Find the one fix that gets your first hundred past week one

You have a small cohort of signups, no distribution, and a founder-level question: is this product worth scaling? North American ecommerce faces an activation crisis, and onboarding may cause a 98% retention collapse. The answer usually hides in early user retention, not in the size of the list. The first growth question is whether a new user comes back after the first week.

You can feel the difference between a signup that is curious and one that is stuck. The curious one asks a question. The stuck one goes quiet. The product is a conversation you are running by hand. If you wait for scale before you learn what breaks, you will make the problem bigger.

Two rates tell you if the first hundred will return

One-week retention and weekly retention are the two rates to use. The first rate catches a broken first task. The second catches a product that cannot hold attention after the first week. A signup that never returns has not earned another look.

Skip paid traffic for now. Hold off on lifecycle campaigns. Avoid tests that need a large sample to mean anything. Those tools assume you already know what the user wants. You are still learning that.

Run the check before you buy more traffic

• Pull the early rates. Export the one-week retention rate and the weekly retention rate for the newest cohort. Done looks like a pair of numbers on a page, not a dashboard. If the numbers are missing, add the minimum event that proves a user came back. A return is a completed action that matches the reason they signed up, not a login.
• Compare the drop to a relevant benchmark. Use a benchmark from a similar product stage and user type, not a global average. You are done when you can state the delta plainly: your week-1 rate is above or below the benchmark by a clear amount. A benchmark should be close enough to force action.
• Find the single fix. Interview the users who left and the users who stayed. Ask what they expected before they clicked, what they tried after they clicked, and what would make them come back. The output is a single change you can ship soon, not a roadmap. If onboarding is the main task, treat it as the likely failure point.
• Ship it to the next batch. Put the fix in front of the next signups before you widen the funnel. Finish when you can see a before-and-after on the same two rates. If the rates do not move, the fix was not the real constraint. If they move, you have evidence that the product can be worth more attention.

The benchmark delta is the test

A benchmark is only useful when it is close enough to your product to be uncomfortable. In AI products, one region hits 619 actions per user while others sit near 100. North American AI engagement fell 38% year over year. A strong average can hide a weak segment. If your users are in the weak segment, the average will not save you.

Choose the benchmark that matches the job, the buyer, and the setup burden. A team tool is not the same as a personal tool. A product that needs a setup call is not the same as a product that works after a quick click. The closer the comparison, the more the delta means.

Write the yardstick before you scale

Keep the First Hundred Retention Yardstick to a single page. Write down the week-1 rate, the weekly rate, the benchmark delta, and the single fix to ship this month. If the week-1 rate is below the benchmark, the fix is usually in the first task, the first value moment, or the reason to return. If it is above the benchmark, the next question is whether the weekly rate can hold as you add more users.