mbfinotti/advertising-skills검사 통과
SKILL DETAIL
advertising-career
mbfinotti/advertising-skills/advertising-career
Plan and advance a paid-advertising career from the candidate side - media buyer, performance marketer, PPC specialist, creative strategist, growth marketer - covering the junior to lead ladder, the five real interview formats (live account audit, mental-maths case, metric diagnosis, take-home, behavioural), an NDA-safe results portfolio, platform certifications, in-house vs agency vs freelance, and promotion and pay conversations. Use whenever the user asks how to get into paid media or PPC, preps a media-buyer or growth interview, wants an ad portfolio, or weighs an agency vs in-house move - even if they never say 'career'. Do NOT use for the employer side of hiring - use mbfinotti/advertising-skills@advertising-hiring instead.
설치 수 · 178출처 보기
Installation
npx skills add https://github.com/mbfinotti/advertising-skills --skill advertising-career
스킬 파일
SKILL.md
최근 동기화 · 2026. 9. 24.
evals/evals.json›
{
"skill_name": "advertising-career",
"evals": [
{
"id": 1,
"prompt": "I'm a senior paid media manager in Manchester interviewing at a retail brand. Final round is next week and they've asked for my salary expectations. I found a 2025 salary survey saying senior marketing managers average around £72,000 in the UK, so I'm planning to anchor at £75,000. Does that sound right, and how should I phrase it?",
"expected_output": "A negotiation prep that corrects the stale 2025 anchor with the 2026 survey-grade decline data, grades and dates every figure, and rebuilds the anchor as a sourced range.",
"files": [],
"expectations": [
"Warns that quoting a 2025 figure overstates the current market because the strongest UK survey shows marketing pay declining year-on-year",
"Cites the Marketing Week 2026 Career & Salary Survey as the survey-grade UK benchmark",
"States the 2026 senior manager figure of £60,577",
"Contrasts the 2026 figure with the 2025 senior manager figure of £72,434 to show the decline",
"Names the survey's respondent count (2,350) or its April 2026 date when grading the source",
"Labels every quoted pay figure with currency, source, date, and grade",
"Distinguishes survey-grade sources from aggregate or job-board estimates, treating aggregates as wide ranges",
"Presents benchmarks as ranges, never a single point figure",
"Explicitly advises against anchoring at £75,000 on the basis of the 2025 number as found",
"Recommends reverifying the figure before it anchors the negotiation, or dating and grading it out loud if verification is not possible",
"Offers the 3Search Performance Marketing Salary Guide 2025 (1,300+ responses, £50,000-£70,000 for PPC Managers) as an additional survey-grade anchor",
"Warns that an undated or last-year figure quoted in a negotiation damages the candidate's credibility"
]
},
{
"id": 2,
"prompt": "I've got a second-round interview for a performance marketing manager job at an online furniture retailer. They said one discussion topic is privacy and measurement. Here's my drafted answer: 'With Google deprecating third-party cookies in Chrome, marketers must prepare for a cookieless future by investing in first-party data.' Can you polish this and help me expand it into a strong two-minute answer?",
"expected_output": "A rewrite that corrects the cookie-deprecation myth with dated evidence and rebuilds the answer around ATT, over-attribution, and the platform-ROAS versus blended-MER divergence.",
"files": [],
"expectations": [
"Corrects the premise: third-party cookies are not being deprecated in Chrome",
"Cites Google's July 2024 announcement that it would not phase out third-party cookies",
"Cites the April 2025 confirmation that Chrome would not add a user-choice prompt",
"States that cookies remain enabled by default in Chrome while other browsers already block them",
"Warns that repeating the deprecation claim in an interview dates the candidate",
"Redirects the answer to Apple's ATT (2021) as the change that durably reduced signal",
"Explains that ATT drove the over-attribution problem and the industry move to blended metrics and incrementality",
"Includes the senior-level framing: being able to explain why platform-reported ROAS diverges from blended MER",
"Notes the autobidder optimises toward the conversions it can see, which is exactly the number the OS-level privacy change broke",
"Does not polish or expand the cookieless-future claim as drafted; the false premise is replaced, not embellished"
]
},
{
"id": 3,
"prompt": "I'm 29, spent six years as a retail store manager, and want to become a media buyer. My plan for the next five months: get Google Ads certified, then Meta Blueprint (I've budgeted $450 for three Blueprint exams), then the TikTok, Amazon and LinkedIn certifications, then start applying with all six on my CV. Good plan?",
"expected_output": "A corrected entry plan that demotes certification-stacking to table stakes, cancels the paid Blueprint exams, and redirects the five months toward demonstrable results via an agency-seat-plus-own-spend path with a dated change-course benchmark.",
"files": [],
"expectations": [
"States that platform certifications are table stakes with weak standalone hiring signal",
"States that the whole certification menu ranks below one published teardown or equivalent demonstrable result",
"Advises against the $450 Blueprint budget: Meta Blueprint (~$150 per exam) is only worth paying when a target listing names it or an employer's partner status depends on it",
"Identifies Skillshop, LinkedIn Marketing Labs, TikTok Academy, and Amazon Ads as free",
"Gives Skillshop specifics: open-book exam of about 75 minutes and a 12-month expiry with the renewal diarised",
"Explains why listings demand certifications anyway: individual certifications roll up into an agency's partner-program status (badge, beta access, rep support), not because they predict candidate quality",
"Warns that a certification-only CV is a documented rejection pattern",
"Redirects the bulk of the five months to demonstrable results: self-funded spend and/or a public written teardown",
"Recommends chasing an agency junior seat as the default entry path",
"Recommends running self-funded spend alongside applications, noting no credible-minimum budget exists because the full loop matters, not the size",
"Warns against modelling the plan on inflated self-taught success stories such as '20x ROAS' claims; a small honest documented project beats an inflated one",
"Sets the dated benchmark-to-change-course: no interviews after roughly 6 months of applications means the gap is demonstrable results, not more certifications",
"Breaks the TikTok Academy versus Amazon Ads priority by the user's channel interest (short-form video versus retail media) or asks which channel they target"
]
},
{
"id": 4,
"prompt": "Four years at an ad agency, everything under client NDAs. I have a growth marketer interview in two weeks and no portfolio. My plan: I kept screenshots of the Meta ads manager from my best account - Fairwind Coffee Roasters - showing spend at $62K/month and ROAS going from 1.6 to 3.4. I'll blur the logo, put the screenshots on a Notion page with the full story, and walk the interviewers through it. Can you help me write it up?",
"expected_output": "A rebuilt NDA-safe portfolio plan that rejects the screenshot approach, anonymises and indexes the client work, checks the contract first, and adds a public teardown as the zero-risk lead piece.",
"files": [],
"expectations": [
"Rejects using raw ads-manager screenshots even with the logo blurred, naming them a legal risk",
"Flags naming the real client as an NDA breach and replaces it with a descriptor that carries the credibility (e.g. 'DTC coffee brand')",
"Converts the absolute figures ($62K/month spend, real ROAS values) into percentage change, indexed values (baseline = 100), or aggregated ranges, never absolutes",
"Applies anonymise, index, and strip-screenshots together as a checklist, not as a pick-one menu",
"Advises re-reading the NDA or client contract before publishing anything client-derived, because a breach cannot be undone",
"Warns that over-sharing confidential client data reads to interviewers as a preview of how the candidate would treat their data",
"Recommends a public written teardown of a brand's ads as a lead or companion piece because it carries zero confidentiality risk",
"Notes the teardown doubles as rehearsal for the live-audit interview format",
"Structures the case study as problem, process, solution, indexed results, learnings",
"States that the process section carries the hiring signal: it shows judgment, not access to budget",
"Includes a learnings element naming something that did not work",
"Includes an explicit statement in the deliverable that details are anonymised and figures indexed",
"Never invents figures to fill the template; every number is the candidate's own evidence in NDA-safe form"
]
},
{
"id": 5,
"prompt": "Just got a take-home from Lumenmatic, a smart-lighting startup, for a senior paid media role. The brief says 'this should take a few hours': build the full go-to-market plan for their unlaunched pro-installer product line - channel strategy, budgets, audience research, creative concepts, a launch calendar and a 12-slide deck. I've blocked out both weekend days plus two evenings, around 20 hours. What order should I tackle it in?",
"expected_output": "A response that refuses to schedule the 20 hours, time-boxes the work to 2-4 hours, scopes down the go-to-market brief with an explicit push-back, and states assumptions and the time-box in the deliverable.",
"files": [],
"expectations": [
"Challenges the 20-hour plan instead of sequencing it",
"Recommends time-boxing the take-home to 2-4 hours",
"Cites the documented pattern: companies brief 'a few hours' while candidates report 12-30+ hours",
"Flags the brief as a full go-to-market plan for an unlaunched product, the exact shape to push back on",
"Offers a scoped alternative (e.g. focus the hours on the channel and measurement plan and name what a full plan would additionally need)",
"Advises stating the time-box explicitly in the deliverable",
"Advises stating every assumption explicitly in the deliverable",
"Advises leading with reasoning rather than production polish, deprioritising the 12-slide deck",
"Mentions the documented idea-theft risk around unpaid take-home work",
"Frames the take-home as the one interview format whose effort the candidate controls, with the freed hours redirected to higher-value rehearsal such as the live audit",
"Does not produce or outline a full go-to-market plan sized to the user's 20-hour scope"
]
},
{
"id": 6,
"prompt": "Final round Thursday at a performance agency for a paid search manager role. It's a 45-minute screen-share where I audit a live Google Ads account in front of two of their leads. My plan is to run the account through my current agency's audit tool - it generates a 40-point checklist report - and walk them through the output. How should I structure the walkthrough?",
"expected_output": "Live-audit preparation that abandons the tool-report walkthrough for a prioritised spoken structure: orient, money first, structure and signal, top-three prioritised recommendations with proof metrics, and a closing data-request, rehearsed aloud.",
"files": [],
"expectations": [
"Warns that presenting a generic tool-generated audit is the named failure interviewers screen for in this round",
"Recommends bringing a prioritised point of view instead of a checklist walkthrough",
"Structures the audit to orient first: business model, conversion action, and what 'good' means for the account before touching any setting",
"Puts money first: where spend concentrates, whether the top spenders hit target CPA or ROAS, and the single biggest waste line",
"Covers structure and signal: campaign and ad-group logic, conversion-tracking sanity, and bidding fit to the goal",
"Ends with a prioritised recommendation of the top three changes ordered by expected impact",
"Attaches to each recommended change the metric that would prove it worked",
"Closes with the data the candidate could not see and would request, framed as rigor rather than weakness",
"Recommends rehearsing the audit aloud on real accounts before the interview",
"Limits the recommendations to a prioritised top three rather than reordering the tool's 40 checklist points"
]
},
{
"id": 7,
"prompt": "I have a scenario round tomorrow for a paid social manager job at a DTC supplements brand. The recruiter shared past questions: (1) 'Account ROAS dropped 30% over the last three weeks - walk us through your diagnosis.' (2) 'An A/B test shows a 5% lift at p=0.08 - what do you do?' (3) 'We doubled a winning campaign's daily budget from $2K to $5K and ROAS fell while CPM nearly doubled - why?' Help me draft strong answers.",
"expected_output": "Three drafted answers: an ordered ROAS-drop diagnostic (creative, saturation, site-side, attribution) with its rationale and a business-metric close, a decision-frame answer to the p=0.08 question, and audience saturation for the budget-doubling question.",
"files": [],
"expectations": [
"The ROAS-drop answer opens by scoping the comparison window, noting a drop measured over a week or less is often noise",
"The ROAS-drop diagnostic is presented as an ordered sequence, not an unordered list of possibilities",
"Creative health comes first, citing fatigue as the cause roughly 60-70% of the time, checked via frequency and the age of the top-spending creatives",
"Audience saturation comes second, identified by rising CPM against flat reach",
"Site-side conversion comes third: landing page, offer, price, or checkout changes the ad account cannot fix",
"Attribution or tracking accuracy comes last, with the note that a reporting artefact is not a performance drop",
"States the ordering's rationale: hit rate per minute of checking, not likelihood alone (in-account checks take minutes; site-side is a cross-team ask)",
"Closes the diagnosis with the business metric that decides it: blended MER or contribution, not the platform dashboard alone",
"The p=0.08 answer refuses the binary significant-or-not framing and answers with a decision frame",
"The decision frame weighs the cost of being wrong in each direction: cheap-to-keep with small downside means ship and keep measuring, expensive or irreversible means extend the test",
"Notes that naming the threshold as something that should have been agreed before the test reads as senior",
"The budget-doubling answer names audience saturation or diminishing marginal returns: the algorithm reaches beyond the efficient core of the audience to spend the larger budget",
"Connects the budget-doubling mechanics to the rising-CPM signal or identifies it as the saturation diagnosis in disguise"
]
},
{
"id": 8,
"prompt": "Performance review in three weeks and I want to make the case for Senior Paid Media Manager. My draft pitch: I've been at Bramblewick (outdoor gear retailer) four and a half years, ran over 240 campaigns this year, never missed a launch, covered two colleagues' accounts when they left in March without any extra pay, and everyone hired the same year as me has already been promoted. Can you tighten this up into something persuasive?",
"expected_output": "A rebuilt promotion case anchored on an owned number and mapped to the senior screen, converting the absorbed accounts into an owned-scope argument, with a graded market anchor and a dated change-course benchmark instead of the tenure pitch.",
"files": [],
"expectations": [
"Rejects the tenure, campaign-count, reliability, and peer-comparison framing rather than tightening it, because cases citing years served, hours, campaign counts, or effort fail",
"Rebuilds the case around an owned revenue, pipeline, contribution, or P&L number the user is accountable for",
"Converts the absorbed colleague accounts (scope creep) into a 'what I own now that I didn't' argument: what was taken on and what the owned number did",
"Cites the survey evidence that remit growth without matching pay is widespread (51.2%, Marketing Week 2026 survey), making it a lever rather than a grievance",
"Maps the evidence to what the senior rung is screened for: forecasts and budgets, multi-account portfolio economics, C-level narrative, QA of juniors",
"Names budget and forecast responsibility as the senior-track signal to demonstrate or explicitly request",
"Uses the one-pager structure: what I own now, what the owned number did, evidence against the next rung's screen, market anchor, the ask",
"The ask names title, band, and timing in one sentence",
"Includes a market anchor carrying currency, source, date, and grade, with the grade stated out loud",
"Uses only the candidate's own evidence, indexed if the absolute figure is internally confidential; no invented numbers",
"Warns that an undated or last-year anchor damages credibility while the surveyed market is declining",
"Sets a benchmark-to-change-course: budget or forecast responsibility not granted within an agreed window means the senior track is blocked there and the plan moves to changing seats",
"Advises citing the internal ladder or a real job description for the target title when mapping evidence to the next rung's screen"
]
},
{
"id": 9,
"prompt": "Five years at agencies, now a campaign manager on £41K running eight client accounts and honestly fried. Two options on the table: an in-house senior exec role at a homeware brand for £48K, or going freelance - a former client hinted they'd send me a project, and I keep seeing £300+ day rates for PPC contractors, which works out to £70K+ a year. Which way do I jump?",
"expected_output": "A seat decision applying the mid-level flip toward in-house, refusing the day-rate arithmetic as a reason to go freelance, probing whether the freelance promoting condition actually holds, and re-ranking on what the user is short of.",
"files": [],
"expectations": [
"States that the seat ranking flips at mid-level: from mid onward in-house ranks above freelance, which ranks above agency",
"States that staying agency past mid-level fights the compensation curve (agency pay lowest at junior and capped at mid)",
"Refuses to promote freelance on the day rate alone, because that number is exactly what the ranking already priced in and rejected",
"Corrects the £70K+ arithmetic: the £300 median PPC day rate (itjobswatch, 2025) is down roughly 15% year-on-year and freelance income is volatile",
"Names freelance's real effort: a standing sales job on top of the craft, plus income instability",
"Distinguishes the former client's hint from the condition that actually promotes freelance: a client who has said yes, or a network that reliably converts",
"Re-ranks on what the user is short of: short of money points to in-house early, short of autonomy is the case that takes freelance early despite the order",
"Uses the burnout evidence as survey-grade context for the agency exit (Marketing Week 2026 survey figures such as 65.3% overwhelmed or 55.1% emotionally exhausted)",
"Treats agency and freelance effort as a tie in size only, costing different things: hours and burnout risk versus income stability and a standing sales job",
"Quotes pay figures with source, date, and grade, and treats the £48K offer against dated ranges rather than points",
"States the ordering out loud as a default rather than a law, and names which of the user's circumstances moved which option",
"Asks clarifying questions or explicitly conditions the recommendation on the unanswered constraints (deadline, one-off versus compounding goal, effort ceiling and reversibility)"
]
},
{
"id": 10,
"prompt": "I'm the paid media manager (mid-level, in-house) at a B2B logistics SaaS spending $180K/month, reporting to a marketing director. Goal: head of growth somewhere within three years. I've shortlisted courses: an MMM course, an incrementality certification, a forecasting bootcamp and an advanced attribution course - about 240 hours of study. Build me the plan.",
"expected_output": "A roadmap that replaces the course list with negotiated-access asks carrying dates, sequences unit economics first, deliberately promotes incrementality for the lead goal, and anchors everything on the two observable threshold signals plus a dated change-course benchmark.",
"files": [],
"expectations": [
"Rejects the study-only plan: from cross-channel allocation downward the roadmap step is negotiating access (budget, a second channel, enough spend to hold a geo out), not studying",
"Writes the access steps as asks with dates, not learning tasks",
"Builds the roadmap by diffing the user's current evidence against what the rung above is screened for, not as a generic curriculum",
"Sequences unit economics first on efficiency, because a week of study changes how every answer is read at any employer",
"Names incrementality as the option the efficiency order starves (first on value and on effort) and promotes it deliberately because the user targets head of growth and sits on real spend",
"Names budget and forecast responsibility as the senior-track signal: if the employer is not handing it over, the user is not on the senior track there whatever the title says",
"Names at least one completed incrementality test (geo holdout or conversion lift) as the lead or head-of-growth signal, expected above roughly $10M in spend",
"Every recommended step names the observable signal that proves it happened",
"Sets a dated benchmark-to-change-course, such as no budget or forecast responsibility within the agreed window meaning the plan moves to changing seats",
"Distinguishes what remains genuinely study-able (e.g. unit economics) from what requires negotiated access (incrementality, forecasting, cross-channel allocation)",
"Ties the head-of-growth target to its screen: P&L ownership or accountability for the revenue number, cross-channel allocation, hiring, incrementality",
"States the ordering out loud and names which of the user's answers (three-year horizon, compounding goal, existing spend) moved which item"
]
},
{
"id": 11,
"prompt": "I've been an SEO executive at a travel company for two years and want to move into paid media. Constraints: I can only spend evenings on this, I can't take a pay cut, and I won't quit my current job until something is certain. One friend says freelancing is the fastest way in, another says do the Level 3 apprenticeship. What's my best route?",
"expected_output": "An entry plan that promotes the lateral move to first for the adjacent-seat user, deletes freelance-first and the apprenticeship by name under the stated constraints, and pairs the move with evenings-scale self-funded spend, with orderings and moved options stated out loud.",
"files": [],
"expectations": [
"Promotes the lateral move to first choice because the user already sits in an adjacent seat (SEO)",
"Gives the lateral move's rationale: it keeps the salary, the employer, and the reversibility that every other entry path spends",
"Deletes freelance-first from the menu by name given the evenings-only reversible constraint, because its effort is a standing sales job that runs for years before proof exists",
"Deletes or removes the apprenticeship by name given the no-pay-cut constraint: low pay, 8-15 months, locked to one employer, and UK marketing apprenticeships in decline",
"States deleted options as off the menu rather than ranked last, so they do not return as scope later",
"Recommends self-funded own spend alongside the lateral move as evenings-compatible proof that doubles as the portfolio",
"States that no credible-minimum budget exists for self-funded spend: the full loop matters, not the size",
"Warns of the self-funded path's risks: unmentored bad habits, and no title conferred on its own",
"States the entry ordering out loud as a default, naming which constraint moved or deleted which option",
"Expresses effort in hours, years committed, and reversibility, never in money",
"Explains why the agency junior default is displaced here, referencing its pay level (UK postings clustering £24,000-£30,000) against the no-pay-cut constraint or the adjacent-seat promotion rule",
"Notes the lateral move still requires proving platform execution despite the adjacent credibility"
]
},
{
"id": 12,
"prompt": "Seven years running LinkedIn and Google search campaigns for B2B software companies. Two worries: a recruiter approached me about a head-of-performance role at a DTC cookware brand and I'm scared I'd be found out, and honestly with all the automated campaign types doing everything themselves I wonder if this career has three years left. Should I double down on hand-optimisation skills to stay valuable, and can I even make the B2B-to-DTC jump?",
"expected_output": "An answer that treats B2B and DTC as markets moved between rather than ranked options, names the transferable set and the narrow re-learning, rejects hand-optimisation as commoditised, and positions the user on the appreciating skills with the efficiency-durability contradiction made explicit.",
"files": [],
"expectations": [
"Refuses to rank B2B against B2C, framing them as two markets a buyer moves between rather than options competing for the same hours",
"Names several of the skills that transfer unchanged: platform mechanics, bidding logic, tracking and attribution setup, budget pacing, audience construction, testing discipline, unit-economics literacy, diagnostic mindset",
"Names the narrow re-learning for the DTC move: creative cadence and high-volume creative testing",
"Names what changes on the DTC side: channel families (paid social, short-form video, shopping/feed formats) and core metrics (ROAS, MER, contribution margin, CPA rather than pipeline and MQL/SQL)",
"Rejects doubling down on hand-optimisation: keyword-by-keyword bidding, manual audience building, manual bid adjustments, and routine reporting are commoditised and no career case should be built on them",
"Frames the role shift as moving from keyword manager to system optimizer: upstream in the inputs, mid-campaign in the signals, downstream in the measurement",
"Names the appreciating areas: creative strategy and testing, measurement and incrementality, first-party data and feed/catalogue work, AI-creative workflows, and system judgment",
"Notes the practitioner consensus that creative roadmapping has no AI replacement yet",
"States the deliberate contradiction: the cheapest area to pick up (AI-creative workflows, an evening) is also the one commoditising fastest, so it wins an interview this quarter and expires by the next rung",
"Conditions the lead choice: AI-creative workflows on a hard deadline, creative strategy and testing on a compounding mandate, and never the cheap end as a career case on its own",
"Positions the user's seven years of watching automated systems as progress toward system judgment, the appreciating skill built from exactly that experience",
"Expresses the effort axis in time units (an evening, a week, a quarter, years), never in money"
]
}
],
"trigger_queries": [
{ "query": "How do I get into paid media with no experience?", "should_trigger": true },
{ "query": "I have a media buyer interview at a DTC brand on Thursday, help me prep", "should_trigger": true },
{ "query": "Three years as a PPC exec, how do I get to senior?", "should_trigger": true },
{ "query": "Should I go agency or in-house for my next paid media role?", "should_trigger": true },
{ "query": "How do I build an ad portfolio when all my client work is under NDA?", "should_trigger": true },
{ "query": "Is the Google Ads certification actually worth it?", "should_trigger": true },
{ "query": "What should a growth marketing manager earn in the US right now?", "should_trigger": true },
{ "query": "I'm an SEO specialist and want to switch to PPC, where do I start?", "should_trigger": true },
{ "query": "Help me prepare for a live account audit interview", "should_trigger": true },
{ "query": "They gave me a take-home case study for a paid social role, how long should I spend on it?", "should_trigger": true },
{ "query": "How do I ask for a promotion as a paid media manager?", "should_trigger": true },
{ "query": "What does the career path from media buyer to head of growth look like?", "should_trigger": true },
{ "query": "Should I quit my agency job and freelance as a media buyer?", "should_trigger": true },
{ "query": "What certifications do I need to become a PPC specialist?", "should_trigger": true },
{ "query": "Is a career in paid advertising still viable with all the automation?", "should_trigger": true },
{ "query": "Prep me for a performance marketing interview at a fintech", "should_trigger": true },
{ "query": "What skills do I need to move from campaign manager to paid media director?", "should_trigger": true },
{ "query": "How much should I ask for as a senior paid social manager in London?", "should_trigger": true },
{ "query": "I keep getting rejected for media buyer roles, what am I doing wrong?", "should_trigger": true },
{ "query": "Moving from B2B ads to ecommerce, what do I need to relearn?", "should_trigger": true },
{ "query": "The interviewer will screen-share their ad account and watch me audit it live, how do I not blow it?", "should_trigger": true },
{ "query": "Got asked 'high CTR but low conversion rate, why?' in an interview and froze, help me practice these", "should_trigger": true },
{ "query": "How do I show campaign results in my portfolio without breaching client confidentiality?", "should_trigger": true },
{ "query": "Meta Blueprint or Google Skillshop first?", "should_trigger": true },
{ "query": "My remit doubled this year with no raise, how do I make the promotion case?", "should_trigger": true },
{ "query": "What does a junior PPC executive earn in the UK these days?", "should_trigger": true },
{ "query": "I want to become a media buyer, where do I even start?", "should_trigger": true },
{ "query": "Is it too late to get into performance marketing at 35?", "should_trigger": true },
{ "query": "Interview tomorrow for a paid search role and they mentioned a mental maths round", "should_trigger": true },
{ "query": "How do I go freelance as a PPC consultant without starving?", "should_trigger": true },
{ "query": "Help me write an anonymised case study from my Meta ads results for my portfolio", "should_trigger": true },
{ "query": "This take-home for a growth role is basically their whole marketing plan, is that normal?", "should_trigger": true },
{ "query": "What's the real difference between agency and in-house life as a media buyer?", "should_trigger": true },
{ "query": "Give me a roadmap to head of paid in three years", "should_trigger": true },
{ "query": "Do hiring managers actually care about platform certifications?", "should_trigger": true },
{ "query": "I run my partner's Shopify store ads on evenings, does that count as experience for a real job?", "should_trigger": true },
{ "query": "They offered me $95K for a growth marketing role, how do I negotiate?", "should_trigger": true },
{ "query": "Which interview questions should I expect for a paid social manager position?", "should_trigger": true },
{ "query": "I'm a teacher wanting to retrain into digital ads, what's a realistic path?", "should_trigger": true },
{ "query": "How do I stay valuable now that Google automates most of the bidding?", "should_trigger": true },
{ "query": "Behavioural interview prep for an ad agency final round", "should_trigger": true },
{ "query": "Should I take the marketing apprenticeship or the agency junior role?", "should_trigger": true },
{ "query": "What separates a mid-level buyer from a senior paid media strategist?", "should_trigger": true },
{ "query": "How do I answer 'what would you do with a $50K per month budget' in an interview?", "should_trigger": true },
{ "query": "My boss won't give me budget ownership, am I stuck at this level?", "should_trigger": true },
{ "query": "Can I put client ROAS numbers on my personal site?", "should_trigger": true },
{ "query": "Is retail media a good specialism to build my career on?", "should_trigger": true },
{ "query": "Compare PPC contractor day rates to a permanent salary for me", "should_trigger": true },
{ "query": "I need a portfolio for growth marketing roles but I've only ever done unpaid side projects", "should_trigger": true },
{ "query": "Tips for the case study round at a performance agency?", "should_trigger": true },
{ "query": "How do I move from account management into media buying at my own agency?", "should_trigger": true },
{ "query": "What should I learn first, incrementality or creative strategy?", "should_trigger": true },
{ "query": "am i underpaid as a ppc manager on 42k", "should_trigger": true },
{ "query": "Recruiters keep passing on me for paid media roles, is my positioning the problem?", "should_trigger": true },
{ "query": "What does a creative strategist career path look like?", "should_trigger": true },
{ "query": "Preparing to ask for a raise after running our ads for two years", "should_trigger": true },
{ "query": "In-house offered me 55k, the agency counter is 48k, which is better for my career long term?", "should_trigger": true },
{ "query": "How do I prove I can run paid ads without a marketing degree?", "should_trigger": true },
{ "query": "Verbal offer as growth lead at a seed startup, what equity is normal?", "should_trigger": true },
{ "query": "Should I bother with the TikTok Academy certification?", "should_trigger": true },
{ "query": "How do I answer 'tell me about your biggest campaign failure'?", "should_trigger": true },
{ "query": "I want to move up but my agency keeps me on pure execution work", "should_trigger": true },
{ "query": "Help me plan the next two years of my paid media career", "should_trigger": true },
{ "query": "My interview has an account teardown stage, what even is that?", "should_trigger": true },
{ "query": "Going out on my own as a media buyer, how much can I actually charge?", "should_trigger": true },
{ "query": "Is it smart to spend my own money running ads just to learn the job?", "should_trigger": true },
{ "query": "How do I switch from email marketing into performance marketing?", "should_trigger": true },
{ "query": "Write a job description for a senior media buyer", "should_trigger": false },
{ "query": "Design an interview scorecard for hiring a PPC specialist", "should_trigger": false },
{ "query": "What should we pay the growth marketer we're hiring in Austin?", "should_trigger": false },
{ "query": "Build a 30-60-90 ramp plan for our new paid media hire", "should_trigger": false },
{ "query": "What work-sample test should we give media buyer candidates?", "should_trigger": false },
{ "query": "How many interview rounds should our performance marketing hiring loop have?", "should_trigger": false },
{ "query": "Screen these five media buyer CVs for our open role", "should_trigger": false },
{ "query": "Should our startup hire a media buyer or keep the agency?", "should_trigger": false },
{ "query": "What are red flags when interviewing paid media candidates for my team?", "should_trigger": false },
{ "query": "Our new PPC hire is struggling, how do I ramp them faster?", "should_trigger": false },
{ "query": "My CPA doubled overnight, help me figure out why", "should_trigger": false },
{ "query": "Audit my Google Ads account, performance fell off a cliff", "should_trigger": false },
{ "query": "Why did my ROAS drop after I increased the budget yesterday?", "should_trigger": false },
{ "query": "Our ads stopped converting last week, diagnose the account", "should_trigger": false },
{ "query": "Meta says 40 conversions but Shopify shows 12, explain the gap", "should_trigger": false },
{ "query": "Platform ROAS says 4.2 but blended MER is 1.8, reconcile these numbers", "should_trigger": false },
{ "query": "Is a CAC of $230 healthy for our B2B SaaS?", "should_trigger": false },
{ "query": "Benchmark our ROAS against industry averages", "should_trigger": false },
{ "query": "How should I split $60K a month between Google and Meta?", "should_trigger": false },
{ "query": "What PPC newsletters and podcasts should I be following?", "should_trigger": false },
{ "query": "Where do I start with paid advertising for my startup?", "should_trigger": false },
{ "query": "Is my ad creative fatigued or is this seasonality?", "should_trigger": false },
{ "query": "Verify my conversion pixel fires only once before we launch", "should_trigger": false },
{ "query": "When can I scale my winning campaign from $2K to $5K a day?", "should_trigger": false },
{ "query": "Build a negative keyword list from this search terms report", "should_trigger": false },
{ "query": "Audit the landing page my paid traffic goes to", "should_trigger": false },
{ "query": "Which ad platform fits a B2B SaaS with a $10K monthly budget?", "should_trigger": false },
{ "query": "Should I switch this campaign from manual CPC to target ROAS?", "should_trigger": false },
{ "query": "How many customers do I need for a lookalike seed list?", "should_trigger": false },
{ "query": "Write me five ad copy variants for this value proposition", "should_trigger": false },
{ "query": "Script a UGC ad for my skincare brand", "should_trigger": false },
{ "query": "Draft a creative brief for our Q4 campaign", "should_trigger": false },
{ "query": "Set up a competitor ad swipe file for my niche", "should_trigger": false },
{ "query": "Plan a campaign promoting our founder's LinkedIn posts as ads", "should_trigger": false },
{ "query": "We're 60% through the month but 80% through budget, fix the pacing", "should_trigger": false },
{ "query": "Map the buying committee for our ERP product so we can target it", "should_trigger": false },
{ "query": "Build an audience targeting plan for my DTC coffee brand", "should_trigger": false },
{ "query": "Should I merge my 40 ad campaigns into 5?", "should_trigger": false },
{ "query": "Is carousel the wrong format for an awareness objective?", "should_trigger": false },
{ "query": "Rank these five video ad hooks for test budget", "should_trigger": false },
{ "query": "Design a creative test plan with kill and scale rules", "should_trigger": false },
{ "query": "Set a maximum CAC guardrail for our paid spend", "should_trigger": false },
{ "query": "Adapt my ad copy for AI assistant answer placements", "should_trigger": false },
{ "query": "Design a retargeting sequence with recency windows for my store", "should_trigger": false },
{ "query": "Help me prep for a software engineering interview", "should_trigger": false },
{ "query": "How do I become an SEO specialist?", "should_trigger": false },
{ "query": "Career advice for a data analyst who wants to become a data scientist", "should_trigger": false },
{ "query": "Prep me for a product marketing manager interview", "should_trigger": false },
{ "query": "How much do social media managers earn?", "should_trigger": false },
{ "query": "Help me build my UX design portfolio", "should_trigger": false },
{ "query": "Negotiate my software developer salary offer", "should_trigger": false },
{ "query": "How do I become a marketing analytics manager?", "should_trigger": false },
{ "query": "What's a good career path in content marketing?", "should_trigger": false },
{ "query": "Help me get promoted as an account executive in sales", "should_trigger": false },
{ "query": "Review my CV for a graphic design job", "should_trigger": false },
{ "query": "Interview prep for a growth engineer role", "should_trigger": false },
{ "query": "How do I become a brand manager at a CPG company?", "should_trigger": false },
{ "query": "Is the HubSpot inbound certification worth it for a content marketer?", "should_trigger": false },
{ "query": "Write a cold email to a recruiter about media buying jobs", "should_trigger": false },
{ "query": "Find me open media buyer roles in Manchester", "should_trigger": false },
{ "query": "Fill in this application form for a PPC role for me", "should_trigger": false },
{ "query": "How should our agency price PPC management retainers?", "should_trigger": false },
{ "query": "Build a training curriculum for my junior media buyers", "should_trigger": false },
{ "query": "What questions should I ask candidates in a media buyer interview?", "should_trigger": false },
{ "query": "How do I get my first freelance web design client?", "should_trigger": false },
{ "query": "Freelancing versus full-time employment for a copywriter, which is better?", "should_trigger": false },
{ "query": "What salary range should we budget for the paid media manager role we're opening?", "should_trigger": false }
]
}
references/anonymised-case-study-example.md›
# NDA-safe case study: template, example, negative example
Every number in a case study must be the candidate's own evidence, converted to NDA-safe form (percentage change, index with baseline = 100, or an aggregated range). Never invent figures to make a template look complete - an interviewer probing one fabricated number sinks the whole portfolio.
## Template
```
# [Descriptor client] - [one-line outcome, indexed or percentage]
**Quick facts:** role · engagement type (agency/in-house/freelance) · timeline ·
team size · channels · one-line impact (indexed).
## Problem
What was broken or missing, in business terms. 2-3 sentences.
## Process
What you investigated, what you found, what you decided to test and why.
This section carries the hiring signal - it shows judgment, not access to budget.
## Solution
What you actually changed: structure, targeting, creative approach, measurement.
Category concepts, not one vendor's menu paths.
## Results (indexed, baseline = 100)
| Metric | Before | After | Change |
|---|---|---|---|
| [metric] | 100 | [index] | [+/-X%] |
## Learnings
One thing that worked, one that did not, one thing you would do differently.
The failure line is deliberate - it shows self-awareness, not weakness.
*All client details anonymised; figures indexed to protect confidential data.*
```
## Worked example (illustrative - every value below is a placeholder shape, not a real result; replace with your own evidence)
# DTC apparel brand - cut acquisition cost by [X]% while holding volume
**Quick facts:** Paid social buyer · agency engagement · [N] months · team of 2 · paid social + retargeting · CPA index 100 → [index], conversion volume held within [±Y]%.
**Problem.** Acquisition cost had risen for [N] consecutive months while spend held flat; the account was concentrated in a handful of aging creatives and one broad audience.
**Process.** Audited spend concentration and creative age; found [share]% of spend in creatives older than [N] weeks with frequency climbing. Ranked hypotheses: creative fatigue first, audience saturation second, landing page third - tested in that order.
**Solution.** Built a weekly creative-testing cadence ([N] net-new concepts/week), restructured toward consolidated campaigns to pool learning, moved reporting from platform dashboard to blended cost per new customer.
**Results (indexed, baseline = 100).**
| Metric | Before | After | Change |
| --------------------- | ------ | ------- | ------ |
| CPA (blended) | 100 | [index] | [-X%] |
| New-customer volume | 100 | [index] | [±Y%] |
| Creative refresh rate | 100 | [index] | [+Z%] |
**Learnings.** The testing cadence mattered more than any single winning ad; the first restructure moved too fast and briefly destabilised delivery - the second was staged over two weeks.
_All client details anonymised; figures indexed to protect confidential data._
## Negative example - what disqualifies a case study
The same story, told the way that loses interviews and can breach an NDA:
> "At **[real client name]** I managed their Meta account spending **$85K/month**. Here's a screenshot of the ads manager showing the account name and revenue. ROAS went from 1.4 to 3.8 - screenshot attached. I can share the full account audit deck I made for them."
What is wrong, line by line:
- The real client name: NDA breach.
- The absolute spend and revenue figures: confidential data.
- The raw ads-manager screenshot with account name visible: legal risk, and interviewers read it as a preview of how their data would be treated.
- Offering a client's internal deck: breach on its face.
The anonymised version above carries the same credibility - "DTC apparel brand", indexed results, described process - with none of the exposure.
references/interview-question-bank.md›
# Interview question bank
Verified questions from candidate reports (Glassdoor), hiring-manager-sourced lists (Growth.Talent, compiled from 20+ growth leaders, 2026; Senator We Run Ads, 2024), and named practitioners. Grouped by format, with sample answers where the source material supports one.
## Live audit / account teardown
The interviewer screen-shares an account (real or example) and asks the candidate to audit it and recommend. Agency owner Jason Parks (The Media Captain): "ask the person you're interviewing to do a screen share with you where they audit your Google Ads account... It's a vote of confidence to see someone confidently dissecting your Google Ads account." The failure he screens for: "It's easy to send off a generic audit generated by third-party software" - a tool audit read aloud fails the round.
A defensible spoken structure, ~10 minutes:
1. Orient (1 min): business model, conversion action, what "good" means for this account before touching any setting.
2. Money first (3 min): where spend concentrates, whether the top spenders hit target CPA/ROAS, the single biggest waste line.
3. Structure and signal (3 min): campaign/ad-group logic, conversion tracking sanity, bidding fit to the goal.
4. Prioritised recommendation (2 min): the top three changes, ordered by expected impact, each with the metric that would prove it worked.
5. What you'd need next (1 min): the data you could not see and would request - this reads as rigor, not weakness.
## Mental-maths / unit-economics case
Glassdoor Paid Media Analyst report: "The case interviews ask you to use mental math to calculate metrics down the purchasing funnel such as CTR, CPC and then to compare 2 companies' revenues and gross profits... Group interview is an excel case with similar questions."
Drill the chain both directions until fluent: impressions → CTR → clicks → CPC → spend → conversion rate → conversions → CPA → AOV/ACV → revenue → gross margin → profit. Practise saying the intermediate numbers out loud; the round tests composure as much as arithmetic.
## Metric-diagnosis scenarios
Verified questions:
- "Your campaign has a high CTR but a low conversion rate. What could be the possible reasons and how would you address them?"
- "A client's campaign is not delivering impressions. What could be the reasons and how would you fix it?"
- "You notice a decline in the Quality Score of a top-performing keyword. What steps would you take?"
- "Here's our funnel data. Where's the biggest opportunity?" - the tell: look for the biggest **absolute** drop-off, not the biggest percentage drop.
- "This A/B test shows a 5% lift with p=0.08. What do you do?"
- "What would you do with a $50K/month paid media budget?"
- Scaling diagnosis: "I duplicated the campaign and set the new budget to $5,000 per day [from $2,000]. The higher budget yielded a much lower ROAS and almost doubled our CPM... why?" - expected answer: audience saturation / diminishing marginal returns; the algorithm reaches beyond the efficient core of the audience to spend the larger budget.
- B2B / professional-network variants: "The client wants to target decision-makers in a specific industry. How would you set this up?"; "You are tasked with creating a retargeting campaign. What audience segments would you target?"
- DTC / paid-social variant: "You notice that ad fatigue is affecting performance. How would you address this?"
"CPA doubled overnight" and "ROAS up, revenue flat" framings appear in agency thought-leadership rather than confirmed interview transcripts - still worth preparing with the same structure.
### Sample answer: paid-social ROAS drop
The practitioner-sourced diagnostic order, worth stating as an order rather than a list - and worth saying out loud in the round that it is ordered by hit rate per minute of checking, not by likelihood alone:
- efficiency: creative health > audience saturation > site-side conversion > attribution accuracy
- effort: site-side conversion (a cross-team ask, someone else's release log) > attribution accuracy (a tracking audit) > audience saturation > creative health (both visible in the account in minutes)
1. **Creative health** - fatigue is the cause roughly 60-70% of the time; check frequency and the age of the top spenders first.
2. **Audience saturation** - rising CPM against flat reach; the scaling-diagnosis question above is this in disguise.
3. **Site-side conversion** - did the landing page, offer, price, or checkout change? The ad account cannot fix a site problem.
4. **Attribution accuracy** - did tracking, consent, or the attribution window change? A reporting artefact is not a performance drop.
Open by scoping the window: a drop measured over a week or less is often noise - establish the comparison period before diagnosing. Close by naming the business metric that decides it (blended MER or contribution, not the platform dashboard alone).
### Sample answer: 5% lift at p=0.08
Do not answer "significant" or "not significant" - answer with a decision frame: what is the cost of being wrong in each direction, and can the test run longer to reach the pre-agreed threshold?
- If the variant is cheap to keep and the downside is small, ship it and keep measuring.
- If the decision is expensive or irreversible, extend the test.
Naming that the threshold should have been set before the test reads as senior.
## Take-home case study
The verified controversy: companies brief "a few hours" or "max 4 hours"; candidates report 12 to 30+ hours, and one priced a task at $2,800 of billable time. Idea theft is documented - candidates report seeing unpaid presentation ideas implemented after being ghosted. Ask A Manager's standard: "It's reasonable to ask candidates to spend an hour or two... it's not reasonable to ask them to complete complex projects, at least not without pay."
Rules:
- Time-box to 2-4 hours, and say so in the deliverable.
- State every assumption explicitly.
- Lead with the reasoning, not production polish.
- Push back on any brief that amounts to a full go-to-market plan for an unlaunched product - offer a scoped alternative ("I've focused the 3 hours on the channel and measurement plan; a full plan would need X, Y, Z from your data").
## Behavioural / client round
A reported three-stage shape: HR screen, then a campaign-manager technical round ("What KPIs would you look for when assessing the results and how would you then report it?"), then an executive character round.
Prepare each story in three lengths - full (2 minutes), short (60 seconds), one line (15 seconds) - so the same evidence survives a follow-up or a rapid-fire round. Bank stories across ownership, conflict/client management, failure and what changed, and a result tied to a business number.
Structure each as situation (1-2 sentences), task (1), action (2-3), result with metric (1-2). Use the candidate's own evidence only - never invent figures to fill a story.
## PPC / SEM interview questions (grouped by category)
### Platform and technical knowledge
- What is the difference between PPC and SEO?
- Explain Google Ads' auction mechanism and how Quality Score affects ad rank and cost.
- What bidding strategies have you used, and when do you switch from manual to automated?
- How do Enhanced CPC, Target CPA, Target ROAS, and Maximize Conversions differ, and what are the trade-offs?
- Describe your experience with Google Ads scripts and automation rules.
- What's your understanding of Google's Smart Bidding and when does it fail?
- How does the Google Ads interface differ from Microsoft Advertising / Bing Ads?
- What ad extensions have you used and which ones drive the most value?
- Explain the difference between Search, Display, Shopping, and Video campaigns.
- How do you set up conversion tracking: website events, app events, phone calls, imports?
### Campaign structure and strategy
- Walk me through how you structure a new Google Ads account from scratch.
- How do you decide campaign vs. ad-group granularity for a new client?
- What's your approach to keyword match types and when do you use broad vs. exact?
- How do you build a negative keyword list and at what frequency do you review it?
- Describe your process for keyword research and selection.
- How do you determine an appropriate budget for a PPC campaign?
- What's your approach to managing multiple PPC accounts simultaneously?
- How do you optimize ad placements on the Search and Display networks?
- What's your strategy for international PPC campaigns?
- How do you handle seasonal budget pacing?
### Optimization and analysis
- What are the key metrics you track in a PPC campaign?
- How do you ensure that your PPC ads comply with platform policies and guidelines?
- How do you handle click fraud or invalid clicks?
- How would you troubleshoot a sudden drop in campaign performance?
- What's your approach to A/B testing ad copy and landing pages?
- How do you analyze and report on campaign performance?
- How do you collaborate with other digital marketing teams?
- How do you optimize ad creative for different devices and audiences?
- What's your approach to budget pacing and mid-month adjustments?
- How do you measure incrementality of paid search vs. organic?
### Client and account management
- How do you communicate underperformance to a client?
- Walk me through a client escalation you handled.
- How do you set expectations around PPC results and timeline?
- Describe your experience with agency vs. in-house client dynamics.
- How do you balance client requests with data-driven recommendations?
- What's your experience presenting campaign results to stakeholders?
## Paid social interview questions (Meta, TikTok, LinkedIn, X)
### Platform and technical knowledge
- Imagine you're explaining Meta Ads to your grandma. How would you describe it?
- What are the main parts of a Meta Ad, and why is each part important?
- What's a pixel, and how does it help with your ads?
- Explain the difference between awareness, consideration, and conversion campaigns in Meta Ads.
- How have you adapted your Meta Ads strategies to address changes in iOS privacy policies?
- How does TikTok's auction differ from Meta's?
- What are LinkedIn Campaign Manager's bidding options and when do you use each?
- How do you set up conversion tracking on each platform?
- What's your experience with TikTok Spark Ads vs. in-feed ads?
- How does X (Twitter) Ads targeting differ from Meta?
### Campaign strategy
- If a client has a tiny budget, how would you suggest they spend it wisely on Meta Ads?
- How would you figure out who a client's perfect customer is on Meta, even if they don't know themselves?
- How do you ensure ads are showing to the right people on Meta?
- What's your approach to targeting for a client selling handmade jewelry on Meta?
- How would you set up a campaign to promote a mobile app on Meta?
- How do you structure campaigns for a DTC vs. B2B brand?
- What's your approach to testing new creative at scale?
- How do you decide which platform gets the budget?
- How do you approach prospecting vs. retargeting budget split?
- What's your experience with lead gen campaigns vs. traffic campaigns?
### Optimization and analysis
- How do you measure if a Meta Ad is successful or not? What numbers do you look at?
- How do you handle negative feedback or comments on your Meta Ads?
- How would you use retargeting in Meta Ads to reach people who have previously interacted with a client's business?
- How do you monitor and report on the performance of a Meta Ads campaign to a client?
- How do you determine the appropriate bidding strategy for a campaign based on its objectives?
- What's your approach to creative fatigue and when do you refresh?
- How do you analyze ROAS by audience segment?
- How do you handle a budget increase that tanks performance?
- What metrics matter most for paid social vs. paid search?
- How do you calculate CAC from paid social data?
### Behavioural and case
- Tell me about a time you had to solve a problem, even a small one. How did you figure it out, and what did you learn?
- How would you explain the Facebook Ads auction in simple terms?
- How would you target ads for a client selling handmade jewelry on Meta?
- How would you handle a situation where a client's ad budget suddenly increases?
- How would you use lookalike audiences in Meta Ads?
- Describe a paid social campaign you ran that underperformed and what you changed.
## Programmatic / display advertising questions
### Platform and technical knowledge
- What is programmatic advertising?
- How does real-time bidding work?
- What are the different types of programmatic buying? (Direct, Open Market, Private Marketplace, Preferred Deals)
- Can you explain the difference between first-price and second-price auctions?
- What's your understanding of viewability in programmatic advertising?
- What are SSPs, DSPs, and ad exchanges, and how do they relate?
- What is a DMP and how does it connect to programmatic?
- How does header bidding work and why does it matter?
- What are the IAB content taxonomies and how do you use them?
- What's your experience with Google Ad Manager / DV360 / The Trade Desk?
### Campaign strategy
- How do you optimize programmatic campaigns for better performance?
- What data points are essential for programmatic targeting?
- How would you explain programmatic advertising to a beginner?
- How do you measure the success of a programmatic campaign?
- How do you approach audience segmentation in programmatic advertising?
- How do you decide between programmatic direct and open auction?
- What's your approach to creative optimization in programmatic?
- How do you handle frequency capping in programmatic campaigns?
### Brand safety and quality
- How do you ensure brand safety in programmatic campaigns?
- What's your experience with ad verification tools in programmatic advertising?
- How do you measure viewability and ensure ads are actually seen?
- How do you handle fraud detection in programmatic?
- What's your approach to data privacy and GDPR in programmatic targeting?
### Behavioural and case
- Describe a successful programmatic campaign you've managed.
- How have you adapted your programmatic strategies to new technologies like AI and machine learning?
- Walk me through how you'd set up a brand awareness campaign programmatically.
## Ad operations / trafficking questions
### Platform and technical knowledge
- What is your experience with ad servers like Google Ad Manager, Freestar, etc.?
- What's your understanding of the trafficking process in advertising?
- How do you ensure compliance with advertising regulations and policies?
- What tools do you use for ad operations management?
- How do you handle ad inventory management?
- What's your experience with S2S (server-to-server) tracking vs. pixel-based?
- How do you handle creative specs across different placements?
- What's your understanding of VAST/VPAID tags for video?
### Campaign strategy
- How do you ensure ads are placed in the right contexts?
- How do you handle ad scheduling and rotation?
- How do you ensure creative compliance in ad operations?
- How do you approach mobile ad operations and in-app advertising?
- What's your experience with programmatic direct sales?
### Optimization and analysis
- What metrics do you track to measure ad operations performance?
- How do you troubleshoot technical issues in ad operations?
- How do you ensure data accuracy in trafficking?
- How do you handle API integration and technical errors?
- How do you measure the effectiveness of your trafficking process?
### Behavioural and case
- Describe a complex trafficking project you've managed.
- How do you ensure your trafficking meets campaign objectives?
- Walk me through a time you caught a trafficking error before it went live.
## Campaign manager / strategist questions
### Platform and technical knowledge
- What's your experience with campaign management platforms like DoubleClick, Google Campaign Manager, etc.?
- How do you approach cross-channel attribution?
- How do you structure and organize large-scale campaigns?
- What's your understanding of marketing mix modeling?
- How do you ensure campaign ROI?
- What tools do you use for campaign reporting and dashboards?
- How do you approach multi-touch attribution vs. last-click?
- What's your experience with data studios / Looker / Tableau for reporting?
### Campaign strategy
- How would you design a cross-channel attribution model?
- What strategies do you use for scaling campaigns across markets?
- How do you balance short-term results with long-term brand building?
- How do you approach campaign budgeting and resource allocation?
- How do you ensure campaigns meet business objectives?
- How do you structure a full-funnel campaign?
- How do you coordinate launch campaigns across paid, owned, and earned?
### Optimization and analysis
- How do you measure and report campaign performance to stakeholders?
- How do you use data to optimize campaign performance?
- How do you approach testing and iteration in campaign strategy?
- How do you handle campaign pivot decisions?
- How do you ensure campaigns stay within budget while maximizing results?
- How do you build a campaign forecast?
### Behavioural and case
- Can you walk me through a complete campaign lifecycle you managed?
- Describe a campaign where you had to pivot strategy mid-flight.
- How do you balance multiple competing priorities in campaign management?
- Tell me about your biggest campaign failure and what you learned.
## Performance marketing questions
### Platform and technical knowledge
- What's your understanding of performance marketing KPIs?
- How do you approach attribution in performance marketing?
- What's your experience with performance marketing platforms and tools?
- How do you ensure campaign scalability?
- How do you approach testing and optimization in performance marketing?
- What's your understanding of incrementality testing?
- How do you track and report on multi-channel performance?
- What's your experience with affiliate marketing management?
### Campaign strategy
- How do you structure a performance marketing funnel?
- How do you approach audience targeting in performance marketing?
- How do you balance acquisition cost with lifetime value?
- How do you approach channel selection in performance marketing?
- How do you ensure campaigns are compliant with regulations?
- How do you structure budget allocation across channels?
- How do you approach market entry with performance marketing?
### Optimization and analysis
- How do you measure the true ROI of performance marketing campaigns?
- How do you handle attribution modeling challenges?
- How do you approach A/B testing in performance marketing?
- How do you ensure data quality in performance marketing?
- How do you handle budget optimization in performance marketing?
- How do you identify and scale winning creatives?
### Behavioural and case
- Can you share an example of a high-performing performance marketing campaign you've managed?
- How have you balanced short-term performance with long-term growth in performance marketing?
- Describe a time you had to shut down a channel that wasn't performing.
## Agency-side vs. in-house interview questions
### Agency-side
- How does client management differ between agency and in-house roles?
- What are the main challenges of working on multiple client accounts?
- How do you handle conflicting priorities from multiple clients?
- What's your experience with presenting to and managing client stakeholders?
- How do you stay current with industry trends while managing multiple clients?
- How do you scope and pitch new business?
- How do you handle scope creep from clients?
- What's your experience with retainer vs. project-based billing?
- How do you manage a team of planners/buyers/optimizers?
- How do you onboard a new client account?
### In-house
- How does your day-to-day work differ when working in-house vs. agency?
- What are the main challenges of working only on one brand's campaigns?
- How do you approach campaign strategy without the diverse client experience?
- What's your experience with cross-functional collaboration in an in-house role?
- How do you measure success differently in agency vs. in-house roles?
- How do you influence other departments without authority?
- What's your experience working with an agency as the client?
- How do you build internal political capital for your campaigns?
## Common behavioral / situational questions (all advertising roles)
- Tell me about a time you had to meet a tight deadline for a campaign launch.
- Describe a situation where a client pushed back on your recommendation. How did you handle it?
- Walk me through a time you identified a significant waste in a budget and what you did.
- Tell me about a campaign that failed. What went wrong and what would you do differently?
- Describe a time you had to learn a new platform or tool quickly under pressure.
- How do you stay current with platform updates and industry changes?
- Tell me about a time you had to explain a complex metric to a non-technical stakeholder.
- Describe a time you had to work with incomplete data to make a recommendation.
- Tell me about a time you influenced a decision without formal authority.
- How do you handle the pressure of monthly/quarterly reporting deadlines?
references/promotion-case-one-pager.md›
# Promotion-case one-pager
The ladder pays for accountability, not time served: cases citing owned revenue or P&L impact succeed; cases citing tenure, hours, campaign counts, or effort fail. Scope creep is the modal experience - Marketing Week's 2026 Career & Salary Survey (2,350 respondents) found 51.2% of marketers saw their remit grow without matching pay - which makes it a legitimate lever, but only once converted into an owned-number argument.
Every figure in the one-pager is the candidate's own evidence; market anchors carry source, date, and grade (survey vs aggregate). An undated or last-year anchor in a declining market damages credibility.
## Template
```
# Case for [target title], [name], [date]
## What I own now that I didn't at [start of period]
- [Scope item 1: e.g. budget/forecast responsibility for X accounts / £Y band]
- [Scope item 2: e.g. measurement ownership - tracking, attribution, first
incrementality test]
- [Scope item 3: e.g. QA or mentoring of N juniors]
## What the owned number did
[1-3 lines: the revenue, pipeline, contribution, or efficiency number you are
accountable for, and its movement over the period - your own evidence,
indexed if the absolute figure is confidential internally.]
## Evidence against the next rung's screen
[Map each item above to what the next level is screened for: forecasts and
budgets, C-level narrative, cross-channel allocation, incrementality - cite
the internal ladder or a real JD for the target title.]
## Market anchor
[Target-title range, with currency, source, date, and grade. State the grade
out loud: "survey-grade" or "aggregate estimate, treat as a range".]
## The ask
[Title, band, and timing. One sentence.]
```
## Positive example (shape)
"Since January I have owned the paid-media budget and quarterly forecast for our three largest accounts - responsibility that sat with my director a year ago - and ran our first geo-holdout test, which changed how we allocate between search and social. The number I'm accountable for is [owned metric and its movement]. That maps to the senior screen - forecast ownership, measurement depth, portfolio economics. The 3Search 2025 survey (1,300+ responses) puts UK PPC Managers at £50,000-£70,000; I'm asking for [band] effective [date]."
Why it works:
- Every claim is an owned scope item or an owned number.
- The evidence is mapped to the next rung's documented screen.
- The anchor is dated and graded.
## Negative example
"I've been here four years, I've never missed a deadline, I ran over 200 campaigns this year and regularly stay late during launches. Everyone else at my level has been promoted. I'd like the senior title."
Why it fails:
- Four years is tenure, not evidence.
- 200 campaigns and late nights are activity, not evidence.
- Peer comparison is not evidence.
- Nothing maps to what senior is screened for: no owned budget, no forecast, no number.
This case reads as a request for recognition, not a demonstration of accountability, and it fails even when the underlying work was genuinely strong.
SKILL.md›
---
name: advertising-career
description: "Plan and advance a paid-advertising career from the candidate side - media buyer, performance marketer, PPC specialist, creative strategist, growth marketer - covering the junior to lead ladder, the five real interview formats (live account audit, mental-maths case, metric diagnosis, take-home, behavioural), an NDA-safe results portfolio, platform certifications, in-house vs agency vs freelance, and promotion and pay conversations. Use whenever the user asks how to get into paid media or PPC, preps a media-buyer or growth interview, wants an ad portfolio, or weighs an agency vs in-house move - even if they never say 'career'. Do NOT use for the employer side of hiring - use mbfinotti/advertising-skills@advertising-hiring instead."
license: MIT
metadata:
author: Maya-Beth Finotti
version: "1.2.5"
---
# Advertising Career
You are a paid-media career coach. Help one person land and grow paid-advertising roles:
- Interview preparation
- A skill-gap roadmap keyed to the seniority ladder
- NDA-safe portfolio positioning
- Progression from junior buyer to lead/head of growth
- The in-house vs agency vs freelance decision
Advise in channel categories and craft skills, never one vendor's UI. Vendor names appear below only where a certification itself is the subject.
Out of scope: the hirer's side of this same market lives in the sibling skill mbfinotti/advertising-skills@advertising-hiring (role definition, interview scorecards, work-sample design, 30-60-90 ramp plans). Hand off, never cover it here.
This skill is also not a job-search or application tool. It does not:
- Find job listings
- Fill in application forms
- Write cold outreach to recruiters
- Submit anything on the user's behalf
Every pay figure in this skill carries its currency, source, date, and grade:
- Survey: a named survey with respondent count.
- Aggregate: a job-board or aggregator estimate - treat as a wide range, never a point figure.
The strongest 2026 survey showed UK marketing pay declining year-on-year, so quoting a 2025 figure can overstate the current market and cost the candidate credibility.
If you can browse the web, reverify any figure before it anchors a negotiation. If you cannot, date the figure out loud and state its grade.
## Interview
Ask one question at a time, multiple-choice where possible. Skip anything already answered. Do not build a roadmap or prep plan until questions 1-2 and the questions relevant to the stated goal are answered.
1. Where are you now? (a) no paid-media experience yet (b) adjacent role - SEO, email, content, analytics, account management, ecommerce ops (c) junior buyer/executive (d) mid-level manager (e) senior strategist/director (f) lead / head of growth (g) freelance/contract.
2. What is the goal? (a) land a first paid-media role (b) prepare for a specific interview process (c) get promoted or reach the next rung (d) decide in-house vs agency vs freelance (e) build or reposition a portfolio (f) plan the next 1-3 years.
3. Target role and seniority - what title, at what level, at what kind of company?
4. Channels of interest: (a) search (b) paid social (c) short-form video (d) retail media (e) B2B professional networks (f) broad / undecided.
5. B2B, B2C/DTC, or both?
6. Geography for pay benchmarks: (a) UK (b) US (c) EU/France (d) other.
7. By what date must the result have landed? Ask for a real date when one exists - an offer deadline, a contract or visa end, a review cycle: (a) interviewing right now (b) 1-3 months (c) 6-12 months (d) no deadline.
8. One-off win or compounding asset? (a) one-off - a named interview, offer, or review to win (b) compounding - capability and evidence that keeps paying across employers (c) both, in that order.
9. Effort ceiling: how many unpaid hours a week, over how many months, and how reversible must the move stay? (a) evenings only, must stay reversible (b) a day a week, willing to change employer (c) full commitment, willing to drop income for a period.
10. Seat preference: (a) in-house (b) agency (c) freelance (d) undecided - that decision is part of the job.
11. Is your past work under NDA or client confidentiality?
12. Spend under management to date: (a) none (b) own small spend (c) under $10K/month (d) $10-100K/month (e) over $100K/month.
13. Do you hold budget or forecast responsibility today? (This is the senior-track signal - see the ladder.)
14. Have you ever run an incrementality test - geo holdout or conversion lift?
15. Which platform certifications do you hold, and are they current?
Route by goal:
| Goal | Route to |
| ------------------------------------------ | ---------------------------------------- |
| (a) land a first paid-media role | Entry paths + Certifications + Portfolio |
| (b) prepare for a specific interview | Interview preparation + Portfolio |
| (c) get promoted / next rung | Skill-gap roadmap + Promotion case |
| (d) decide in-house vs agency vs freelance | In-house vs agency vs freelance |
| (e) build or reposition a portfolio | Portfolio |
| (f) plan the next 1-3 years | Skill-gap roadmap + Compensation |
Example invocations:
- "I have a media-buyer interview at a DTC brand on Thursday" → route (b)
- "three years as a PPC exec, how do I get to senior?" → route (c)
Every ordering in this skill is a default for someone with no unfair advantage, not a law - say the ordering out loud, then re-rank it:
- A hard date (7) promotes the fast-acting options - anonymising work that already exists, drilling the live audit, applying to agency seats - and demotes anything measured in quarters.
- A compounding answer (8) promotes the slow ones: self-funded spend, negotiated incrementality access, creative-testing reps.
- A low effort ceiling (9) deletes rather than demotes every option whose effort is a standing job: freelance client-getting, an apprenticeship, an unbounded take-home. Say which options that removes - "freelance-first is off your menu, not last on it" - because an option the user cannot act on, parked at the bottom of a list, comes back as scope a month later.
- Re-rank again on answers 1-6 and 10-15: an adjacent seat promotes the lateral move to first, an analytics background promotes measurement, an existing network or a built portfolio demotes certifications. A market with no reachable agency seats deletes agency-first outright - rebuild the menu without it rather than leaving it ranked last.
- Name which answer moved which option, and which options the answers deleted, so the user argues with the order instead of inheriting it.
If your harness has persistent memory, store the target role, the roadmap, and the benchmark-to-change-course dates so later sessions resume the same plan instead of re-interviewing.
## Entry paths
Six paths appear in real postings and practitioner accounts, ranked below by what each returns per unit of effort. Effort on this menu is years committed, reversibility, and hours nobody pays for - never money. The pay figures in the table are facts about each path, not its cost, and the ordering does not depend on their precision.
- efficiency: agency junior > self-funded own spend > lateral move > in-house coordinator > apprenticeship > freelance-first
- value (speed to a track record the next employer screens for): agency junior > in-house coordinator > lateral move > self-funded own spend > freelance-first > apprenticeship
- effort: freelance-first > apprenticeship > agency junior > in-house coordinator > self-funded own spend > lateral move
Default: chase an agency junior seat while running self-funded spend alongside it - the second is what makes the first's applications answerable, and it costs evenings rather than years. Promote the lateral move to first choice when the user already sits in an adjacent seat, because it keeps the salary, the employer and the reversibility that every other rung spends.
What the efficiency order starves: freelance-first, the path with the highest ceiling and the longest runway. Its rate ceiling is the top of the menu and its autonomy is uncapped, yet it loses every round because its effort is a standing sales job that runs for years before any proof exists - a ratio can only ever rank it last.
Promote it above the default, not merely off the bottom, when either condition holds:
- The user already holds the client side of the work (a client has said yes, or a network that reliably converts).
- Autonomy rather than reps is what they are short of, and answer 9 allows dropping income for a period.
Never promote it on the day rate alone: that number is exactly what the ratio already priced in and rejected.
| Path | Trade-off | Anchor (source, date) |
| ------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------- |
| Agency junior/executive | Fastest, broadest hands-on learning; lowest pay, highest volume pressure, but reversible at any point | UK postings cluster £24,000-£30,000 (itjobswatch-listed postings, 2025) |
| Self-taught own spend | Most credible no-experience proof, doubles as the portfolio; risk of unmentored bad habits, and confers no title on its own | No credible-minimum budget exists - the full loop matters, not the size |
| Lateral move | Adjacent credibility (SEO, email, analytics, account management) at near-zero income risk; must still prove platform execution | Analytics/SQL transfers best into measurement-heavy growth roles |
| In-house coordinator | Broader remit, closer to business metrics; fewer seats, less mentorship - often the only paid person, so capability accrues slower | Real postings span search, social, copy, testing in one role |
| UK Level 3 apprenticeship | Paid training, zero tuition; low pay, locked to one employer for the duration, and UK marketing apprenticeships are in decline (Marketing Week) | 8-15 months, funding band £11,000-£11,500 |
| Freelance/contract | No brand name needed, highest day rate; volatile income, and a standing sales job on top of the craft before any proof exists | UK median PPC day rate £300, 25th-75th £275-£325, down ~15% YoY (itjobswatch, 6 months to 17 July 2025) |
Agency-to-in-house is the dominant industry path: agencies teach breadth fast, in-house pays for depth. Warn anyone tempted by "20x ROAS" self-taught success stories that practitioners publicly call those claims not credible - a small honest project with documented process beats an inflated one.
## The ladder and the skill-gap roadmap
What hiring managers screen for at each rung, from real job descriptions:
| Level | Typical title | Screened for |
| ------ | ----------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------- |
| Junior | Paid Media / PPC Executive | Platform basics, keyword/audience research, ad copy, reporting cadence, coachability |
| Mid | Paid Media / Campaign Manager | End-to-end campaign ownership ("full lifecycle... strategy... execution, optimisation, and reporting"), client presentation, attribution and tracking |
| Senior | Senior Strategist / Director | Multi-account portfolio economics, "manage forecasts and budgets", C-level narrative, QA of juniors |
| Lead | Head of Paid / Head of Growth | P&L ownership ("accountable for the revenue number"), cross-channel allocation, hiring, incrementality |
Where each skill enters - build the roadmap by diffing the user's current evidence against the rung above them:
| Skill | Enters at |
| ---------------------------------------------------------------------------------------- | --------------------------------------------- |
| Platform execution, keyword/audience build, reporting | Junior - table stakes, increasingly automated |
| Creative strategy and creative testing | Mid - dominant on paid social |
| Unit economics: CAC, LTV, contribution margin, payback | Mid to senior |
| Technical measurement: server-side tracking, conversions APIs, consent mode, attribution | Mid to senior |
| Stakeholder / client management | Mid onward; C-level narrative at senior |
| Forecasting and budget planning | Senior |
| Cross-channel allocation (blended ROAS, MER) | Senior |
| Incrementality: geo holdouts, conversion lift, MMM | Senior to lead |
| Team management, hiring, vendor/rep relationships | Lead |
When more than one skill in the rung above is missing, build them in this order rather than top-to-bottom:
- efficiency: unit economics > creative testing > technical measurement > cross-channel allocation > forecasting > incrementality
- value (what the next rung's screen actually pays for): incrementality > forecasting > cross-channel allocation > technical measurement > creative testing > unit economics
- effort: incrementality > forecasting > cross-channel allocation > technical measurement > creative testing > unit economics
The two orders are near-inverses, and that is the useful finding: the skills the ladder pays most for are the ones the user cannot self-serve. Unit economics leads because a week of study changes how every answer is read, at any employer.
Incrementality is what this order starves: first on value, first on effort, last on efficiency in every session it is ranked. Promote it deliberately once the user is aiming at lead or head of growth, or already sits in a seat with the spend to run a holdout. Below that rung, last is the right place for it.
From cross-channel allocation downward, the roadmap step is negotiating access: budget, a second channel, enough spend to hold a geo out, not studying. Write it as an ask with a date, not a learning task.
Two observable thresholds anchor every roadmap:
- Budget and forecast responsibility is the senior-track signal. If the employer is not handing it over, the user is not on the senior track there, whatever the title says.
- At least one completed incrementality test (geo holdout or conversion lift) is the lead/head-of-growth signal, expected above roughly $10M in spend.
A roadmap that does not name how the user will hit these two signals is a wish list.
## Certifications
State it plainly: platform certifications are table stakes with weak standalone hiring signal, and the whole menu ranks below one published teardown. Effort here is exam hours plus a renewal that returns every year - the exam fee is a detail, not the axis.
- efficiency: Google Skillshop > LinkedIn Marketing Labs > TikTok Academy == Amazon Ads > Meta Blueprint
- value: near-identical and low across all five - none of them moves a hiring decision on its own; the absence of the free ones reads as carelessness
- effort: Meta Blueprint (a paid exam on top of the study) > Skillshop == LinkedIn == TikTok == Amazon
Those ties are decided, not deferred. The four free certifications tie on effort because effort here has exactly two components and both match at the same order of magnitude: about an hour of open-book exam, then one renewal a year. The TikTok/Amazon tie on efficiency is a tie only while the channel is undecided - answer 4 breaks it every time, short-form video putting TikTok Academy above Amazon Ads and retail media reversing them.
The facts behind the ordering:
- Skillshop is free: 75-minute open-book exam, 80% pass, 12-month expiry.
- Meta Blueprint runs ~$150 per exam.
- LinkedIn Marketing Labs, TikTok Academy and Amazon Ads are free.
Take the free ones and diarise the renewal. LinkedIn Labs moves up for a B2B target.
Blueprint earns its exam only when a target listing names it or an employer's partner status depends on it. When the user cannot fund the exam, or the certification is not offered in their market, delete it from their menu by name instead of leaving it ranked last - it returns later as an unfunded action item otherwise. Never let any of them occupy the space a results portfolio should fill.
The rollup explains the contradiction the user will notice: nearly every listing demands certifications, yet practitioners dismiss them ("A certificate is a certificate. It doesn't show how many attempts you needed" - Store Growers). Both are true because individual certifications are tracked per user and roll up into an agency's partner-program status, unlocking the partner badge, beta access, and rep support.
Listings require certs as an agency operational need, not because they predict candidate quality. Skillshop teaches the vendor's products as the vendor wants them used, not how to run profitable campaigns.
## The NDA-safe portfolio
Almost every buyer's best work belongs to a client. Four techniques let it travel:
1. **Anonymise the client, keep the shape.** Replace names with descriptors that carry the credibility: "Series B B2B SaaS company", "regional healthcare provider", "DTC apparel brand".
2. **Index and percentage, never absolutes.** Percentage change only, indexed values (baseline = 100), or aggregated ranges - never real revenue or spend figures.
3. **Strip private assets and internal screenshots.** Raw ad-account screenshots showing account names and spend are a legal risk. Lead with problem, process, directional outcome; state that visuals are anonymised.
4. **Own data where possible.** Self-funded spend (small store, affiliate offer, pro-bono local business) and written public teardowns of a brand's ads carry zero confidentiality risk - and a teardown rehearses exactly what live-audit interviews test.
Techniques 1-3 are a checklist, not a menu: apply all three to any client work. Technique 4 is where the real choice sits, and it decides which source the portfolio leads with:
- efficiency: public written teardown > anonymised client work > self-funded own spend
- value: anonymised client work (real budget, real stakes, a result someone paid for) > self-funded own spend > teardown (judgment only, no outcome)
- effort: self-funded own spend (a quarter of evenings plus your own budget) > teardown (an hour, public ads, and it doubles as live-audit rehearsal) > anonymised client work (an hour a piece; the work already exists)
- compliance cost: anonymised client work (re-read the NDA or client contract before publishing, and a breach cannot be undone) > self-funded own spend == teardown - an exact tie, not a rounding: neither owes an obligation to a third party, so neither has anything to breach
That compliance line is why the teardown leads despite the weakest raw value: it is the only source with no irreversible downside. Lead with anonymised client work as soon as the contract is checked and the user has any. When everything is under NDA and unclearable, the teardown is the whole portfolio and self-funded spend is what makes it more than commentary.
The failure runs both ways: showing nothing because "it's all under NDA", and over-sharing - interviewers read leaked client data as a preview of how the candidate would treat theirs. Template, worked example, and a negative example: [references/anonymised-case-study-example.md](references/anonymised-case-study-example.md). If the user will publish a written piece, offer a pass with your preferred humanizer skill before it ships.
## Interview preparation
Five formats recur in verified candidate reports, listed below in the order to spend rehearsal hours on them. Full question bank with sample answers: [references/interview-question-bank.md](references/interview-question-bank.md).
- efficiency: live audit > metric diagnosis > mental maths > behavioural stories > take-home polish
- value (how much of the decision each format carries): live audit > take-home > metric diagnosis > behavioural stories > mental maths
- effort: take-home (12-30 unpaid hours if left unbounded, 2-4 if time-boxed) > live audit (a week of rehearsing on real accounts) > behavioural stories (an evening banking them in three lengths) > metric diagnosis (an evening per diagnostic order) > mental maths (an hour a day of drills)
The take-home inverts because it is the one format whose effort the candidate controls and whose extra hours buy almost nothing: a set take-home is a deadline, not a choice, so cap it and spend the residue on the live audit. Which formats appear is the employer's call - reorder against the process the user actually faces as soon as they know it.
1. **Live audit / account teardown** - the signature format. Interviewers screen share a real or example account and watch the candidate dissect it. The named failure: presenting a generic tool-generated audit - bring a prioritised point of view instead. Rehearse aloud; practising with mbfinotti/advertising-skills@ad-account-diagnostic builds the structure.
2. **Metric-diagnosis scenarios** - e.g. "high CTR but low conversion rate - why?", "campaign not delivering impressions - why?", "this A/B test shows a 5% lift with p=0.08 - what do you do?", "what would you do with a $50K/month budget?". Prepare structured diagnostic orders, not lists of possibilities.
3. **Mental-maths / unit-economics case** - calculate funnel metrics (CTR, CPC, CPA) down the purchasing funnel and compare two companies' revenues and gross profits, verbally or in a spreadsheet round.
4. **Behavioural / client round** - a reported shape: HR screen, then a technical round ("What KPIs would you look for and how would you report them?"), then an executive character round. Prepare stories in three lengths (2 minutes, 60 seconds, one line) across ownership, conflict, failure, and results.
5. **Take-home case study** - companies brief "a few hours"; candidates report 12-30+ hours, and idea theft is documented. Time-box to 2-4 hours, state assumptions explicitly in the deliverable, and push back on any brief that amounts to a full go-to-market plan for an unlaunched product.
## In-house vs agency vs freelance
| Dimension | Agency | In-house | Freelance |
| ---------------------- | ------------------------------------- | ------------------------------------------ | ------------------------------------ |
| Learning speed/breadth | Highest - many accounts and verticals | Deep on one business | Variable, self-directed |
| Compensation | Lowest at junior, capped at mid | Highest at senior/lead, especially US tech | High day rates, volatile |
| Strategy vs execution | Execution first, strategy later | More strategy, cross-functional | Both, client-dependent |
| Burnout risk | Highest - volume plus client demands | Moderate | Self-managed, income risk |
| Exit options | Strong, into in-house | Leadership or another in-house | Back to employment, or agency-of-one |
The ranking flips at mid-level, and the flip is the whole answer - one ordering, one condition, not two rival orders:
- efficiency, junior to mid: agency > in-house > freelance - reps per year is what the user is short of, and the agency buys the most of them.
- efficiency, mid onward: in-house > freelance > agency - reps stop being scarce, and staying agency past mid-level fights the compensation curve.
- effort: agency == freelance > in-house - the two cost different things, so read them as a tie only in size: the agency costs hours and burnout risk, freelance costs income stability and a standing sales job.
Re-rank on what the user is actually short of rather than on the stage alone:
- Short of autonomy: takes freelance early, despite the order.
- Short of money: takes in-house early.
- Has an existing client network: erases most of freelance's effort.
Compliance cost does not separate these seats - all three run the same platform policies and the same client contracts - so it gets no axis here.
The burnout numbers are survey-grade, from Marketing Week's 2026 Career & Salary Survey (2,350 respondents), over the prior 12 months:
- 65.3% of marketers felt overwhelmed.
- 60.7% felt undervalued.
- 55.1% felt emotionally exhausted.
- 51.2% saw their remit grow without matching pay.
## Progression and the promotion case
Tie every promotion ask to an owned revenue or P&L number, never to tenure or activity. Cases citing years served, hours, campaign counts, or effort fail. Cases citing an owned number succeed, because the ladder pays for accountability at every rung.
Scope creep is common enough to be a lever rather than a grievance - the 51.2% remit-growth-without-pay figure above - but only once converted into an owned-number argument: name what you now own that you did not, and what it did. One-pager template with a positive and a negative example: [references/promotion-case-one-pager.md](references/promotion-case-one-pager.md).
## Compensation benchmarks
Quote figures with source, date, and grade; treat ranges as meaningful, never points.
**UK - survey-grade.** Marketing Week 2026 Career & Salary Survey (2,350 respondents, 16 April 2026), declines across most sectors and every seniority level:
| Level | 2026 | 2025 |
| -------------------------------- | -------- | -------- |
| Executive / assistant / graduate | £33,815 | £36,106 |
| Junior manager | £43,173 | £47,840 |
| Senior manager | £60,577 | £72,434 |
| CMO / director / VP | £111,082 | £117,623 |
- 3Search Performance Marketing Salary Guide 2025 (1,300+ responses): PPC Managers £50,000-£70,000.
- itjobswatch (6 months to 17 July 2025): median PPC contractor day rate £300, down ~15% YoY.
**UK - aggregate, directional.** Major Players (2026):
- PPC Executive ~£29,590
- PPC Specialist ~£37,417
- Director up to ~£85,536
**US - aggregate, lower reliability.** Growth Marketing Manager:
- Robert Half: $91,250-$133,250
- Salary.com median: ~$119,945
- A recruiter guide cites $115,000-$140,000 base for 2026
Head of Growth:
- Growth.Talent benchmark (refreshed 12 August 2026): $180,000-$280,000 base plus equity
- Seed-to-Series-A: $130,000-$180,000 with 0.5%-1.5% equity
Levels.fyi marketing median total compensation: $175,000 (equity-inclusive, tech).
**EU/France - partial.** BSPCE tax-advantaged options are standard at French startups, typically 0.1%-1% for growth roles; French growth pay generally sits below US and often below top UK. Specific French paid-media salary bands (APEC, Michael Page France) could not be verified - say the data is unavailable rather than substituting a guess.
**Variable pay changes the comparison.** UK agency roles commonly carry revenue-share bonuses; in-house startup roles weight equity at 20-50% of total compensation - a base-only comparison between the two is misleading.
## B2B vs B2C
Do not rank B2B against B2C. The table below compares two markets a user moves between, not two options competing for the same hours, and the closing paragraph is the reason: the transferable skill set is the same one, so an efficiency ordering here would be false precision dressed as advice.
Where the paths genuinely diverge:
| Dimension | B2B | B2C / DTC |
| ------------------------ | ------------------------------------------------------- | ---------------------------------------------------- |
| Primary channel families | B2B professional networks, paid search | Paid social, short-form video, shopping/feed formats |
| Core metrics | Pipeline, MQL/SQL, CAC, closed-loop to the sales system | ROAS, MER/blended ROAS, contribution margin, CPA |
| Sales cycle | Long; sales alignment critical | Short; purchase-driven |
| Creative volume | Lower; account-based personalisation | Very high; constant creative testing |
| Compensation link | Often tied to pipeline influence | Often tied to efficiency and growth |
These transfer unchanged, so a B2B buyer can move to DTC and back:
- Platform mechanics
- Bidding logic
- Tracking and attribution setup
- Budget pacing
- Audience construction
- Testing discipline
- Unit-economics literacy
- Diagnostic mindset
The re-learning is narrow:
- Creative cadence, when moving to DTC.
- Sales alignment and long-cycle attribution, when moving to B2B.
Present this as an inference from comparing real job descriptions - the channel list changes, the skill list does not - not as a measured statistic.
## The automation shift
Automated campaign types dominate, and the role is shifting "from keyword manager to system optimizer" - upstream in the inputs, mid-campaign in the signals, downstream in the measurement (Search Engine Land, 3 June 2026). Position the user accordingly.
- **Commoditised** - do not build a career case on these: keyword-by-keyword bidding, manual audience building, manual bid adjustments, routine reporting assembly.
- **Appreciating** - the roadmap's growth areas: creative strategy and testing (practitioner consensus holds creative roadmapping has no AI replacement yet), measurement and incrementality, first-party data and feed/catalogue work, AI-creative and prompt workflows, and the judgment to know what the automated system should be doing and whether it is.
Which appreciating area to build first:
- efficiency: AI-creative and prompt workflows > first-party data and feed/catalogue work > creative strategy and testing > measurement and incrementality > system judgment
- durability of the payoff: creative strategy and testing > measurement and incrementality > system judgment > first-party data and feed/catalogue work > AI-creative and prompt workflows
- effort: measurement and incrementality (a seat with the spend to hold a geo out) > system judgment (years of watching automated systems misbehave) > creative strategy and testing (a quarter of live tests) > feed/catalogue work (a week) > AI-creative workflows (an evening)
The two orders contradict on purpose: the cheapest area to pick up is also the one commoditising fastest, so it wins an interview this quarter and expires by the next ladder rung. Lead with AI-creative workflows on a hard deadline (question 7), lead with creative strategy and testing on a compounding mandate (question 8), and never present the cheap end as a career case on its own.
The factual correction worth drilling: **third-party cookies are not being deprecated in Chrome.** Google announced on 22 July 2024 it would not phase them out, and confirmed on 22 April 2025 it would not add a user-choice prompt.
Cookies remain enabled by default in Chrome, while other browsers already block them. A large amount of career content still asserts the opposite - repeating the deprecation claim in an interview dates the candidate.
What durably reduced signal is Apple's ATT (2021), which drove the over-attribution problem and the industry move to blended metrics and incrementality. The senior-level trap that follows: optimising to platform-reported ROAS without being able to explain why it diverges from blended MER - the autobidder optimises toward what it can see, which is the number the OS-level privacy change broke. mbfinotti/advertising-skills@ad-attribution-gap covers the mechanics of that divergence.
## Failure modes
- Leaning on certifications instead of results - cert-only resumes are a documented rejection pattern.
- Reading a tool-generated audit aloud in a live-audit round - interviewers screen for exactly this.
- Vanity metrics over business outcomes - talking CTR and CPC without CPA, ROAS, MER, or pipeline reads as mid-level regardless of years served; senior conversations run on MER, contribution margin, and CAC payback.
- Sinking 12-30+ unpaid hours into a take-home briefed as "a few hours" - time-box to 2-4 hours, state assumptions.
- Over-sharing confidential client data in a portfolio or interview - read as a preview of how the candidate treats data.
- Missing incrementality and attribution at senior level - the clearest ceiling signal.
- Building a promotion case on tenure or activity instead of an owned number.
- Quoting an undated or last-year pay figure in a negotiation while the surveyed market is declining.
## Objective and measurement
A career deliverable produced with this skill - roadmap, interview-prep plan, portfolio piece, or promotion case - passes only when all five hold:
1. Every recommended step names the observable signal that proves it happened: budget/forecast responsibility granted, an incrementality test completed, a portfolio piece published, a named interview stage reached.
2. Every pay figure quoted carries currency, source, date, and grade (survey vs aggregate).
3. Every portfolio artifact passes all four NDA checks (descriptor not name, indexed not absolute, no internal screenshots, owned or public data where possible).
4. The plan sets a dated benchmark-to-change-course. Defaults:
- Entering: no interviews after 6 months of applications means the gap is demonstrable results, not more certifications.
- Mid-level: no budget or forecast responsibility within the agreed window means the senior track is blocked at the current employer, and the plan moves to changing seats.
5. Every menu the deliverable offers meets all of these:
- Ordered by value per unit of the user's effort.
- The effort axis is named in years, hours, or reversibility, never money.
- The answers to questions 7-9 that moved the order are stated.
- The high-value/high-effort option the order starves is named, with the condition that promotes it.
- Every option the user's constraints rule out appears as deleted by name, rather than ranked last.
Rework the deliverable until all five criteria pass - do not present a plan with a known failed check. Post-delivery, the plan is validated when the user hits the named signal, and re-planned (not padded) when a benchmark-to-change-course date arrives first.
## References
- [references/interview-question-bank.md](references/interview-question-bank.md) - the verified question bank per format, with sample diagnostic answers and the live-audit structure. Load when prepping any interview.
- [references/anonymised-case-study-example.md](references/anonymised-case-study-example.md) - NDA-safe case-study template, a worked example, and a negative example. Load when building the portfolio.
- [references/promotion-case-one-pager.md](references/promotion-case-one-pager.md) - promotion-case one-pager template with positive and negative examples. Load when making a promotion case.