If you've followed this series, you know I've been running our microtest framework on us.

Sprint 1 found the audience. Sprint 2 found the headline. Sprint 3 found the offer format

Something unexpected came out of it: the research methodology itself became the product.

Quick recap: one sprint, one variable, one answer.

Each costs ~$35 in ad spend and runs for 48 hours. Pain points first, then headlines, then offer format, then creative. Before any sprint runs, there's a research phase—dig into the prospect's market, find their customers' actual language, and write 13 "True or False?" ad variations. After five sprints on our own business, I realized that the research phase alone is more valuable than most agencies' entire pitch decks.

So we packaged it.

We call it an Offer Intelligence Report—a personalized document built from scratch for one business before we ever reach out.

Two-track research. We scrape their website—positioning, pricing, and CTAs. Then go wider: Google reviews, Reddit, G2, forums. Track A: what the internet says (unfiltered). Track B: what their website says (curated). When those contradict—and they often do—that's where the real insight is.

Sprint 1 copy. From the research, we write 13-15 "True or False?" ads—the exact copy we'd run on Meta. Each test one pain point in the first person, in their customer's language.

These come from independent research—real language from Reddit, G2, and Google reviews.

The prospect reads them and thinks: "how did you know this about my market?" The math. If Sprint 1 finds a 5%+ CTR winner at their industry's CPL with conservative close rates—here's their monthly pipeline. Each report costs ~$0.15 to generate.

A traditional agency charges $2K-$5K for the equivalent.

When Sprint 1 runs, data flows into a dashboard we built.

Every variant ranked — CTR, CPC, impressions, winner flag. Pain points clustered by theme.

The surprise: clients study the losers more than the winners. "I need a better funnel" at 1.8% CTR next to "I don't know where my last 5 clients came from" at 7.2%—assumptions break apart in real time.

We're building more. Another week.

Your first touch should be work, not a pitch. A 2-minute Loom tearing apart their landing page. A list of 5 headlines you'd test. Anything that shows you already thought about their specific problem.

Everyone is sending AI-generated cold emails. The founder who sends actual work wins by default.

Want to see what an OIR looks like for your business? Reply with your URL — I'll show you the first thing I'd test.

Or book a 48-hour microtest here.

Next week: Sprint 4 — Creative. Does the image actually matter? Five visual concepts. One answer