Experimentation planning consultation
A focused review of your product and data that ends in a prioritized experiment plan the team can start on.
Book an intro call
Who it’s for
Tests have no direction
Experiments run, but not against the biggest questions.
Results are hard to trust
Setup and measurement leave doubt about what a test shows.
No one owns the backlog
Ideas pile up with no order and no method.
Fit
Suitable when
You have a live product with enough traffic or usage to test against, and data the team can pull. You want a plan, not one-off tests.
Unsuitable when
You have too little traffic or usage to test, or no live product yet. Other methods fit that stage better.
How it works
1
Kickoff call
Your goals, the product, the stack, and what the team has tested or wanted to test.
2
Data and tooling
Access to analytics and any testing tools, with table notes or someone who knows them. First data request.
3
Hypothesis work
Tuesday and Thursday check-ins on candidate tests, sample sizes and open questions.
4
Plan presentation
Each proposed experiment, the question it answers, how to run and read it, and a ranked order.
5
Optional: run the tests
Stay on to run and read the experiments, scoped separately.
Your own hunches get framed as testable questions, not assumed. The idea with the most backing is often not the one worth testing first.
What you receive
Metrics and KPI review
Experiment candidate list
Test priority map
Design and measurement guidance
90-day testing calendar
If the data is too thin to support a test, I will say so instead of forcing one.
Examples
Which tests to run first with limited traffic
Test ideas lived in chat threads with no agreed order. I brought them onto a shared roadmap and scored impact and effort against the weakest parts of the business, giving the team a clear basis for choosing what to test.
ASSUMPTIONS CHECKED AGAINST REAL COMPARABLES
Planned features came with an expected lift, and the plan reran as they changed. I checked the assumptions behind it against real market comparables, so the team tested the claims that mattered most first.
Questions