I built the test library I needed and could not find.
GrowthLayer was built by Atticus Li, a growth and experimentation operator. He built the product to keep the record, evidence limits, workflow status, and program reporting in one reviewable place.
The problem I kept hitting
After running hundreds of experiments across five brands at NRG Energy, and across the Startup Banking division at Silicon Valley Bank, I kept hitting the same four walls. None of them were about running tests. They were about everything around the tests.
No institutional memory
Experiment results can remain scattered across chat threads, slides, spreadsheets, and testing platforms, making prior context difficult to review.
Learning never compounded
Without consistent fields and evidence limits, teams have to reconstruct what a prior result meant before they can use it in a new decision.
Reporting was manual every time
Quarterly program reporting often requires manual aggregation when outcomes, learnings, and impact estimates live in separate files.
Statistical rigor was inconsistent
Arbitrary timelines and winner labels can obscure power, sample-ratio mismatch, effect size, and the limits of the available evidence.
I looked for a tool. It did not exist.
The platforms I used were great at running tests and useless at remembering them. Enterprise tools focus on execution, not institutional memory. Generic wikis and spreadsheets had the memory but none of the statistical rigor or revenue reporting.
Nothing sat on top of whatever testing tool you already use and kept the result, evidence limits, workflow status, and program reporting together. So I built it.
What GrowthLayer is
The institutional-memory layer for any team running experiments — CRO, growth, product, or agency.
Analyze with real rigor
Paste in your test data and get significance, confidence intervals, and sample-ratio-mismatch checks in seconds.
Remember every result
Save each test with consistent fields for its hypothesis, outcome, status, and learning so the history can be reviewed later.
Compound your learning
Review structured outcomes, coverage, and quarterly program summaries instead of reconstructing history from scattered files.
Report in revenue
Review quarterly summaries of recorded outcomes, supplied revenue impact, and key learnings.
What I believe
Growth should not run on opinions. The teams that win are the ones that test customer behavior across UX, messaging, offers, and funnels, and remember what they learn. My whole career has been about turning vague growth opinions into tested revenue decisions. GrowthLayer is that belief, turned into a product.