How to measure a branded game
Decide the one action first. Then count who was invited, who started, who finished, how long they played, and how many did the action. Report those five numbers and the ratios between them on one page.
Real gameMost branded games end as a screenshot and a guess. The game looked good, people played, and nobody can say what changed. Knowing how to measure a branded game fixes that, and it is easier than most teams think.
Decide the one action first
Before the game is designed, write one sentence: we want people to ___. Sign up. Try the product. Visit the store. Come back next week. Everything you count should serve that sentence.
If the sentence has three actions in it, you have three games or no clear goal. Pick one. Everything else is a supporting number.
Five numbers that measure a branded game
- Invited. How many people had a real chance to play: messages sent, links opened, or people at the event.
- Started. How many distinct people began. Count people, not plays.
- Finished. How many reached the end, or the level you defined as the end.
- Time played. How long they stayed. Look at the median, because one obsessive player distorts the average.
- The action. How many did the thing you wrote in that first sentence.
For a game people return to, add one more number: came back, meaning how many played again after a set gap such as seven days.
Gamification metrics that diagnose the problem
The ratios tell you more than the totals. Each one points at a different fix.
| Ratio | If it is low, look at |
|---|---|
| Started ÷ Invited | The invitation. The hook, the channel, the prize, or too many steps before play |
| Finished ÷ Started | The game. Too long, too hard, or slow to load on real phones |
| Action ÷ Finished | The link between game and goal. The action is hidden, unclear or unrelated to the fun |
| Came back ÷ Finished | The reason to return. No new content, no rhythm, no stakes |
There is no universal good score, so we will not give you one. Your first run is your baseline. Set a target for the action before launch, and judge the second run against the first.
AI visualisationGame analytics for brands: what changes by type
The five numbers stay the same. What counts as each one changes.
Digital campaign games. Invited comes from your media and message data, so tag every link by channel. That lets you compare channels by action rather than by clicks. Decide early how a game on a brand site connects to your CRM or sales data, if the action happens outside the game. See digital games.
Mixed digital and physical games. People play on a phone or a screen in a real place, so Invited is harder to count. Use a figure you can defend, such as registered attendees or scans of the join code, and say which one you used. Count by session and by hour, because a queue can hide a good game behind a slow one.
Player data needs extra care
A game can see who played and how they scored, so the rules have to be clear before launch. Game analytics record plays, completions and scores. What is collected, and who can see it, is agreed in the brief.
- Is a leaderboard opt-in, and can it show nicknames rather than full names?
- What does a player agree to when they sign up, and is it written in plain words?
- How long is the data kept, and who deletes it?
- Does the prize draw or reward follow the rules in each country where people play? Saudi Arabia and Egypt each have their own.
- Does this meet the personal data law in each market? Have your legal team review the plan.
People play honestly when they trust what happens to their data.
How to report it on one page
Marketing teams do not need more data. They need a result they can show.
| Line | What goes in it |
|---|---|
| The action | The sentence you wrote first, and the target |
| Five numbers | Invited, started, finished, time played, the action |
| Ratios | The three or four that matter, against your baseline |
| What changed | Anything altered during the run, and what happened next |
| One recommendation | The single change you would make next time |
If a number on the page does not change a decision, take it off. Downloads, total plays and impressions look large and prove little without the ratios behind them.
Build measurement in, do not bolt it on
Measurement costs least when it is part of the design. Agree the action and the events to record in the brief, then build them with the game: invitation opened, game started, game finished, action done. Test them before launch, because a run with broken tracking cannot be repeated.
Adding it later is harder, and sometimes impossible. If the game has no way of knowing who was invited, no later report can tell you the start rate.
Reporting is planned from the start of every game we make. A live dashboard is a separate, optional element, useful when someone will act on the numbers during the run. For a short campaign, a report at the end may be all you need. Read how this fits our work on measurement in games.
As a guide to scope, a bilingual web mini-game with analytics plans at SAR 12,500 to 23,000 (USD 3,300 to 6,100) Calculator estimate, before VAT. Try your own mix in the budget calculator.
How to measure a branded game
What is a good completion rate for a branded game?
There is no number that holds across games, audiences and lengths. Use your first run as the baseline, and compare similar games only.
How do we measure a game that runs at an event?
Use a defensible count of who could play, such as registered attendees or scans of the join code. Then count starts, finishes and the action as usual.
Do we need a live dashboard?
Only if someone will change something during the run because of it. Otherwise a one-page report after the run does the job.
Can a game show a return on investment?
It can if the action connects to something with a value, such as a sign-up, a trial or a completed course. Agree that link and how it will be tracked before you build.
