These notes come from post-course forms, letters, and two longer case write-ups we were allowed to publish. Attribution styles differ because people differ in how they wish to be named.
I came for Retention Cartography and stayed for the argument about when a week starts. Our grocery app treats Sunday night as a second Saturday; the seven-day charts we had copied from a US blog were quietly lying. The map we left with is uglier and finally matches how the shops actually close.
Owen Hale · Head of product, regional grocer · Cardiff
Solid on activation. Thin on monetisation in the four-week desk — we should have booked Revenue Leak Mapping instead of hoping it would sneak in.
Verified platform-style note · 4/5 · March 2026
Harriet made me throw away a north-star slide I had used for two board packs. I was annoyed. She was right. The replacement is a paragraph.
Amelia · Edinburgh
The Signal Desk method is slower than our sprint cadence. We now write the note on Thursdays and present on Mondays. The gap is the point. I would not send a junior who still needs a tour of the tool; they would sit politely and miss the knife.
Client in digital banking · United Kingdom
Store Page Forensics taught me to stop treating a ranking jump as a story. We now report impression share and custom page conversion as two different animals. Reviews still move in ways we cannot explain; the course did not pretend otherwise.
Nia Okonkwo · Growth, independent fitness app
I missed two live Thursdays to a launch. The recordings are clear. They are not the same as being in the room when someone dismantles your window. Budget the hours or wait.
Samir, Birmingham · Reading Product Signal
A mid-size travel organiser in the North of England arrived with a single D30 number they presented as health. In Reading Product Signal they split the population into people who had booked, people who had only searched, and people who opened the app because a trip was already happening.
The “healthy” line was the third group. The first group — the one that paid — was already thinning after a painful checkout experiment. The team paused a planned homepage redesign and repaired the fare-hold flow instead. We did not “increase retention”; we stopped mixing guests with customers. That is the kind of win the house prefers to advertise.
Published with permission · names withheld · winter 2025 table
An independent language app showed a sharp lift in trial starts after a new paywall. Revenue Leak Mapping — taken as a private desk after the public flagship — followed the same users to the second invoice. Trial starts were up; second-month remaining was not. The new wall was recruiting the curious, not the committed.
They kept the wall, changed the trial length, and wrote a Signal Note that a trustee could read without a glossary. One founder later told us the course felt “slightly monastic.” We accepted the description.