mbfinotti/advertising-skills / skills/ad-buyer-group-mapper
ad-buyer-group-mapper
VerifiedMap the likely buying committee for an offer, deal size, and industry, and give each role a messaging angle plus the ad-targeting proxy - job function and seniority, account list, account-level intent - that actually reaches them, sequenced by buying stage. Use whenever the user mentions a buying committee or buying group, decision makers, influencers or blockers, who signs off, targeting the economic buyer or a champion, per-role ad messaging, or whether one broad message beats per-role targeting - even if they never say 'buying committee'. Covers B2B committees and multi-decider household purchases. Do NOT use to size or build the audiences themselves - use mbfinotti/advertising-skills@ad-audience-targeting instead.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 29.0 KB |
| evals/evals.json | config | 27.7 KB |
| references/worked-example.md | markdown | 14.4 KB |
references/worked-example.md
View on GitHub# Worked Examples
Filled-in outputs in the shape defined by SKILL.md's Output shape section, plus starter per-role patterns. All angles below are illustrations of the method, not benchmarks - verify every pattern against the user's own closed-won data, call notes, and reviews before use.
## Starter per-role patterns (defaults to verify)
Practitioner-claimed defaults, not controlled evidence. Use them to seed the map when the user's own evidence is thin, and mark every row built from them as `hypothesis`.
Read the `Proof / offer type` column against the offer-type ranking in SKILL.md § Per-role output: this table says which offer fits a role, that ranking says which one to build first when you cannot build them all.
| Role | Measured on | Personal risk | Proof / offer type that fits |
| ------------------- | ---------------------------------------------------------- | --------------------------------------------------------------- | ----------------------------------------------------------------------- |
| Champion | The pain metric the offer fixes | Sponsored a failed project; spent political capital for nothing | Comparison guides, business-case builders they can circulate internally |
| End user | Daily throughput / quality of their own work | Forced onto a tool that makes their job worse | Demo, free trial, hands-on content - no sales gate |
| Economic buyer | Budget efficiency, ROI of the line item | Approved spend that produced no return; CFO scrutiny | ROI calculators, benchmark reports, customer proof with numbers |
| Technical evaluator | System reliability, integration cost | Owning the integration when it breaks | Technical docs, architecture pages, security whitepapers |
| Gatekeeper-blocker | Risk avoided: breaches, compliance findings, bad contracts | Being the one who signed off before an incident | Compliance certifications, audit reports, security documentation |
## Example 1 - B2B, per-role map justified
**Input:** compliance-automation SaaS, $85K ACV, selling to mid-market fintech (200-1,000 employees), ~350 named target accounts, closed-won notes from 22 deals available, professional network + broad social, $18K/month.
**Committee summary.**
- Committee size: closed-won notes show a median of 4 distinct roles per deal (range 3-6), consistent with the TrustRadius mid-market band, well under the "6 to 10" complex-purchase figure.
- Purchase type: mostly net-new category.
- Precision verdict: 350 named accounts and $85K ACV clear both thresholds (<500 accounts, approaching $100K ACV); per-role creative is defensible.
- Trade-off declared: this map takes the precision side, since the 95-5 objection applies less when reach is already bounded at 350 accounts.
**Role map.**
| Role | Evidence | Measured on | Personal risk | Likely objection | Messaging angle | Proof/offer | Targeting proxy | Proxy size vs. floor |
| ----------------------------------------- | ---------------------------------------------------------------------------------------- | -------------------------------------------- | ----------------------------------------------------------- | -------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Champion (Head of Compliance) | Evidenced - present in 22/22 won deals | Findings closed per audit cycle | Sponsoring a tool that fails the next audit | "We already handle this in spreadsheets" | "Your next audit prep in days, not quarters - without adding headcount" (verbatim from 3 call notes: "audit prep eats my quarter") | Audit-readiness checklist, ungated | Account list + function: Legal & Compliance, seniority: Director+ | ~2,100 - below practical floor; acceptable only because account-list ABM campaigns run on engagement objectives, flagged for mbfinotti/advertising-skills@ad-audience-targeting |
| Economic buyer (CFO / VP Finance) | Evidenced - 18/22 won deals show finance sign-off | Cost of compliance program vs. fine exposure | Approving a five-figure line that duplicates existing spend | "What does this replace?" | "One line item that retires three tools and caps fine exposure" | ROI model comparing tool consolidation + penalty avoidance | Account list + function: Finance, seniority: VP+ | ~1,400 - same flag as above |
| Technical evaluator (Head of Engineering) | Hypothesis - appears in 9/22 notes; disproof: cut if absent from next 10 wins | Integration and maintenance load | Owning a brittle integration | "Another vendor API to babysit" | "Read-only connectors, no schema changes, sandboxed in an afternoon" | Architecture doc + sandbox access | Account list + functions: Engineering AND Information Technology (cross-functional role straddles both), seniority: Director+ | ~3,800 |
| Gatekeeper (Security review) | Evidenced - validation-stage security review in 14/22 notes, reopened requirements twice | Vendor risk accepted on their signature | "New vendor, new attack surface" | "The vendor assessment is pre-filled - SOC 2 Type II, pen-test report, DPA on request" | Security documentation pack, ungated | Same account list + function: Information Technology, seniority: Manager+, reached from supplier-selection stage onward - before validation | ~2,900 |
**Sequencing notes.**
- Champion and end users (merged into champion here, same people in this segment per call notes): problem identification.
- Economic buyer: business-case content engagement.
- Security pack: delivered during supplier selection, not validation, because two deals stalled when security arrived late.
- Economic-buyer and gatekeeper angles weight risk reduction over urgency: 15/22 lost-deal notes cite "decided to wait", matching the indecision pattern.
**Handoffs.** Audience construction, floors resolution, and budget: mbfinotti/advertising-skills@ad-audience-targeting. Copy per angle: mbfinotti/advertising-skills@ad-copy-variants.
## Example 2 - multi-decider household purchase
**Input:** residential solar installation, $22K average ticket, metro region, no platform decided, review mining of 180 installer reviews available.
**Committee summary.**
- Household type: two-decider (typically both partners), occasional third voice (adult child or contractor friend advising).
- Purchase characteristics: high-consideration, personal money, no formal review stages.
- Precision verdict: household-graph isolation is imprecise; run role-differentiated creative on one shared household audience instead of trying to target each member.
**Role map.**
| Role | Evidence | Measured on (cares about) | Personal risk | Likely objection | Messaging angle | Proof/offer | Targeting proxy |
| -------------------------------- | --------------------------------------------------------------------------------------------------------------- | ------------------------------------------------ | --------------------------------------------------- | --------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------- |
| Initiator-researcher (partner A) | Evidenced - 130/180 reviews name one partner as the researcher | Monthly bill reduction, doing the homework right | Championing a purchase the household regrets | "Payback math never works out" | "See your actual payback year from your last three bills" (verbatim review language: "the calculator matched our real bills") | Bill-based savings calculator, ungated | Geo + homeowner signals + category in-market behavior; retargeting pool from calculator use |
| Co-decider (partner B) | Evidenced - reviews repeatedly cite "my wife/husband was worried about..." | Not being locked into a bad contract | "20-year contract with a company that might vanish" | "Rated installers, transferable warranty, no lien on the house" | Warranty terms one-pager, third-party review scores | Same household audience - connected-TV and shared-device placements carry the risk-reduction creative to the second decider |
| User (whole household) | Hypothesis - reviews rarely mention post-install experience; disproof: cut if next review pass confirms silence | Nothing changes day-to-day | (none - merged into co-decider's risk frame) | - | - | - |
The user row is cut at the structural pass: no evidence, and its concerns are absorbed by the co-decider. Final map: two roles, two creative variants, one shared audience.
## Negative example - what not to ship
**Input:** $6K ACV email-deliverability tool, SMB buyers, no closed-won analysis, professional network, $3K/month.
**The bad map:** eight roles (CEO, CMO, VP Marketing, Marketing Ops, IT, Procurement, Legal, "Influencers") copied from a committee framework, each with its own campaign at ~$375/month, titles targeted exactly ("Director of Email Marketing"), each audience 800-3,000 people.
Why it fails every gate:
- No evidence: all eight roles are title-based inference - the weakest source - asserted, not hypothesized, with no disproof tests. TrustRadius's SMB band (2-3 people) and the ACV both predict a two-role committee.
- "Influencers" is not a role: it maps to no attribute and no angle.
- Every audience is below the ~5,000 significance floor and the 20,000-50,000 practical floor; $375/month cannot fund a learning phase anywhere.
- Exact-title targeting misses every variant of "Director of Email Marketing" the platform doesn't recognize.
- Procurement and legal do not exist as stages in a $6K self-serve purchase - they were imported from an enterprise framework, not from evidence.
- The precision thresholds point the other way: neither <500 named accounts nor >$100K ACV - one strong message to a broad marketing-function audience wins on economics.
**The fix:** two roles: champion-user in marketing, and economic buyer only as a proof point in the creative, not a separate audience. One consolidated campaign, function + seniority targeting, and a disproof test on the economic buyer ("add a finance-facing variant only if call notes show finance blocking deals").
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |