Three steps, at the pace of your evidence.

Join the queue. When your place is ready, start with your workbook.

We release pilot places in batches and email the workbook matched to your situation, the Pilot Guide and The Method when your access is ready. Time in the queue doesn’t count toward your four weeks. You can switch to a different workbook at any time.

Start with your workbook. The Pilot Guide explains the timeline. The Method is there if you want the thinking behind it. Do Step 1 on your own, before discussing your answers with your team.

See inside the workbooks

• Step 1. The leader answers six questions alone, in writing, in about an hour. Each answer is marked Known or Assumed.
• Step 2. Five people from five functions answer the same questions on their own, about 20 minutes each. Optional for solo founders.
• Where answers differ is a gap. No gaps doesn’t mean the team is right, so everyone goes on to Step 3.

Write down what you believe. See if the team agrees.

Aim to complete Steps 1 and 2 within four weeks of receiving your materials. When you finish, request your outside read from your pilot page. Reports are scheduled separately.

Check it against customers.

Collect what customers say without being asked, log what they do that doesn’t fit the plan, and run the four signal checks: repeated, consistent, costly, lasting. Until all four are Yes, a behavior is a lead, not a pattern.

You finish with a one-page Brand Brief: the customer role your evidence supports, what remains uncertain, and what to test next.

Seal your findings. Then see the outside read.

Once we start your report, we research your market from public evidence only: a Brand Intelligence Report if you have customers, an Idea Intelligence Assessment if you don’t. We never receive your workbook or internal findings.

See a specimen outside read (fictional company)

See a sample outside read for each edition · Follow one fictional company through the whole pilot

When your report is ready, we ask you to confirm your completed internal findings are saved and dated. Only then do we release it. Keep the original; record any later revisions separately.

Run the Crosscheck yourself.

Set your sealed findings beside the outside read, claim by claim, using the Crosscheck worksheet and the optional AI assistant file.

Crosscheck is participant-led. A facilitated review isn’t part of the pilot.

• Supported. Both reads point the same way. That strengthens a claim; it doesn’t prove it.
• Challenged. The reads differ. The better-supported evidence wins, whichever side it came from, and the reason is written down.
• Unresolved. Neither side can support it yet. It stays open, with the next test named.