St. Pierre · Builder SOW · Phase 1

Custom Audit Engine

The keystone build. Turn a prospect's basic info into a personalized set of six assets — automatically.

★ AUDIT-1 · owner: Collens · reused across onboarding, outbound & lead magnets
Why this is the keystone Build this once and it powers three things: nurturing $1 buyers toward the first $975, sending value to cold high-value prospects (outbound), and future lead magnets. Everything in Phase 1 and 2 leans on it — so it's the highest-leverage thing you'll build.

Objective

When a prospect pays the $1 or books, the engine generates and delivers a personalized pack — six audits/assets specific to their business — with no human assembling each one. During your proof phase everything is a draft for Akash to review, not an auto-send.

The six assets it produces

AI-Readiness AuditWhere the contractor is vs. an AI-run demand engine; the gaps.
Bottleneck AuditThe one constraint throttling booked jobs right now.
Brand AuditHow their brand reads to a $1M–$10M owner; quick wins.
1M→10M Growth SystemThe path from their number to the next tier, mapped to their business.
The GuaranteeThe St. Pierre promise, framed to their risk profile.
Custom VSL / DemoA short walkthrough personalized to their market and service.

How it works (build stages)

  1. Intake — capture business name, market, service, website, and the form answers (from the $1 flow / Typeform).
  2. Enrich — pull what's publicly knowable (site, reviews, market) into a per-prospect record.
  3. Generate — produce the six assets from templates + the enriched record. Same input → same repeatable output.
  4. Assemble — drop them into one personalized result page / pack per prospect.
  5. QA — checks for missing fields, off-brand claims, and unsupported proof.
  6. Hand back — the pack + a manifest + a Loom, for Akash to review (no live send during proof).

Inputs from Akash

Definition of done

RepeatableOne prospect record in → all six assets out, on a personalized page, without manual assembly.
On-brandReads like St. Pierre, uses only approved claims, no fabricated proof.
DocumentedManifest of inputs/outputs + a ≤5-min Loom; another session can re-run it.
Reusable hooksSame engine can be fed a pre-filled record for outbound (see OUT-1) — not a one-off.
Guardrails No live sends, posts, or DMs during proof — outputs are drafts. No fabricated results or client names. No credentials in files. If you think you need a live account or key, stop and write the exact ask.
Hand-back format 1) Output link · 2) What you built · 3) How to re-run · 4) What failed · 5) What you need from Akash.
St. Pierre · Phase 1 keystone · feeds the $1→$975 machine, outbound (OUT-1), and future lead magnets. Full board: team.stpierre.ai/builderpipeline