The audience dashboard
Audience shows who visits and plays your games, with no setup on your side. The same view is on each game’s Stats page. An assistant reads it through get_audience and get_retention.
Pick a period (today, 24 hours, 7, 30 or 90 days, 12 months, or all time) and a game, or look at all of them. Each figure is compared with the period before. Tables export to CSV. A day is a day in your timezone.
The headline numbers
Section titled “The headline numbers”| Number | What it counts |
|---|---|
| Unique visitors | People who opened a game page |
| Plays | Sittings started, on Alldle or on your site |
| Players | Unique people who played |
| Page views | Every time a game page was opened |
| Play rate | Share of visitors who pressed Play |
| Time played | Visible time after pressing Play, per player, per day |
| Returning | Share of visitors who had been before |
Plays are also split into those inside Alldle and those sent to your site, which is how you see what a listing is worth.
How people are counted
Section titled “How people are counted”- Everyone counts, signed in or not. A browser gets a random id, kept for at most 13 months. There are no cookies, and IP addresses are never stored.
- A browser that opts out, with Global Privacy Control or from Alldle’s privacy page, keeps no id. It is counted through a daily-changing hash instead.
- A play is counted once per visitor, game and kind of play per 30 minutes, so reloading the page does not inflate it. Pressing Play and then opening your site counts as two.
- Signed-out plays are capped at 25 per game per day from one address, so a script cannot push a game up the rankings.
- Bots are dropped by their user agent.
- Your own visits are left out while you are signed in. Testing your page signed out still counts as a visitor.
Where they come from
Section titled “Where they come from”Sources are split into On Alldle and Elsewhere.
| Source | What it means |
|---|---|
| Alldle home page | The home page |
| Categories | A category page |
| “Games like” pages | A page suggesting similar games |
| Search on Alldle | Alldle’s search |
| Players’ daily ritual | Their ritual list |
| Daily reminders | A reminder notification |
| Top games, New games | Those lists |
| Discovery queue | The queue of suggestions |
| Another game’s page | The suggestions on another game |
| Google, Bing and others | Search engines |
| Other sites | The site that linked to you, or your utm_source or ref |
| Direct / none | No source known |
To track a campaign of your own, add ?utm_source=your-name to your links to the Alldle page.
Who they are
Section titled “Who they are”Countries and cities (on a globe you can turn), devices, browsers, operating systems and languages. Plus a weekday-by-hour heatmap of when people show up, which tells you when a new puzzle matters most.
Live shows how many people are on your games right now, meaning the last 5 minutes, with the last 30 minutes by country and by game. It shows totals only.
In daily rituals is how many signed-in players keep your game in their list, and how many added it this period. Signed-out rituals stay in the browser, so they are not in this number.
Habit shows how many different days each player played in the period.
Do players come back?
Section titled “Do players come back?”This is the panel to watch on a daily game. It is measured on plays, signed in or not.
| Number | What it asks |
|---|---|
| Next day | Of those who played, how many played again the next day? |
| Within a week | How many came back within seven days? |
| After a month | How many were still playing a month later? |
| Stickiness | Daily players as a share of monthly players |
Below these, cohorts: each row is the players whose first play fell in that week, and each column the share of them who played again N weeks later. A row with very few players swings wildly, so read the size next to the percentage.
Comparing with earlier periods
Section titled “Comparing with earlier periods”A comparison is left out when the previous period reaches back before data existed for your game, because it would read as growth that never happened.
What this does not show
Section titled “What this does not show”The audience dashboard knows who opened your game, not how they did in it. Win rate, guesses and puzzle difficulty come from the SDK.
No individual player ever appears anywhere in this data, for you or for an assistant.