Flagship · 14 weeks
The Cohort Atlas
A full reading of retention, conversion, and paywall instrumentation for consumer and B2B apps.
Ridsdale atelier · est. 2018
Product data, read with the patience of a conservator.
We do not sell dashboards. We teach the grammar that makes them honest.
Connectvectorgrid trains product and growth teams to instrument mobile and web apps so that a retention curve can be trusted in a board pack. The work is slow on purpose: naming events, drawing cohorts, and refusing vanity totals that collapse the week a marketer asks a sharper question.
From the studio ledger
Most app analytics programmes fail at the taxonomy, long before anyone argues about a north-star metric. We start there, and we stay there until the names can survive a rebrand.
61product teams taught since the first Ridsdale sitting
486event schemas reviewed in critique, including the ugly ones
37live studios held in London, Manchester, and remotely
The journal
Recent studio notes
Event names that survive a rebrand
Why “Signup_v3_final2” is a debt instrument, and how UK teams keep a catalogue when marketing changes the wrapper.
Why DAU misleads subscription apps
Daily actives flatten trial, billing retry, and genuine habit into a single polite number.
Reading a retention curve without panic
The first dip after day two is rarely a product emergency. Sometimes it is a tracking seam.
Programmes
What the house currently teaches
Each programme is a studio, not a video dump. Seats are limited because critique does not scale past a table you can still see across.
How the house works
We treat instrumentation as a literary problem.
An event name is a sentence. If two teams would gloss it differently, the chart that follows is already a negotiation, not a measurement. Our tutors come from product analytics desks in retail, media, and fintech; they have lived the moment a Mixpanel property is overloaded until it means nothing.
Students leave with a written taxonomy, a cohort plan, and a short list of questions they are willing to leave unanswered. That last part is taught on purpose. App analytics is full of numbers that look decisive and are not.
The Signal Studio method is the house philosophy behind every programme.
Voices from the table
What students actually used
The Cohort Atlas module on paywall exposure finally explained why our “activated” users looked healthier than finance said they were. We had been counting a preview screen as a start.
Event Taxonomy Studio was slower than I wanted in week two. Imogen would not let us ship the catalogue until the “purchase” event stopped meaning three different things. Annoying then. Correct now.
An invitation, not a funnel
If your board pack still leans on a single daily-active line, write to the house.
We answer within three working days. Fees are listed without theatre. There is no checkout on this site; a seat is confirmed by correspondence after we are sure the programme fits the work you actually do.