all notes
2026-08-02Naman Barkiya

What to Measure in the First 90 Days After Your MVP Ships.

Track three numbers in the first 90 days after an MVP ships: activation (do new signups reach the core action), retention (do they return at day 1, day 7, day 30), and one real monetization signal (a checkout started, a pre-order paid, an unprompted pricing question). Traffic, social shares, and raw signup counts are noise until those three hold up.

Signups are the easiest number to get and the least useful one. Here are the three metrics that actually predict whether an MVP survives its first 90 days, and the table that shows when to check each one.

Track three numbers in the first 90 days after an MVP ships: activation (do new signups reach the core action, not just create an account), retention (do they come back on day 1, day 7, day 30), and one real monetization signal (a checkout started, a pricing question asked, a pre-order paid). Everything else — traffic, social shares, feature requests — is noise until those three hold up.

Most founders open a dashboard the week after launch and watch signups. Signups are the easiest number to get and the least useful one — they measure curiosity, not whether the product works. We've shipped 15 products across gaming, travel, healthcare, and AI, and the pattern repeats: the founders who obsess over top-of-funnel traffic in month one are usually avoiding the harder question of whether anyone who arrived actually got value.

What should you actually track in the first 90 days?

Three signals, in this order of weight:

  1. Activation — the percentage of signups who complete the one workflow the MVP exists for.
  2. Retention — whether the users who activate come back, measured at day 1, day 7, and day 30.
  3. A real monetization signal — not revenue yet, necessarily, but an action that costs the user something: a checkout started, a card entered, a pre-order paid, a specific pricing question asked in a support thread.

This is a tight version of the AARRR framework Dave McClure popularized at 500 Startups — acquisition, activation, retention, referral, revenue — collapsed to the two that actually predict survival at 90 days. Referral and top-line revenue are lagging indicators; at this stage they mostly measure how good activation and retention already are.

What is activation, and why do founders usually measure the wrong thing?

Activation is not "created an account." It is the first moment the product's value became real to the user — the workflow it was built to deliver, completed once, unassisted. For a marketplace MVP, that might be one completed transaction. For a SaaS tool, it might be the first report generated from real data, not a demo dataset.

The mistake we see most often: founders define activation as something the product does easily (a login, a tour completed) rather than something the user chose to do that proves the core loop works. A high login rate with a low activation rate means the onboarding funnel is fine and the value proposition, or the workflow itself, is not landing. Define activation as one specific, named action before you launch — not after you're staring at a dashboard trying to explain a number that already happened.

How do you read a retention curve without lying to yourself?

Plot the share of activated users still active at day 1, day 7, and day 30. A curve that keeps dropping toward zero means the product is a leaky bucket — no acquisition strategy fixes that. A curve that flattens, even at a modest number like 15-20%, means a real core of users found something worth returning for. That flattening point, more than any vanity metric, is the signal that a product has product-market fit worth funding further iteration on.

Read it in cohorts, not in aggregate. A single retention number blended across three months of changing onboarding hides whether last week's fix actually worked. Compare the cohort that signed up in week 1 against the cohort in week 8 — if week 8's curve sits above week 1's, the product is getting better, which is the only claim worth making to an investor or to yourself.

What counts as a monetization signal before you have real revenue?

Willingness to pay, demonstrated, not stated. A user who says "I'd pay for this" in an interview is being polite. A user who enters a card number, requests an invoice, or asks "what does the paid tier cost" unprompted is showing you something real. On one of our client builds, LaunchProd, the signal that mattered most in the first quarter wasn't signups — it was the rate at which trial users asked about usage-based pricing before the sales team brought it up.

MetricWhat it measuresRead it atIgnore if
Activation rate% of signups completing the core workflow onceWeekly, per cohortDefined as login or onboarding-tour completion
D1 / D7 / D30 retention% of activated users still active at each markWeekly, per cohortBlended across changing onboarding versions
Monetization signalCheckout started, invoice requested, unprompted pricing questionAs it happensVerbal "I'd pay for this" with no action behind it
Traffic / signupsTop-of-funnel volumeMonthly, for context onlyTreated as the headline number
Feature requestsWhat users say they want nextOngoing, qualitativeActed on before activation is solved

What should you ignore for the first 90 days?

Social shares, press mentions, and the raw signup count. All three feel like progress and none of them predict whether the product survives contact with real usage. We tell founders the same thing we told ourselves the week LaunchProd shipped: a screenshot of a growth chart with no retention behind it is a vanity metric with good production values.

Feature requests deserve a specific caveat. They arrive early and often, and they are tempting to act on because they feel like customer validation. Most of them are noise until activation and retention are solid — a feature request from a user who never activated is a request from someone who hasn't experienced the product's core value yet, and building for them first usually adds scope the MVP doesn't need.

The one call this decides

If activation and retention hold up at day 30, the next conversation is about a post-launch retainer to fund the iteration that turns early traction into a fundable metric. If they don't, the next conversation is the one we'd rather have before a build starts: go back to validating demand, because no amount of post-launch polish fixes a workflow nobody wanted in the first place. Ninety days of the right three numbers tells you which conversation you're actually in.


Heuristics


Written 2026-08-02 by Naman Barkiya.

FAQ

Questions this usually surfaces.

What metrics should I track after launching my MVP?
Three, in order of weight: activation (the percentage of signups who complete the one core workflow the MVP was built for), retention (the share of activated users still active at day 1, day 7, and day 30), and a real monetization signal (a checkout started, an invoice requested, or an unprompted pricing question). Traffic and raw signup counts are context, not the headline number.
What is a good retention rate for a new MVP?
There is no universal number — what matters is whether the retention curve flattens instead of dropping toward zero. A curve that levels off even at a modest 15-20% at day 30 means a real core of users found lasting value. Read it in weekly cohorts, not one blended average, so you can tell whether a recent onboarding fix actually worked.
How do I know if my MVP has product-market fit in the first 90 days?
Look for a retention curve that flattens rather than decaying to zero, paired with an unprompted monetization signal — a user asking about pricing, entering a card number, or requesting an invoice before your sales process asks them to. Verbal interest ('I'd pay for this') is not the same signal; it costs the user nothing to say.