts and newsletters should I follow", "should_trigger": false },
    { "query": "should we go product-led or sales-led", "should_trigger": false },
    { "query": "pods vs assembly line for our sales team", "should_trigger": false },
    { "query": "how big is the addressable market for this product", "should_trigger": false },
    { "query": "which sales skill should I start with", "should_trigger": false },
    { "query": "I have a broad sales problem and don't know where to begin", "should_trigger": false },
    {
      "query": "customer discovery interviews to validate my startup idea",
      "should_trigger": false
    },
    {
      "query": "set up a continuous discovery habit with weekly customer interviews",
      "should_trigger": false
    },
    {
      "query": "write user research interview questions for our onboarding flow",
      "should_trigger": false
    },
    {
      "query": "questions to ask in a stakeholder discovery workshop before a software build",
      "should_trigger": false
    },
    {
      "query": "what should I ask participants in a usability study to avoid biasing them",
      "should_trigger": false
    },
    {
      "query": "draft the discovery questionnaire we send clients before a consulting engagement",
      "should_trigger": false
    },
    {
      "query": "questions for a churn interview to find out why a customer left",
      "should_trigger": false
    }
  ]
}

# Discovery call sheet - fill-in template

One page. The rep reads it 20 minutes before the call; it must survive being skimmed.

```
DISCOVERY CALL SHEET
Prospect   : <persona, title, company>          Call: <type>, <length>
Origin     : <inbound / outbound + what they responded to>
Known      : <2-4 bullets from prior touches - never re-ask these>
Next step  : <the success outcome this call closes toward, with a date>

HYPOTHESES (test, don't confirm)
H1: <suspected pain - ranked first>
H2: <alternate>
(H3: <fallback>)

FRAME
"<one-line upfront agenda - purpose, time, decide-together outcome>"

QUESTION PATH
[Situation - max 2, only what research couldn't answer]
Q: <question>
   listen for: <signal>            exit: current process in their words

[Pain]
Q: <primary, tied to H1>
   ladder: <rung 1> / <rung 2> / <rung 3>
   listen for: <a recent, specific example>   exit: named problem + example

[Root cause]
Q: <primary>
   ladder: <prior attempts> / <did it work>
   exit: cause ≠ symptom; attempts known

[Impact - need a figure, ≥2 of technical/business/personal]
Q: <business-impact question>
Q: <technical- or personal-impact question>
   listen for: their number, who else is touched
   exit: at least one quantified figure, in their words

[Urgency]
Q: <why-now question>
Q: <cost-of-inaction question>
   exit: dated/triggered reason to act - or its honest absence, noted

[Decision context - complex B2B only]
Q: <process question>   Q: <other-voices question>
   exit: path to a decision sketched (feeds the scorecard - no score here)

CLOSE
Summary: "So if I've got it - <pain>, because <cause>, costing <figure>..."
Ask    : "<specific next step> on <date>?"

BRANCHES
- No pain after H1+H2 → <fallback question>; do not manufacture pain
- Impact volunteered early → bank it, loop back for root cause
- Executive tone → halve the questions, lead with sharpest hypothesis + POV
- Price/demo asked mid-spine → brief answer, trade back one stage

DO NOT ASK
- Timeline or budget before impact
- "Would you buy/use...?" hypotheticals
- <deal-specific leading question to avoid>

AFTER THE CALL (fill in)
Stages unmet: <which exit conditions failed - these are the next-touch gaps>
```

# Frameworks and evidence

The four questioning frameworks with a publicly documented structure, what each one costs and buys as a coverage lens, and the honest state of the evidence. Rows below are ordered by efficiency; the selection logic ranks them in this same order.

## The four frameworks, in efficiency order

| Framework           | Origin                                                                      | Structure                                                                                                                    | Costs (call minutes, rehearsal, buyer patience)                                                                                                                                  | Buys                                                                                                                                                                                                         |
| ------------------- | --------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| GAP Selling         | Keenan, 2018                                                                | Current state → future state → gap; PIC chart: Problem / Root Causes / Impacts; impact split technical / business / personal | An hour of per-deal rehearsal; two impact questions inside a stage already open                                                                                                  | Root-cause-before-impact ordering, the three impact dimensions, and the quantified figure everything downstream runs on; the logistics-first negative examples                                               |
| Sandler Pain Funnel | Sandler (sequence published by a vendor; Sandler's own pages don't show it) | Vague → specific example → duration → prior attempts → cost → emotional weight                                               | The same hour of rehearsal and the same handful of minutes - ties GAP on effort; spends the most goodwill per stage, since every rung asks the buyer to stay on one sore subject | Depth on one pain: the ladder that separates a complaint from a problem, inside the pain and impact stages                                                                                                   |
| SPIN                | Neil Rackham, 1988 book                                                     | Situation → Problem → Implication → Need-payoff                                                                              | Near-zero - re-sequences questions already written, adds none                                                                                                                    | Near-zero on top of the spine, which has already absorbed its macro ordering: context before problem, problem before consequence. Free, never budgeted for                                                   |
| MEDDPICC            | Originated at PTC, 1990s (Dick Dunkel); popularized by MEDDICC (Andy Whyte) | 8 elements incl. Metrics, Identify Pain, Decision Process; the Three Whys: "Why anything? Why you? Why now?"                 | Highest: a whole extra stage, re-gathered per persona across the cycle, and minutes spent on process rather than on pain                                                         | Decision-context coverage and the Three Whys for urgency - the path by which a quantified pain reaches someone who can approve it. "MEDDIC is not a checklist" (MEDDICC) - elements are gathered, not ticked |

Convergence, not any single acronym, is the skill's spine: establish the situation, isolate a real problem, dig to root cause, quantify impact across several dimensions, then establish why now.

## Applying the chosen lens

Step 4 already decides which lens and which ones get deleted; this section only covers how the survivor is applied.

- Mandated methodology → map every stage to its vocabulary (e.g. MEDDPICC shop: label impact questions "Implicate the Pain", metrics questions "Metrics") so the call notes land in their CRM fields.
- Decision-context answers feed `meddpicc-scorecard` downstream wherever that stage survives; this skill never scores.
- Challenger-style outbound → not a fifth lens but a contingency on top of whichever one you kept: open pain with a point-of-view statement to react to. Insight-led selling de-emphasizes questions up front, it doesn't remove the need to diagnose.

## Evidence caveats - read before quoting any number

- SPIN's research base ("35,000 sales calls") is proprietary and was never published for independent replication; Rackham himself publicly acknowledged skepticism about the figure. Treat SPIN as a well-established practitioner ordering, not a proven law.
- The "ideal question count" is a vendor claim:
  - One conversation-intelligence vendor (Gong) published 11-14 questions as correlating with success.
  - Its own 2025 refresh (519,000+ calls) reported won deals at ~15-16 questions and lost deals at ~20.
  - The rule breaks down against C-suite buyers.

  Correlational, non-auditable corpus, never independently replicated. Coach the behavior - fewer, deeper, evenly spread questions - never the number.

- The "43:57 talk-to-listen ratio" (same vendor, 2016, n≈25,500 calls) was revised by the vendor's own 2025 data to a much smaller gap (~57% rep talk on won vs ~62% on lost). Same caveats. Use talk-listen only as a directional coaching signal.
- Independent peer-reviewed work supports the general premise, not the numbers: a 2011 meta-analysis (Verbeke, Dietz & Verwaal, Journal of the Academy of Marketing Science) found adaptiveness and selling-related knowledge among the strongest performance predictors. Treat any specific question-count or ratio threshold as unvalidated by that literature.
- Buyer-indecision findings (Dixon & McKenna, published book research on ~2.5M conversations): a large share of deals end in no-decision, and added pressure worsens indecision - the basis for this skill's "record the absence of a compelling event honestly" rule. Book-published vendor-adjacent research; same replication caveat.
- Buying-group size:
  - A Gartner survey (2017, n=750, Gartner's own figure) puts the median B2B buying group at 6-10 decision makers.
  - A 2024 Forrester survey reports averages around 13.

  Analyst figures from paid research - directionally solid, precise numbers vary by study.

- If a claim appears elsewhere and has no traceable source listed here or in the question bank, use "not verified" rather than asserting it.

# Question bank by spine stage

Wordings are starting points - rewrite every question into the prospect's vocabulary and the deal's specifics before it goes on a call sheet.

## Table of Contents

- [Frame (1 line)](#frame-1-line)
- [Situation (cap at 1-2)](#situation-cap-at-1-2)
- [Pain](#pain)
- [Root cause](#root-cause)
- [Impact](#impact)
- [Urgency / compelling event](#urgency-compelling-event)
- [Decision context (complex B2B only - questions, never scores)](#decision-context-complex-b2b-only---questions-never-scores)
- [Close](#close)
- [B2C high-ticket variants](#b2c-high-ticket-variants)
- [Do-not-ask list](#do-not-ask-list)

## Frame (1 line)

- "Fair if we spend 25 minutes on where things break down for you, then decide together whether a deeper look makes sense?" (applies Sandler's upfront-contract concept: agree purpose, time, and possible outcomes upfront.)

## Situation (cap at 1-2)

Research replaces situation questions - asking facts the company publishes burns credibility. Keep only what research cannot answer.

- "Walk me through how your team handles [process] today." (SPIN Situation wording as restated by a vendor publisher.)
- "Who touches that process day to day, and what tools does it live in?"

## Pain

- "What's the biggest hurdle you're facing right now with [goal]?" (SPIN Problem restatement.)
- "What's the hardest part of [process] as it works today?"
- "When did it last go wrong? Walk me through what happened." (past-specific anchor.)
- Outbound variant - offer a point of view first, then test it: share the pattern seen at similar companies in one sentence, then "how close is that to your world?" (outbound buyers need problem awareness built, not assumed.)

Follow-up ladder (Sandler Pain Funnel ordering, published by a vendor; the ordering vague → specific → duration → attempts → cost → emotion is the transferable part):

1. "Tell me more about that."
2. "Can you be more specific? Give me an example."
3. "How long has that been a problem?"

## Root cause

- "Why do you think that keeps happening?"
- "What have you tried to do about it?" then "Did that work?" (Sandler funnel rungs.)
- "What would you say is underneath that - people, process, or the tooling?" (closed follow-up, only after the open version.)
- Listen-for: a cause stated as a symptom ("we're just too busy") - one more rung down before accepting it.

## Impact

Quantify on at least two of the three GAP Selling dimensions. Which dimension gets the next question, when the call cannot fund all three:

- efficiency: `business > personal > technical`
- value: `business > technical > personal`
- effort: `technical > business == personal`

- Business and personal tie on effort: each costs one question and one follow-up asked inside a stage the buyer is already in, the same minute either way, and the same small draw on their patience.
- Business wins efficiency: its answer is the figure the close, the recap and every downstream deal skill run on.
- Technical costs more than both on effort: it pulls the buyer into explaining a system, and most of what it returns the root-cause stage already surfaced.

Starved: technical impact, genuinely high value (it is the only dimension an architect or a security reviewer can answer with a real number) but it loses every round on effort with anyone else.

- Promote it to first when the persona _is_ the technical evaluator, or when the pain is an outage, a reliability problem, or a data loss.
- Delete it, rather than compressing it, on a short call: two funded dimensions beat three rushed ones, and the gate wants two.

This ordering is a default, not a law; re-rank it against the persona in front of you before drafting. B2C runs `personal > business > technical` - household time, money and stress _is_ the business case there.

- Technical: "When that fails, what does it do to everything downstream?"
- Business: "What do you think that costs you per month - time, money, or deals?" (Sandler cost rung, merged with figure-first framing.)
- Business: "If [necessary task] doesn't happen, what's the impact on [team]'s numbers?" (SPIN Implication restatement.)
- Personal: "What does that mean for you and your team, week to week?"
- Personal / emotional weight: "How do you feel about what this has already cost you?" (Sandler funnel rung.)
- Spread: "Why do you want it, and who else does it impact?" (Samantha McKenna, via a vendor roundup.)
- Framing note, a vendor claim: mentioning the seller's ROI math during discovery correlates with worse next-step rates; cost-of-inaction framing - their number, their loss - performs better.

## Urgency / compelling event

MEDDPICC's Three Whys: "Why anything? Why you? Why now?" - discovery owns "why anything" and "why now"; "why you" is positioning, later.

- "What happens if this isn't solved in the next six months?" (cost-of-inaction form.)
- "What's changed recently that makes this worth looking at now?" (hunts the trigger event.)
- "Where does fixing this sit on this quarter's priority list?"
- "Is there a date this becomes a real problem - renewal, launch, audit, season?"
- Caveat: if nothing is found, record "no compelling event" honestly. Manufactured urgency deepens buyer indecision (Dixon & McKenna's published buyer-indecision research, their own claim).

## Decision context (complex B2B only - questions, never scores)

Gather these to feed the scorecard skill; never output a qualification verdict here.

- "If you decided this was worth solving, what would the path to a yes look like?" (decision process.)
- "Who else would want a say - and who could veto it?"
- "What would you need to see to know it worked?" (metrics.)
- "Have you evaluated anything for this before? What happened?" (competition incl. do-nothing.)
- Regulated buyers: add "what does security / compliance / procurement review look like for a purchase like this?"

## Close

- Summarize in their words: "So if I've got it - [pain], because [root cause], costing you [their figure]. What did I miss?"
- "Given that, would [specific next step] on [date] make sense?" (dated, specific, tied to the impact just quantified.)

## B2C high-ticket variants

- Pain: "What finally made you book this conversation?" (the trigger event, asked first.)
- Root cause / prior attempts: "What's kept you from doing this until now?"
- Personal impact leads: "What would getting this handled change for you day to day?"
- Household-business impact: "What is living with it costing you - time, money, stress?"
- Couples: ask pain and impact to each person separately - "what do you each most want out of this?" Two decision makers, two spines.
- Urgency: "Why now, after [duration they gave]?"
- Regulated categories (financial advice, health): flag phrasing constraints to the user rather than guessing the rules.

## Do-not-ask list

Put these on every call sheet as written warnings:

- "What's the timeline for this project?" and "Is there budget allocated for this?" as openers - negative examples published by GAP Selling practitioners: they harvest logistics before establishing that a problem exists. Both become legitimate late, after impact.
- Hypothetical purchase questions - "would you use / buy / pay for X?" Answers are politeness, not data; ask about past behavior instead.
- Leading questions - "that must be costing you a fortune, right?" The prospect's number persuades; the rep's number gets argued with.
- Compound questions - "what's broken and what have you tried?" Ask one, ladder to the other.
- Filler check-ins - "does that make sense?" - and any question asked to fill silence rather than to learn.

# Worked examples

Rewrite specifics for the actual deal - never reuse verbatim across deals.

## Table of Contents

- [Example 1 - B2B: AE discovery, 30 min, outbound](#example-1---b2b-ae-discovery-30-min-outbound)
- [Example 2 - B2C high-ticket: kitchen remodel consult, 60 min, couple (abbreviated)](#example-2---b2c-high-ticket-kitchen-remodel-consult-60-min-couple-abbreviated)
- [Example 3 - negative: the logistics-first sheet (do not build this)](#example-3---negative-the-logistics-first-sheet-do-not-build-this)

## Example 1 - B2B: AE discovery, 30 min, outbound

Context given:

- VP Operations at a 200-person logistics company.
- Replied to a cold email about shipping errors.
- Product is order-orchestration software.
- Mid-market, ~4 stakeholders expected.
- Success = a scoped demo with her ops lead present.

```
DISCOVERY CALL SHEET
Prospect   : VP Operations, NorthHaul Logistics (200p)   Call: AE discovery, 30 min
Origin     : outbound - replied to cold email on shipping errors
Known      : runs 3 warehouses; hiring a "fulfillment systems analyst" (job post);
             replied "this is a constant headache" - do not re-ask whether errors happen
Next step  : scoped demo with her ops lead, within 2 weeks

HYPOTHESES
H1: mis-ship rate is driven by manual re-keying between OMS and WMS
H2: peak-season temp staffing breaks their picking process
H3 (fallback): reporting lag hides errors until customers complain

FRAME
"You mentioned errors are a constant headache - fair if we spend 25 minutes on
where they actually come from, then decide together if a deeper look makes sense?"

QUESTION PATH
[Situation - 1 only; site + job post answered the rest]
Q: Walk me through what happens between an order landing and a label printing.
   listen for: manual handoffs, swivel-chair steps   exit: process in her words

[Pain - H1]
Q: You called errors a constant headache - when did one last really hurt?
   ladder: tell me more / can you give me a specific example / how long has it
   been like this
   listen for: a named incident with a customer attached
   exit: specific problem + recent example

[Root cause]
Q: Why do you think that one happened?
   ladder: what have you tried so far / did it work
   listen for: "we added a checking step" = symptom patch, go one rung deeper
   exit: cause ≠ symptom; prior attempts known

[Impact]
Q: What do mis-ships run you per month - reship cost, credits, anything else?
Q: What does it do to your team when a big one lands - who gets pulled in?
   listen for: her figure, not mine; names of other stakeholders
   exit: one quantified figure + who else is touched

[Urgency]
Q: What's changed that made you answer a cold email about this now?
Q: If nothing changes before peak season, what does November look like?
   exit: dated/triggered reason - or noted absence

[Decision context]
Q: If you decided to fix this properly, what would the path to a yes look like?
Q: Who else would want a say - and who could kill it?
   exit: process + voices sketched (feeds scorecard; no score)

CLOSE
Summary: "So - mis-ships from re-keying between systems, patched with a checking
step that didn't hold, costing about <her figure>/month and worst in peak. Miss
anything?"
Ask    : "Worth a 45-minute demo scoped to that flow, with your ops lead - week
after next?"

BRANCHES
- H1 dead → pivot to H2 via "where does peak season strain the process most?"
- She volunteers the cost figure early → bank it, return for root cause
- Her boss (COO) joins → drop situation entirely, open with H1 as a POV statement
- Asks pricing → range in one sentence, then "depends on the flow - can I ask two
  more questions about it?"

DO NOT ASK
- "What's your timeline?" / "Is there budget?" before impact
- "Would you use a tool that...?" - the product stays out until the close
- "Errors must be killing your margins, right?" - her number, not mine

AFTER THE CALL
Stages unmet: ______________________
```

## Example 2 - B2C high-ticket: kitchen remodel consult, 60 min, couple (abbreviated)

Context: in-home consultation; two decision makers; success = a paid design agreement.

- Frame: "Let's spend the hour on what you want this kitchen to do for you, and by the end decide together whether a design plan is the right next step - fair?"
- Pain, asked to each: "What finally made you book this visit - and was it the same thing for both of you?" Ladder: specific example ("when did the kitchen last really frustrate you?") / duration.
- Root cause / prior attempts: "What's kept you from doing it until now?" - listen for money vs disruption vs disagreement; each has a different close.
- Impact, personal leads: "What would a finished kitchen change about an ordinary Tuesday night?" Household-business: "What is waiting costing you - hosting, resale timing, workarounds you pay for?"
- Urgency: "Why now, after eight years?" - listen for a trigger (event, sale, milestone). No trigger found → note it honestly; don't invent a deadline.
- Close: summarize each person's answer in their own words, then: "The next step would be a design agreement so you can see it before you commit to a build - does <date> work to review it together?"
- Branch: partners disagree on pain → surface it as a question ("you two see this differently - which matters more this year?"), never pick a side.

## Example 3 - negative: the logistics-first sheet (do not build this)

```
1. What's your timeline for this project?
2. Is there budget allocated for this?
3. Who's the decision maker?
4. What tools are you using today?
5. Would you be open to a demo next week?
```

Why it fails, line by line:

- Lines 1-2: the two canonical logistics-first offenders (published GAP Selling negative examples). They harvest deal admin before establishing that a problem exists, and signal the rep sells to a quota, not a gap.
- Line 3: an authority question with no pain context reads as "are you worth my time". The same information falls out naturally from the decision-context stage after impact.
- Line 4: a situation question research should have answered.
- Line 5: a close with nothing to close on, no pain, no cause, no figure, no why-now - the demo, if accepted, is a tour, not a next step.

Every question here becomes legitimate late in the spine. The failure is sequence, not wording - which is why the quality gate checks order, not just phrasing.

---
name: sales-discovery-questions
description: Builds a sequenced sales discovery question set - pain, root cause, impact, urgency - for a specific deal or persona, sized to the call length, with follow-up ladders, branch triggers, and a do-not-ask list. Covers B2B (SDR qualification through multi-stakeholder AE discovery) and high-ticket B2C consultations. Use whenever the user mentions discovery call prep, qualification questions, SPIN or pain-funnel questions, a first-meeting agenda, or "what should I ask this prospect", even without the word discovery. Do NOT use for rebuttals (mbfinotti/sales-skills@sales-objection-handling), the cold-call opening (mbfinotti/sales-skills@cold-call-opener), or grading a finished call (mbfinotti/sales-skills@sales-call-review).
license: MIT
metadata:
  author: Maya-Beth Finotti
  version: "1.0.7"
---

# Discovery Questions

Build the sequenced question set a rep carries into a sales discovery conversation, plus the follow-up ladders and branch logic to adapt it live.

Four questioning frameworks with a publicly documented structure converge on one spine:

- SPIN (Rackham, 1988)
- Sandler's Pain Funnel
- MEDDPICC's Three Whys
- GAP Selling's Problem / Root Cause / Impact chart

The spine: isolate a real problem, dig to its root cause, quantify the impact, then establish why now. This skill encodes that convergence, not any one acronym. Vendor question-count and talk-ratio benchmarks are directional at best - never set a number to hit.

## Interview

Ask before drafting. One question per message; offer multiple-choice answers when possible; skip anything the user already answered or that their notes make obvious.

- B2B, or B2C? (B2C here means considered, high-ticket consumer sales - coaching, remodeling, financial advice - not impulse retail.)
- Who is the persona, and how senior? (End user, manager, executive - seniority changes question count and depth.)
- What does the product do, and what problem does it solve, in one sentence each?
- What call is this: SDR qualification (15-30 min), AE deep discovery (30-45 min), persona-specific follow-up in a running deal, or a single B2C consultation? How long?
- Deal size and cycle complexity? (Transactional single-buyer / mid-market few stakeholders / enterprise buying committee.)
- Inbound or outbound origin? (Inbound buyers arrive with intent to confirm and deepen; outbound buyers need problem awareness built first.)
- What is already known from prior touches - outreach replies, notes, a prior call, a deal record?
- Does the team mandate a methodology (SPIN, MEDDPICC, Sandler, GAP, other)? If so, the set must map to its vocabulary.
- Is the buyer in a regulated context (healthcare, finance, government)? (Adds decision-process and procurement questions; constrains what can be asked about data and finances.)
- What pain does the rep already suspect? (Becomes the hypothesis to test - never a fact to confirm.)
- What next step would make this call a success? (The set must close toward it.)
- By what date must this call have produced a booked next step? (A hard date promotes the impact triad - a figure won on this call is what a dated ask is argued from - and demotes decision-context coverage, which pays back across the cycle, not today.)
- One-off win or compounding asset: a sheet for this one call, or a reusable persona template the team reruns? (Compounding promotes decision-context coverage and the full follow-up ladder, since both amortize across calls and personas; one-off keeps the bare spine.)
- Effort ceiling: how many of the call's minutes are yours to spend on questions, how much rehearsal will the rep actually do, and how much of the buyer's patience can you spend? (A low ceiling deletes the decision-context stage outright and caps every ladder at two rungs.)

## Workflow

1. Run the Interview; collect every answer before writing a question.
2. Gather context. If you can browse the web, research the company and persona; otherwise work from the user's notes. Never fabricate account facts - research answers situation questions so the call doesn't have to.
3. Write 2-3 candidate pain hypotheses from the inputs, each stated as a hypothesis to test on the call. Confirm with the user which to prioritize.
4. Pick the coverage lens layered on the spine below. Call minutes are the budget every lens is priced in, and the buyer's patience is spent alongside them - a lens that costs more goodwill is expensive even though it is free to the rep.
   - efficiency: `GAP > Sandler > SPIN > MEDDPICC`
   - value: `GAP > MEDDPICC > Sandler > SPIN`
   - effort: `MEDDPICC > Sandler == GAP > SPIN`
   - Default GAP: about an hour of per-deal rehearsal and two impact questions - buys the figure in the prospect's own words that the close, the recap, and every downstream deal skill run on.
   - Sandler: costs the same hour and the same handful of minutes inside a stage already open, the genuine tie on effort with GAP - buys depth on one pain instead of a number, so it ranks second.
   - SPIN: near-zero on both axes - it only re-sequences questions already written, and the spine has absorbed its ordering. Take it free, never budget for it.
   - Starved: MEDDPICC's decision context. High value in a real buying group, highest effort - a whole extra stage, re-gathered per persona across the cycle - so efficiency buries it, and reps ship sheets that quantify a pain nobody can approve. Promote it to first whenever the buying group runs past a single economic buyer, or the buyer is regulated. Gather its elements as questions only; scoring belongs to the scorecard skill.
   - Delete, never demote:
     - A mandated methodology removes the other three from the sheet and maps every stage to its own vocabulary.
     - A transactional or B2C deal removes the decision-context stage outright.
     - A ruled-out lens parked at the bottom reappears as scope mid-call.
   - Re-rank before writing a question - this order is a default, not a law, and it moves with who runs the call and who answers:
     - Fifteen SDR minutes leave only the top of the line.
     - An hour with a practitioner promotes Sandler to par.
     - An executive promotes whichever lens reaches a figure fastest and deletes the rest.
5. Size the set to the call: roughly one substantive question plus its follow-ups per 3-4 minutes of questioning time; fewer and deeper for executives. Spend those minutes in step 4's order - the lens at the head of the efficiency line gets its questions first, and whatever the budget cannot fund is cut from the tail, never trimmed evenly across all four. Published "ideal question counts" are correlational and internally inconsistent, so size by call length, never a magic number.
6. Draft from [references/question-bank.md](references/question-bank.md), sequenced per the spine, one stage at a time. Attach a 2-3 rung follow-up ladder and a listen-for note to each primary question.
7. Add branch triggers (below) and the do-not-ask list from the question bank.
8. Assemble into the call sheet from [references/call-sheet-template.md](references/call-sheet-template.md).
9. Run the Quality gate; iterate - rewrite and re-check - until every item passes.
10. Deliver, and remind the user: the sheet is a map, not a script - follow the prospect's answers, use the branches.
11. If your harness has persistent memory, memorize the persona, the hypotheses that proved true, and which questions produced signal - the next set for the same persona starts warmer. Otherwise hand the user a short recap to keep.

## The pain → impact → urgency spine

Every question belongs to a stage. Every stage has an exit condition. Movement is one-way - never retreat to logistics before the spine is walked.

| Stage            | Job                                                                                | Move on when                                              |
| ---------------- | ---------------------------------------------------------------------------------- | --------------------------------------------------------- |
| Frame            | One-line agenda agreement (Sandler's upfront contract, Sandler's own concept)      | Purpose, time, and possible outcomes agreed               |
| Situation        | Minimal context you could not research - 1-2 questions max                         | You can describe their current process in their words     |
| Pain             | Surface a real problem behind the hypothesis                                       | A named, specific problem with a recent example           |
| Root cause       | Why it happens and what they already tried                                         | Cause distinguished from symptom; prior attempts known    |
| Impact           | Quantify across technical, business, and personal dimensions (GAP Selling's triad) | At least one figure, plus who else it touches             |
| Urgency          | Why now - the compelling event                                                     | A dated or triggered reason to act, or its honest absence |
| Decision context | (Complex B2B) how a decision would actually get made                               | Process, other voices, evaluation path sketched           |
| Close            | Prospect's-words summary, then a dated next step                                   | Specific next step proposed against the impact uncovered  |

Sequencing rules - the logic the whole set obeys:

- Open before closed. Open the stage with an open-ended question; use closed questions only to confirm or quantify what was said. Approach discovery "with a learning mindset, not a selling one" (Chris Riley, via a vendor roundup).
- Layer before advancing: every meaningful answer earns 2-3 follow-ups before a new topic. Sandler's Pain Funnel (published by a vendor; Sandler's own pages don't show it) orders the ladder: vague → specific example → duration → prior attempts → cost → emotional weight. That ordering, not the exact wordings, is the transferable part.
- Root cause before impact. "You can't solve a problem you don't understand" (Keenan). Quantifying a symptom's cost quantifies the wrong thing - and "pain can be misleading" (Keenan).
- Impact before urgency. Urgency asked before impact is quantified is just pressure - and published buyer-indecision research (Dixon & McKenna's JOLT work, their book's own claim) found pressure deepens indecision rather than resolving it.
- One idea per question. Never compound, never stack. A stacked question gets the answer to its easiest half.
- Past-specific over hypothetical. "When did that last happen, and what did it cost?" beats "what would it cost if...". Answers about real events are data; answers about imagined ones are politeness.
- Their words, not the product's. Questions stay product-agnostic until the close; the product enters only after the gap is established. "Never sell to need" (Keenan).
- Quantify with their number. "Value is a figure, not an adjective" (Command of the Message, Force Management). Let the prospect say the number; a number the rep supplies is an objection waiting to happen.

Branch triggers - the live-adaptation half of the deliverable, written into every call sheet. These carry no ranking and never get one: each fires on its own condition rather than competing with the others for the same minute, so ordering them would invent a precision the call does not have.

- No pain surfaces after two hypotheses → switch hypothesis, or ask the honest fallback: "What would have to be true for this to be worth your time?". Do not manufacture pain.
- Prospect volunteers impact early → skip forward, bank it, return for root cause. The spine orders exit conditions, not scripts.
- Executive in the room → halve the question count, lead with the sharpest hypothesis, trade situation questions for a point-of-view statement. "Top-performing reps ask fewer questions during discovery calls" (Gong), and its published rules-of-thumb explicitly break down for C-suite buyers.
- Answers stay surface-level → drop down the Sandler ladder one rung at a time rather than moving on; a stage exited without its condition met is a gap the recap email will expose.
- Prospect asks for price or a demo mid-spine → answer briefly, then trade: one more stage before going deeper. Premature pitch is the top failure mode below.

## Invocation and expected output

Typical invocations:

- "Build my discovery question set for tomorrow's 30-minute call with a VP Ops at a 200-person logistics company - outbound, she replied to a cold email about shipping errors."
- "I sell kitchen remodels; couples book a 60-minute in-home consult. Give me the question set."
- "SDR qual call, inbound demo request, IT manager at a mid-market accounting firm - 15 minutes."

Deliver one call sheet (full template in [references/call-sheet-template.md](references/call-sheet-template.md)):

```
DISCOVERY CALL SHEET - <persona> at <company>, <call type>, <length>
Hypotheses     : 2-3 suspected pains, ranked - to test, not confirm
Frame          : one-line upfront agenda
Question path  : per spine stage - primary question, 2-3 follow-up rungs,
                 listen-for note, exit condition
Branches       : if no pain / if exec / if impact early / if price asked
Do-not-ask     : leading, logistics-first, and hypothetical questions to avoid
Close          : summary prompt + the dated next-step ask
After the call : which stages went unanswered → gaps for the next touch
```

## B2B and B2C

- Same for both (the frameworks converge here, and none of them splits the spine by business model):
  - The spine itself
  - The sequencing rules
  - The follow-up ladders
  - The quality gate

  What changes is who answers and which impact dimension leads.

- B2B: buying groups run large - a Gartner survey (2017, n=750, Gartner's own figure) puts the median at 6-10 decision makers - so discovery recurs per persona. Generate persona-specific variants (workflow pain for end users, business outcomes for the economic buyer, architecture and security for technical evaluators); route stakeholder strategy itself to the deal-champion-mapping skill.
- B2B: decision-context questions are load-bearing in complex deals; in regulated buyers, procurement and compliance path questions expand and personal-finance-adjacent questions disappear.
- B2C high-ticket: one conversation, one or two decision makers - usually the only shot at the full spine, so the close matters more than coverage breadth. Business impact becomes household time, money, and stress, which is why the question bank re-ranks the impact dimensions here. With a couple, ask each person the pain and impact questions separately - two decision makers means two spines.
- B2C: prior-attempts questions ("what's kept you from doing this until now?") carry the root-cause stage; regulated categories (financial advice, health) constrain phrasing - flag for the user, don't guess the rules.

## Quality gate

Score the drafted set against all ten items. Pass threshold: 10/10. Iterate until nothing fails.

1. Every question maps to exactly one spine stage; no orphans, no stage empty except a justified skip (e.g. decision context on a first SDR call).
2. Each stage opens open-ended; closed questions only confirm or quantify.
3. Root cause is probed before impact; impact carries at least one quantifying question before any urgency question appears.
4. Impact questions cover at least two of technical / business / personal.
5. No leading questions (the desired answer embedded in the wording) and no compound questions anywhere on the sheet.
6. Every question that can be past-specific is: it references a real, recent instance, not a hypothetical.
7. Budget, timeline, and process logistics appear only after pain and impact - never as openers. The two canonical offenders ("What's the timeline for this project?", "Is there budget allocated?") are on the do-not-ask list, negative examples published by GAP Selling practitioners.
8. The set fits the stated call length at roughly one primary question plus ladder per 3-4 minutes, reduced for executives; no target question-count is cited as a rule.
9. Every question could change what the rep does next. A question whose answer alters nothing is discovery theatre - cut it.
10. The sheet ends with a prospect's-words summary prompt and a dated next-step ask, and carries branch triggers and the do-not-ask list.

## KPIs and measurement

- Per call, the sheet worked if:
  - 3-4 distinct problems surfaced, with at least one quantified in the prospect's own number.
  - Every exited stage met its exit condition.
  - A dated next step was accepted.

  A recap email writable from the answers alone (route writing it to the sales-meeting-recap skill) is the honest test - unanswerable sections are the discovery gaps.

- Across calls, track first-call → next-meeting (or → opportunity) conversion per persona and per question set version. Movement there, not any talk-ratio target, is the signal - published talk-listen and question-count benchmarks are vendor claims, correlational, and internally inconsistent across the vendor's own refreshes (details in [references/frameworks-and-evidence.md](references/frameworks-and-evidence.md)).
- If a conversation-recording tool is available, review which questions produced long, specific answers and which produced one-liners; feed that back into the next version of the set. Without one, the rep's after-call gap notes on the sheet serve the same loop.

## Common failure modes

| Failure                                                           | Fix                                                                                                                                                                 |
| ----------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Interrogation feel - rapid-fire list, no layering                 | Fewer primaries, deeper ladders; react to answers ("you said X - tell me more") instead of jumping topics                                                           |
| Happy ears - hearing confirmation of the hypothesis in everything | Hypotheses are written to be tested; require a recent, specific example before a pain counts as found                                                               |
| Surface-level pain - a complaint mistaken for a problem           | Walk the ladder: specific example → duration → prior attempts → cost; a pain without a story is a symptom                                                           |
| Unquantified impact                                               | No urgency questions until at least one figure exists, in the prospect's number, on ≥2 impact dimensions                                                            |
| No compelling event                                               | Ask why-now directly (MEDDPICC's Three Whys: "Why anything? Why you? Why now?"); record its honest absence rather than inventing one - absence is deal intelligence |
| Leading questions                                                 | Strip the embedded answer; "how do you feel about that cost?" not "that must be expensive, right?"                                                                  |
| Premature pitch - demo or product talk mid-spine                  | Product enters after the gap is established; brief answer, then trade back into the spine                                                                           |
| Logistics-first questions - budget and timeline as openers        | They harvest logistics before a problem exists; move them behind impact, per the gate                                                                               |
| Discovery theatre - questions whose answers change nothing        | Cut any question with no downstream consequence; coverage is measured in exit conditions, not questions asked                                                       |

## Reference

- See [references/question-bank.md](references/question-bank.md) for the stage-by-stage question bank with methodology attributions, follow-up ladders, B2C variants, and the do-not-ask list.
- See [references/call-sheet-template.md](references/call-sheet-template.md) for the fill-in-ready call sheet.
- See [references/worked-examples.md](references/worked-examples.md) for a complete B2B call sheet, an abbreviated B2C consult sheet, and an annotated logistics-first negative example.
- See [references/frameworks-and-evidence.md](references/frameworks-and-evidence.md) for the four verified frameworks in efficiency order, what each costs and buys as a coverage lens, and the evidence caveats behind the vendor benchmarks.
- See `mbfinotti/sales-skills@sales-objection-handling` for rebuttals when a question surfaces resistance.
- See `mbfinotti/sales-skills@cold-call-opener` for the opening script before discovery starts.
- See `mbfinotti/sales-skills@sales-call-review` for grading the call transcript afterwards.
- See `mbfinotti/sales-skills@meddpicc-scorecard` for scoring the deal against the elements these questions gather.
- See `mbfinotti/sales-skills@sales-meeting-recap` for turning the answers into the recap email.
- See `mbfinotti/sales-skills@sales-outreach-personalization` for the pre-call research that precedes this skill.