How Discover Works in Fortnite and Who Gets Traffic

18 August 2026 · 7 min read

GameMapa editorial · how we check facts
Checked: Claude (AI) — fact-checked against the Ukrainian original; not proofread by a native speaker

Discover distributes traffic based on player behaviour: the system looks at QPTR, average playtime and retention — and new islands first go through a distribution test period lasting up to two weeks. This is what Epic’s documentation states directly. Everything else you’ve heard about “the algorithm” is speculation, and we’re not repeating it here.

Below is a breakdown of what’s actually known: how a new island’s launch works, which signals are named as key, why genre matters, and what numbers you can actually see in the Insights Dashboard. There’s a separate section on what we don’t claim — it’s just as useful as the list of facts.

The first two weeks: the distribution test period

When an island is first published, it doesn’t immediately drop into a settled exposure pattern. It first goes through a distribution test period — up to 2 weeks.

What this means in practice:

• The island can appear in rows from day one — both in genre rows and personalised ones. In other words, “zero impressions on launch day” isn’t the default norm.
• The first few days aren’t a final verdict. During this time the system is testing distribution, not locking in your ranking for good.
• The upper limit is stated explicitly: up to two weeks. Not “a month or two”, not “whenever the algorithm gets round to it”.

The main takeaway for creators: it’s too early to draw firm conclusions on day two after publishing. But there’s also no basis for assuming something will switch on by itself once two weeks are up — the signals are shaped by player behaviour, not by the calendar.

Key signals: QPTR, playtime, retention

The documentation names three things among the key Discover signals.

QPTR in plain terms

The acronym sounds intimidating, but the idea is simple. An ordinary play-through rate is the “saw it → played it” ratio. The word Qualified here means the metric is weighted by engagement: what matters isn’t the click itself, but how substantial the interaction actually was.

So traffic that came in and immediately left, and traffic that came in and played, aren’t treated the same by the system. That’s exactly why “more impressions” isn’t a goal in itself: an impression that a player didn’t stick around for doesn’t count the same as one that they did.

Exactly how Epic calculates the engagement weighting, with what formula and thresholds, hasn’t been published. So anyone who tells you the precise QPTR formula is giving you their own reconstruction, not the documentation.

Genre isn’t just a label for show

There’s a direct, unambiguous rule here: an island whose content doesn’t match its chosen genre won’t be shown in the matching genre rows or game-style rows.

This matters more than it might seem. The temptation to pick a “more popular” genre to land in a bigger row works against you: a mismatch between content and genre shuts the door on exactly the rows that the genre choice was supposed to unlock.

There’s one practical takeaway: set the genre your island actually is. Not the one with more traffic.

What the Insights Dashboard shows

The Insights Dashboard is a metrics panel for creators. Among the facts we verified, two metrics are clearly described, and both have specific definitions that are easy to mix up.

Ecosystem return. This metric counts players who come back to Fortnite after a break of 28 days or more. Note the level here: this is about returning to the game as a whole, not specifically to your island. It’s a “your content brought someone back into the ecosystem” metric, and reading it as ordinary retention is a mistake.

Daily Retention Rate. This one is about you specifically: it shows how many players come back to your islands on two consecutive days. Not “within a week”, not “ever again” — specifically two days in a row.

The difference in what these two figures actually measure is the main reason not to add them together or compare them directly.

What we don’t claim

This section is deliberately long, because this is exactly where guesswork usually creeps in.

• We don’t know the weighting of the signals. Whether QPTR matters more than average playtime hasn’t been published. Any specific weighting split quoted in guides is made up.
• We don’t know the threshold values. There’s nothing in the documentation about “you need this many minutes of session length” or “this percentage of retention”. If someone quotes you a number, ask for the source.
• We don’t know the full list of signals. The key ones are named. That’s not the same as “all of them”.
• We don’t know how ranking within a row is decided. The only confirmed fact is that rows can be genre-based and personalised.
• We don’t offer advice on gaming the metrics. We don’t describe or evaluate the “effectiveness” of artificially inflating metrics — there’s no basis for that in the documentation.

What this means for creators

An honest action list, based only on the facts above:

1. Set the correct genre. A mismatch between content and genre directly costs you exposure in genre rows and game-style rows.
2. Give it two weeks for the distribution test. Up to 2 weeks is the stated limit of the test period for a new island.
3. Pay attention to the first few minutes of gameplay. QPTR is weighted by engagement, and average playtime is a separate key signal. Both suffer where players don’t understand what to do at the start.
4. Give players a reason to come back tomorrow. Daily Retention Rate measures return specifically on two consecutive days — a very particular window.
5. Read the metrics by their actual definitions. Ecosystem return (28+ day break from Fortnite) and Daily Retention Rate (your islands, two days in a row) answer different questions.

Sources for all the facts in this piece: How Discover Works in Fortnite and Insights Dashboard in Fortnite Creative.

In short

Discover relies on player behaviour — QPTR, average playtime and retention — and a new island first goes through up to two weeks of distribution testing, appearing in rows from day one. Epic hasn’t published exact weightings, thresholds or “formulas”, so it’s worth checking your own progress against the Insights Dashboard metrics rather than other people’s guesswork.