1. Stage contracts
Who may enter, what exit looks like, and which events are forbidden from quietly promoting someone. You draft the contract on your own product, not a toy app.
Flagship · six weeks
Reading the Drop
A Product Funnel Analytics studio for product managers, analysts, and researchers who share a leaking pipeline and a noisy warehouse.
What you leave with
- A written stage contract for one live product, including recode rules.
- A cohort window a GB finance partner can rebuild from event exports.
- A leak hypothesis tied to one interview transcript, not a persona slide.
- A Thursday memo with kill criteria for the next experiment.
This is not a certification mill and not a vendor onboarding. You will spend more time arguing about denominators than choosing chart colours.
Modules
2. Windows that survive audit
Calendar-aligned cohorts, late events, and the temptation of rolling 30-day magic. We keep one boring window public and one experimental window private.
3. Leaky attribution
How last-touch returns through reporting shortcuts. You publish a mix table with uncertainty, not a single winning channel.
4. Qualitative overlay
One stall interview next to the drop. Sampling is ugly on purpose so you do not wait for a research quarter.
5. Experiment readouts
Lift expressed as a stage movement with sample caution. North-star charts stay in an appendix.
6. The Thursday memo
Assessed artefact: one page, kill criteria, owner, and a sentence on what would falsify the leak story.
Instructor
Amira Cole
Amira ran growth analytics for a Birmingham marketplace, then spent four years as an embedded analyst in two UK scale-ups before opening the Anchorhub desk. She still marks every Thursday memo herself in the flagship room.
From people who took this syllabus
Module 3 made me reprint our channel report with ranges. Marketing was annoyed for a week, then used the same table in a budget meeting.
The interview overlay felt thin if you cannot recruit quickly. I still finished the memo, but week four slipped.
Questions we actually get
Do I need SQL?
You need to export events or partner with someone who can. The course does not teach warehouse modelling from zero. That is a real limitation: analysts without export access stall in week two.
Is this a tool training?
No. We are tool-agnostic on purpose. Bring Mixpanel, Amplitude, BigQuery, or spreadsheets. We mark the contract and the memo, not the interface.
Can I take it as a solo founder?
Yes, if you can produce events. Pairing with a colleague is cheaper per head on the Pipeline Seat, but solo seats are welcome.