Key takeaways
- Games need a stable address plus entry points where readers already go: homepage, articles, newsletter, app.
- The New York Times judges its games on how many players come back every day for 7, then 30 days: that is the core metric.
- At Le Parisien, the target for games is 2 to 3% of total audience: a small share, but very loyal visitors (2024).
- Test one variable at a time, measure retention rather than clicks, and let tests run for several weeks.
- Connect games with editorial content so that players also become readers.
Where should games go on your site?
A daily game lives on an appointment: readers need to find it in the same place every day. The foundation is a dedicated page with a stable, memorable address, linked from the main menu. It hosts all your games and the day's puzzles.
That page is not enough on its own: you need entry points where readers already go.
- the homepage: a "Today's games" block visible without scrolling, naming the day's puzzles;
- the end of articles: a link to today's game, right when the reader has finished;
- the newsletter: a daily line reminding readers of today's puzzle;
- the app: a dedicated tab or menu entry.
The New York Times chose the last option: its games sit in a "Play" tab inside its news app, to build the "news plus games" habit in one place (Nieman Lab, September 2023). On apps, the 2026 report from Pugpig, a media app vendor, says users who play games open their app far more often than those who don't. That vendor figure, published as partner content, describes a correlation.
Personalised reminders come next. The Times says it uses its data to work out which puzzles each subscriber enjoys, then promotes that puzzle to them again on their preferred channels at the right time of day (Press Gazette, May 2025).
Which games should you start with?
Two or three well-established games beat a full catalogue nobody can find. Aim for different mechanics to reach different audiences: a word game, a logic game, a quick game.
Crosswords remain a pillar. At Le Parisien, they account for "almost half of games traffic, followed by puzzles and word games", according to La Revue des médias, published by INA, in April 2024. But a short game can attract people who don't play yet: in 2025 The Times launched Quizle, designed to be quick and easy, to reach the three quarters of its subscribers who were not puzzlers.
For a first pick, a Wordle-style word game such as Tusmo is a proven starting point (SIWA's word games are in French); the guide to games like Wordle covers this family of games.
Which KPIs show whether games work?
Set your metrics before launch and record their starting values. Without a baseline, there is no way to tell what the games change.
| Metric | What it measures |
|---|---|
| Daily active players | The size of the appointment |
| Completion rate | Puzzle difficulty and appeal |
| Return at 7 and 30 days | Habit formation |
| Consecutive days played | Depth of engagement |
| Clicks from games to articles | The bridge to your content |
| Sign-ups from games | Conversion into identified readers |
Daily return is the key metric. The New York Times judges new games on the share of users who play every day for 7 days, then every day for 30 days (Nieman Lab, September 2023). The Times, for its part, sees a 9-point retention gap between subscribers who play once a week and those who play two days in a row (Press Gazette, May 2025).
Do not judge games on audience share alone. Le Parisien's target is modest, but it is aimed at high-value visitors.
How do you test what works?
An A/B test shows two versions to two randomly selected groups of readers, then compares their behaviour. Applied to games, it lets you choose a placement, a label, a reminder time or a streak mechanic based on data rather than gut feeling.
Duolingo offers a public example. The company compared two versions of its app, one where the streak depended on the daily goal and one where the two were separate. The second improved Day 14 retention (Duolingo, November 2020). Another test, on an animation celebrating streak milestones, increased the share of new learners still active a week later by 1.7% (Duolingo, January 2022).
To measure the overall effect of games, not just a detail, keep a control group when you can: a small share of traffic that does not see the entry points to games for a few weeks. It is the only way to separate the effect of games from your readers' existing loyalty.
A few rules for reliable tests:
- one variable at a time: placement, or label, or timing, never all at once;
- retention before clicks: a louder block attracts clicks, not necessarily players who come back;
- enough time: habits take weeks to form, as the resource on streaks explains;
- modest gains: Duolingo's published improvements are measured in single-digit percentages; what counts is how they add up.
Launch checklist
Before going live:
- Pick two or three games with different mechanics.
- Create the dedicated page with a stable address and add it to the main menu.
- Plan entry points: homepage, end of articles, newsletter, app.
- Record baselines: returning visitors at 7 and 30 days, time spent, sign-ups.
- Set up tracking: active players, completion, return, clicks to articles.
After going live:
- Announce the launch in your newsletter and on social media.
- Connect games and articles: Ouest-France told INA it wants to "build a bridge between games and articles".
- Compare cohorts of players and non-players after a month.
- Run a first A/B test on the placement of the homepage block.
What SIWA takes care of
SIWA provides 16 daily games in French, word and logic, with a new puzzle every day. The games embed via an iframe or a script, in your colours and fonts; your space is set up in under 24 hours. The client portal tracks daily active players, sessions, completion rate, retention and top games, with CSV export and campaign links. Optional tracking of clicks to your articles measures the bridge between games and content. Details are on the pro page, and the link between games and loyalty is covered in the resource on audience retention.
Sources
- The New York Times finds a match with the word game Connections
- Times adds subscription sharing in bid to improve retention
- More people spending time on mobile apps, new report finds (Pugpig partner content)
- Jeux en ligne : un casse-tête rentable pour les titres de presse
- Improving the streak: Forming habits one lesson at a time
- The habit-building research behind your Duolingo streak