How to find your key action

Your key action is the one thing a user does that means they got what they came for. Sent the message. Published the post. Closed the ticket.

Meterio asks you to pick exactly one, and then measures everything about new users against it: how many reach it, how long they take, and whether reaching it makes them stay. Pick well and the dashboard tells you where your product is losing people. Pick badly and it measures nothing.

What makes an action a good candidate

Three tests. An action has to pass all three.

It delivers value to the user, not to you. A signup is your win. A completed invoice is theirs. Actions that only matter to your funnel will happen for everyone and separate nobody.

It happens in the first session or two. The point is to predict retention early enough to act. An action people take in week three is a result of them sticking around, not a cause of it.

It repeats. A key action people do once is a milestone. One they do weekly is a habit, and habits are what you're trying to build.

Verbs, not places. "Viewed pricing" is a location. "Created report" is a thing someone did.

How to pick between two or three candidates

Write down your shortlist first. Most founders have two or three obvious contenders and pick on instinct. You don't have to.

Set one as your key action and let a few weeks of data collect. Meterio then shows you the number that settles it: the retention gap between users who took the action and users who didn't.

If people who did it come back at 55% and people who didn't come back at 20%, that action predicts retention and it belongs. If both groups return at roughly the same rate, the action is something users do on the way past, not the reason they stay. Try the next candidate.

That comparison needs at least ten users on each side before it means anything, so Meterio hides it until you have them. Below that, the split is noise.

Setting the threshold

Some actions only count once you've done them a few times.

Meterio lets you say the key action must happen N times before a user counts as activated. One is the default and the right choice for most products. Raise it when a single instance is ambiguous: sending one message might be someone testing the box, while sending five means they're using it.

You can also filter on properties, so "published a post" can be narrowed to posts that were public, or over a certain length. Keep this simple. Every condition you add makes the metric harder to explain to yourself six weeks later.

What you get once it's set

Four things, all measured against a rolling cohort of recent signups rather than the 7/30/90 toggle.

Conversion rate. The share of new users who reached the key action within seven days of first showing up. This is the number Meterio checks to decide whether your product has cleared the stickiness stage, and the bar depends on your category: 30% for SaaS, 45% for an AI tool, 20% for a marketplace, 25% for a community, 40% for content.

Median time to activate. How long the typical activated user took, in hours. A long median usually means the path is unclear rather than hard. This one needs at least five activated users before it shows.

The retention gap. Whether activated users come back more than everyone else, at day 7 and day 30.

Drop-offs. The users who arrived and never reached it. This is your list of what to fix.

New users only appear in these numbers once enough time has passed to judge them fairly. Somebody who signed up yesterday isn't counted as a failure to activate within seven days, because their seven days haven't happened.

Getting it wrong is fine

You are allowed to change your key action. Most products change it at least once, usually after the first look at the retention gap tells them the action they picked doesn't separate anyone.

Changing it recalculates the history, so you don't lose anything. What you lose is the weeks you spent measuring the wrong thing, which is exactly why it's worth checking the gap rather than trusting the first instinct.

Once it's set and the gap is real, the work is obvious: get more new users to that action, faster. That's the single most reliable way to lift repeat rate and cut churn.