Insights
Written when we have something worth your time.
Patterns from the work, anonymized and compressed. A post has to be new, applicable, and likely to change what you do, or we skip the week.
Experimentation
Why Most Startup Experiments Fail (and What to Do Instead)
Ten "winning" tests and flat company growth usually means the experiments were framed around outcomes, not behavior. The fix is in how a test starts, not the math.
Retention & Lifecycle
Retention in Marketplaces: Why Liquidity Beats Variety Alone
Liquidity, not variety, is what brings marketplace buyers back. That changes what you measure, and what you stock.
Data Foundations
Why creative tagging fails, and what a working ontology looks like
Two video ads running right now for the same brand look like different teams made them, and a typical creative taxonomy scores them as literally identical. That's the normal state of creative tagging, and it's fixable at ~2 minutes per asset.
Experimentation
How long to run an ecom A/B test
Run length is set by 2 numbers you already know before you build the test: weekly orders and the smallest lift worth acting on. One formula, a table to find your row, and what to do when your row says a year.
Measurement & Attribution
Why your year-over-year sales look terrible the week before a holiday
Same-weekday comparisons fix weekdays, not holidays. Labor Day moved 6 days this year, enough to make a brand pulling ahead read -18% on the Friday before its biggest weekend.
Measurement & Attribution
Why media affiliates look like your best channel, and the 30-minute test that cut one brand's commission by ~$500k
Media affiliates read $8.3 per $1 of affiliate ROAS, beaten only by brand search. A 30-minute test showed a third of it was discount-code pages, worth ~$500k a year.
Data Foundations
How to map ZIP codes to DMA, metro area and region: the one table to join to your orders
Every order already has a ZIP code. Join it to one master table and the same orders can be cut by TV market, metro, region, urban/rural and income, which is where most geo decisions actually get made.
Data Foundations
DMA vs CBSA vs MSA: which geography to use for ecom analysis
DMAs are TV markets, CBSAs are commuting areas and an MSA is the larger kind of CBSA. Each answers a different question, and for store decisions none of them is the right map.
Get in touch.
Tell us what you are trying to figure out. If we can help, we will say how. If we are not the right fit, we will say that too, and point you somewhere better.
01
Free work first
We start by doing a real piece of work on your data, free, so the first thing you judge is output, not a pitch.
02
First build, fixed scope
One piece of work at a fixed price, scoped to a few weeks: a metric layer, a tracking plan, one agent or one pipeline, whichever gap you came with. Small enough to be safe, real enough to matter.
03
Scale when it earns it
Month to month from there, from a few thousand a month. You own every model, dashboard and definition.