Category
220 articles in this category · page 2 of 19
A two-group in-person focus group runs $8,000-$12,000 and online $4,000-$7,000. But the line item that decides your budget is the show rate: at 80% you must recruit 17 to seat 10.
Koji Team · August 19, 2026 · 12 min read
Every post-purchase record requires a customer willing to accuse someone. Regret the customer blames themselves for produces no return, no dispute, no review and no ticket. How to measure the silent keep.
Koji Team · August 18, 2026 · 13 min read
A return rate falls when listings get accurate and when returning gets annoying. Same dashboard, opposite outcomes. The suppression test, the pairing that makes returns reduction falsifiable, and how to research it.
Koji Team · August 18, 2026 · 13 min read
Return reason codes measure your refund policy, not the cause of the return. Here is why the data is biased, what it cannot distinguish, and how to interview returners about the purchase instead.
Koji Team · August 18, 2026 · 13 min read
In a replenishment subscription, one decision produces an unbounded number of purchases - and every purchase after the first has no decider, no moment, and no reason attached to it. Your research instruments have nobody to interview. Here is the one that still works.
Koji Team · August 17, 2026 · 12 min read
A design change that raises conversion by making the decision easier to make also makes it easier to make badly. The best evidence on nudges at scale says the honest effect is 1.4 points, not 8.7 - and the downstream cost lands in a different quarter, on a different team's dashboard.
Koji Team · August 17, 2026 · 12 min read
A product detail page test measures whether a page closed a sale, not what the shopper came to find out. Only 48% of leading desktop sites score decent or better on PDP UX - and 19.3% of online orders come back. Here is how to research the listing properly.
Koji Team · August 17, 2026 · 12 min read
An AI answer has no publish date, no last-updated, no version and no correction channel. Three separate clocks decide what it says about you, you control none of them, and the cost is paid by the buyer who believed it.
Koji Team · August 16, 2026 · 12 min read
Visibility and description accuracy are independent variables. If the model has settled on the wrong description of your product, every extra mention distributes it further. Here is how to find out what it is saying.
Koji Team · August 16, 2026 · 11 min read
AI visibility trackers sample a prompt list you invented. Here is what they can legitimately prove, the metric that systematically lies, and how to measure the part that decides deals.
Koji Team · August 16, 2026 · 11 min read
Digital shelf analytics measures a page, not a shopper. Here is exactly what share of search can prove, the three structural gaps a scrape cannot close, and how to pair it with shopper research that recovers the missing why.
Koji Team · August 15, 2026 · 11 min read
The vast majority of marketplace sales go through one box that an algorithm fills. That makes your sales data a record of ranking decisions, not customer preference - and the two can move in opposite directions. Here is how to tell them apart.
Koji Team · August 15, 2026 · 11 min read