The Activation Metric Problem

The gap is the story

Per Mixpanel’s 2026 State of Digital Analytics, 58% of B2B SaaS companies now run a product-led motion. Only 34% of them track activation. That gap is the entire PLG story compressed into one statistic.

A product-led company that doesn’t measure activation is running a growth engine with the most important gauge missing. It knows how many people walked in the door. It does not know how many people got to the thing they came for. Everything downstream — retention, expansion, the self-serve upgrade path that makes PLG economically work — depends on activation, and most of the market is flying without it.

The fix is not another dashboard. It’s a decision about what the company is actually measuring, and the willingness to instrument the one moment that predicts everything else.

Activation predicts the P&L; signups predict nothing

Activation is the first moment a new user reaches the value the product promised. Not the moment they sign up. Not the moment they poke around. The moment the product delivers on its reason to exist for that specific user.

It is the most predictive metric in a product-led business because it sits upstream of everything that matters. A user who activates retains. A user who retains expands. A user who expands becomes the self-serve revenue that lets PLG undercut sales-led competitors on cost of acquisition. Miss activation and none of that fires. The user churns quietly, and the company never learns why, because the company was watching the signup counter instead.

Signups predict almost nothing. A signup is an intent signal at best and a bot at worst. Two products with identical signup curves can have wildly different revenue outcomes, and the difference is entirely in what happens between signup and value. Activation is that difference, made measurable.

This is the part teams intuitively know and structurally ignore. Ask any PLG product leader whether activation matters and they’ll say yes. Ask them their activation rate and most can’t answer. The 34% number is not an information gap. It’s an instrumentation gap dressed up as a knowledge one.

Teams track signups because signups are easy

Signups are trivial to measure. The event fires itself. Every analytics tool captures it by default. It shows up on day one with zero engineering work, and it goes up and to the right in a way that feels like progress.

Activation is hard to measure because it requires the company to make a claim. You cannot instrument activation until you’ve decided what it is, and deciding what it is forces a fight about what the product is actually for. That fight is uncomfortable. It surfaces disagreement between founders, between product and sales, between the story the deck tells and the behavior the data shows. So teams avoid it, and default to the metric that requires no argument.

There’s also a vanity problem. Signups are the number you put in the board deck and the fundraising email. Activation is the number that tells you how many of those signups were real. One flatters, the other corrects. Under pressure to show momentum, teams reach for the flattering one and call it a north star. It isn’t. It’s a decoy.

The tools reinforce the habit. Most analytics platforms ship with signup and session tracking configured out of the box and activation configured by nobody. The path of least resistance is the wrong metric, so the wrong metric wins.

Defining activation is the actual work

Activation is not a universal event. It’s specific to the product, and defining it correctly is most of the job.

The test for a good activation definition is simple: it’s the earliest action that reliably predicts long-term retention. Earliest, because you want the signal fast enough to act on. Reliably predicts retention, because a definition that doesn’t correlate with users sticking around is a definition of nothing.

The way to find it is to look backward from your retained cohort. Take the users who were still active at 90 days. Trace what they did in their first session, their first week. Find the action that separates them from the users who churned. That action, or that small cluster of actions, is your activation event. It’s the “aha” moment made into a query.

At MoneyGeek, the version of this question is always the same: what is the moment a user goes from browsing to trusting the answer we gave them? That moment is measurable, and it looks nothing like a signup. The specific event differs by product. The method of finding it does not.

Two failure modes to avoid. Defining activation too late — “made a purchase,” “invited three teammates” — gives you a clean correlation with retention but no room to intervene, because by the time the user hits it, they’ve already decided. Defining it too early — “completed onboarding,” “viewed the dashboard” — gives you a number that moves without predicting anything. The right definition sits at the earliest point where the signal is still trustworthy.

The instrumentation checklist

Once activation is defined, the instrumentation is mechanical. Four things have to be true, or the metric lies to you.

The event has to fire server-side or with client-side validation, so ad blockers and dropped sessions don’t corrupt the count. The event has to carry the user identity, so activation can be tied to a cohort and followed forward into retention and revenue. The event has to be defined once, in one place, and referenced everywhere, so marketing’s activation and product’s activation are the same number. And the activation rate has to be reported as a cohort curve, not a lifetime average, because a lifetime average hides exactly the trend you’re trying to see.

Get those four right and activation stops being a vanity slide and starts being a control surface. You can run an experiment against it. You can attribute a drop to a release. You can tell the difference between an acquisition problem and an activation problem, which are the two things signup-obsessed teams constantly confuse and mistreat.

The instrumentation is the easy part. Teams skip it not because it’s technically hard but because the definition work that precedes it never got done.

Where the teams that measure this end up

A company that measures activation correctly runs a different playbook. It stops optimizing the top of the funnel it can see and starts optimizing the middle it couldn’t. It moves acquisition spend toward the channels that produce activated users, not the ones that produce cheap signups, and those are rarely the same channels. It builds onboarding against a metric instead of against opinion.

Most of all, it gets a self-learning loop. Activation defined, instrumented, and tied to retention becomes a signal generator: every cohort teaches the next one, every release moves a number that means something, and the product compounds because the company can finally see the moment where value happens and engineer more of it.

The 58/34 gap is not a reporting oversight. It’s the line between companies running PLG and companies with a PLG story. The ones that close it are measuring the only moment that predicts the rest of the business. The ones that don’t are counting people at the door and calling it growth.