Points issued and badges earned look great on a slide and tell you almost nothing.
Here are the metrics that show whether your gamification is really working, and the ones quietly fooling you.
BRIEFING
Turning real money into a game currency is one of the oldest tricks in commerce, and one of the most effective, because spending money quietly hurts and abstraction numbs the hurt. This piece breaks down the psychology with the research behind it, names the three places a currency tips into a dark pattern, and stays honest about the real reasons both players and studios prefer virtual currency. It ends with how to design one that respects the person spending it.
The dashboard is a wall of green. Badges awarded this month, up. Points issued, up forty percent. Engagement, whatever that means this quarter, up and to the right. Everyone in the room is pleased, until the quiet person at the end asks whether retention or revenue actually moved, and the honest answer is no. This is the trap gamification sets, and almost everyone walks into it at least once.
The good news is that measuring gamification well is not hard once you know which numbers lie to you and which ones do not. That is the whole subject of this piece.
The trap gamification sets for itself
Most product features are hard to measure. Gamification has the opposite problem: it is far too easy to measure, and the easy numbers are the wrong ones. The moment you launch a points system, points-issued goes up, because you are issuing points. Badges-earned climbs, because you are handing out badges. Logins tick up for a week on novelty alone. None of it tells you whether the product became more valuable to anyone. "Engagement" is especially slippery, because as I wrote in how your interface quietly decides whether people stay, the real thing lives in the moment-to-moment feel of the product, which a points counter cannot see.
Eric Ries drew the line precisely in The Lean Startup: a vanity metric makes you look good but does not inform a decision, while an actionable metric shows clear cause and effect. Page views, sign-ups, downloads, and yes, points issued, are the classic vanity metrics. They are the numbers gamification is best at inflating and worst at justifying.
Measure the behaviour you were trying to change
The fix starts before you build anything: name the behaviour you actually want to move. Not "increase engagement," which is vague enough to be satisfied by any twitch of a vanity metric, but something specific: new users reaching their first real success, the weekly return rate, a feature getting adopted, a workflow getting finished. The badge is the intervention. The behaviour is the result. If you cannot say in one sentence which behaviour this gamification is meant to change and how you will see it in the data, you are not ready to build it yet. That sentence is your whole measurement plan in miniature.
The gamification metrics that actually matter
Once you know the target behaviour, a short list of honest metrics carries almost all of the weight.
- Retention. The share of users who come back, tracked by cohort at day 1, day 7 and day 30, and whether the curve flattens into a plateau rather than falling to zero.
- Activation. Of the people who arrive, how many reach the first genuine moment of value. Gamification aimed at onboarding lives or dies here.
- Stickiness (DAU over MAU). What share of your monthly users show up on a given day, a proxy for habit. Sensible targets swing widely by product, so the benchmark that matters is your own trend: climbing, flat, or sliding.
- The business outcome. Revenue, conversion, churn, support load, whatever the product exists to do. If the gamification cannot eventually connect to one of these, it is decoration.
Retention deserves the closest look, because it is the outcome gamification most often promises to move, and the one you most control. A flattening retention curve is the classic signal that people found lasting value rather than a novelty. As I noted writing about creating urgency without frustrating your users, Nicholas Lovell offers a rough rule of thumb for healthy retention, on the order of 40 percent the next day, 20 percent at a week and 10 percent at a month. Treat those as a compass, not a target, and always read your own trend against your own history.
A frame to hang them on: pirate metrics
If that list feels loose, there is a clean frame for it. Around 2007, investor Dave McClure introduced AARRR, the "pirate metrics": Acquisition, Activation, Retention, Referral and Revenue, five stages of a user's life he proposed specifically to drag teams off vanity numbers and onto behaviour that predicts real growth.
Most gamification is really an attempt to improve two of those letters, Activation and Retention, so that is where to measure it. If your points system was meant to help people stick, the honest scoreboard is the retention curve, not the points counter.
Cohorts and controls: how you actually know it worked
Here is the step that separates a real result from a hopeful story. A number going up after you launch does not mean your launch caused it: marketing, seasonality and a dozen other things move the same numbers.
The way to know is a control. Hold a random slice of users out of the gamification, and weeks later compare their retention and target behaviour to the group that got it. The gap, if any, is the actual effect, cleanly attributable to the thing you shipped. Without that comparison you are left with "it went up and we were involved," which is not evidence, it is a coincidence you are hoping to take credit for.
Beware the metric you are gaming
There is a law waiting for anyone who picks the wrong target. Goodhart's law says that when a measure becomes a target, it stops being a good measure.
Reward logins and you will get logins, empty ones, from people opening the app to protect a streak and closing it again. Reward points and users will farm points in whatever way is easiest, which is rarely the way that creates value.
The metrics worth chasing are the ones that are hard to game precisely because they are tied to something real: a person coming back because the product is genuinely useful cannot be faked the way a badge count can.
Pick one number that means value
Finally, give the team a single number to steer by that captures real value delivered, what the growth world calls a North Star metric.
Not points, not badges, but the one measurable thing that means a user got what they came for: nights booked, workouts completed, projects shipped, lessons actually learned.
When the North Star is a genuine value metric, gamification gets aimed at making the product better.
When it is a vanity metric, gamification gets aimed at making the number bigger, and those are not the same thing. Anyone can feel the difference in a product that was built for the second.
TIP
Run a holdout. Before rolling gamification out to everyone, keep a random slice of users out of it. Weeks later, compare their retention and target behaviour to the gamified group. That difference is the real effect, and it is the number to put in front of whoever controls the budget. Everything else is a dashboard telling you what you hoped to hear.
WATCH OUT
Goodhart's law catches nearly everyone. The instant a metric becomes the goal, people optimise the metric instead of the thing it was standing in for.
Treat any number your gamification directly inflates, points issued, badges earned, logins, as an input to watch, never as proof of success. The proof is always downstream, in retention, activation and revenue that a mechanic cannot manufacture on its own.
One question to ask before you celebrate
Before you call a gamification launch a win, ask one thing: if I had held a group of users out of this, would their retention and behaviour genuinely be worse?
If you can answer yes with data, you have proof. If the only thing you can point to is points issued and badges earned, you have a number that was always going to go up, and it has told you nothing you did not already know the day you shipped it.
KEY TAKEAWAY
The easiest gamification numbers to move are the least meaningful. Judge the work by whether it shifted the behaviour you cared about, retention, activation, a real business outcome, ideally measured against a control group.
If you cannot tie it to value, you have not built a better product, only a prettier dashboard.
Glossary
Vanity metric. a number that makes you look good but does not inform a decision (page views, sign-ups, points issued). Easy to grow, hard to act on.
Actionable metric. a metric with clear cause and effect that helps you decide what to do next (Eric Ries, The Lean Startup).
Retention. the share of users who return over time, tracked by cohort at day 1, 7 and 30. The metric gamification most often claims to improve.
Cohort analysis. grouping users by when they started and following each group over time, so you compare like with like instead of a blurred average.
DAU/MAU (stickiness). daily active users divided by monthly active users: a proxy for how habitual usage is. Judge it against your own trend.
Activation. the point at which a new user first reaches real value. Where onboarding-focused gamification succeeds or fails.
AARRR (pirate metrics). Dave McClure's funnel, Acquisition, Activation, Retention, Referral, Revenue, built to replace vanity metrics with behaviour that predicts growth.
Churn. the rate at which users stop using the product. The mirror image of retention.
North Star metric. the single measure that best captures the real value a product delivers, used to keep a team optimising value instead of vanity.
Goodhart's law. when a measure becomes a target, it stops being a good measure, because people start optimising the number itself.
Holdout / control group. a random set of users deliberately kept out of a change, so its true effect can be measured by comparison.
Not sure whether your gamification is working, or just looks like it?
I help teams define the behaviour that matters, instrument it honestly, and measure real retention and activation against a control, so you can prove the ROI or cut what is not earning its place.
Numbers you can act on, not just admire.