Publishing on Fortnite: How Discover, Featured Rows, and Algorithmic Reach Actually Work

Discoverability is its own discipline

You can build the best UEFN map of the year and still see almost no players if your publishing strategy is "ship and hope." Fortnite's Discover surface is a fast-moving, signal-heavy ranking system that rewards specific behaviors and punishes others, and the difference between a map that shows up on the front row and a map that doesn't is rarely the quality of the map itself. It's the shape of the cohort that played it in the first 72 hours.

This is a guide to how the surface actually works, what we know (and don't know) about the ranking signals, and the publishing-day patterns that consistently outperform the alternatives.

What "Discover" actually is

Discover is the primary entry point into UEFN content for most Fortnite players. It's a stack of horizontal rows, each with its own logic: some are algorithmic ("For You," "Trending"), some are categorical ("Action," "Adventure," "Party Royale"), some are editorial ("Featured," seasonal events, partner highlights), and some are persistent surface placements (top-of-screen banners, carousel rotations).

The crucial distinction for a creator is between algorithmic placement and editorial placement. Algorithmic placement is something you earn by performing well on signals the system can measure. Editorial placement is something Epic decides to give you, often based on inputs the algorithm doesn't see (production polish, brand fit, partnership context, narrative significance). They behave very differently and the strategies for getting each are almost entirely separate.

If you're a creator without an editorial relationship, your entire job is to optimize for algorithmic placement. The good news is that the algorithm is more knowable than people assume.

The signals that move algorithmic placement

Epic does not publish the ranking model, but the signals it pays attention to are observable in aggregate. Based on what consistently moves maps up and down across thousands of cohorts, the model is almost certainly weighting some combination of:

Click-through rate from impression. A player sees your map's tile in a row. Do they click it? CTR is the cheapest, fastest signal in the system, and it's the gating factor for almost every other surface. A map with weak CTR will not get the impressions to prove out its retention or session-length numbers, regardless of how good those would be.

First-session completion rate. A player who clicks in, plays for less than 60 seconds, and bails is a strong negative signal. A player who clicks in and plays a full round is a strong positive signal. This is the signal the algorithm uses to filter out maps with great tiles and weak first sessions.

Median session length. Total time spent is a vanity number; the median session length is the one the system seems to weight. A map where the median session is 25 minutes will outrank a map where the median is 4 minutes even if the latter has higher concurrents, because the long-session map is producing more engagement per impression.

Returning-player rate. Recall the previous post: D1 retention is the strongest predictor of a map's long-term shape. The algorithm seems to know this and weights returning-player rate heavily. A map that gets 15% of its day-one cohort back on day two is treated very differently from one that gets 5%.

Reporting and abandonment. Maps that are flagged, that have unusually high quit-out rates, or that show patterns consistent with low-quality content are demoted hard and fast. This is a one-way ratchet; once a map is in the demotion bucket, climbing back out is structurally difficult.

Cohort momentum. New maps with growing daily-active counts get extra impressions for free. The algorithm has a clear bias toward "things that are picking up steam," which is part of why the first 72 hours of a launch matter so disproportionately.

What the algorithm almost certainly does not directly weight: production budget, creator follower count off-platform, marketing spend, social-media reach. These influence the algorithm only through their effect on the signals above (more shares produce more impressions, which produce CTR data, which produces ranking).

The first 72 hours decide most of what happens

Almost every successful UEFN map can be traced back to a strong opening 72 hours. The reason is structural: the algorithm needs cohort data to evaluate a new map, the early cohorts disproportionately shape the model's view of the map, and the maps that win their first three days enter a feedback loop where success generates impressions that generate more success.

This means publishing day is not a release event, it's an optimization window. You're not just turning the lights on; you're producing the data that the algorithm will use to decide your map's ceiling for the next month.

A few patterns that consistently work:

Land your launch with a concentrated push. A diffuse trickle of players over the first week produces weak per-day cohorts that the algorithm reads as "low momentum." A concentrated push that brings in a strong wave of players in the first 24–48 hours produces the cohort-momentum signal you actually want. This usually means lining up creator shares, social posts, and any community channels you have to fire on day one or day two, not spread across the week.

Play to your strongest acquisition source first. If you have a creator audience, that's your strongest source: those players have the highest intent and will deliver the best D1 numbers, which seeds the algorithm with positive data. Algorithmic acquisition (random placements in Discover) brings in players with weak intent and depresses your early metrics, which is exactly what you don't want during the optimization window.

Watch the first-session metrics like a hawk. If you can see that first-session completion is below where you expected, fix it immediately. A patch that improves the first 60 seconds of the experience inside the first 72 hours can dramatically reshape the algorithmic curve. The same patch shipped a week later, after the algorithm has already decided what your map is, has a much smaller effect.

Resist the urge to ship fixes that change the loop in the middle of the launch window. This is counterintuitive but important: substantial gameplay changes in days two and three of a launch reset the cohort-comparison signals the algorithm is using and can cause your map to drop placements while the system re-evaluates. Save large changes for after the initial window stabilizes (typically around day 5–7).

Tiles, names, and the click-through battle

Your tile is the part of your map that 100x more players will see than will ever play. CTR is the gating factor for almost everything downstream, so the tile and name are not afterthoughts; they're the most-leveraged single piece of work in the entire publishing process.

What works for tiles, in our experience:

• One clear focal subject. Multiple competing focal points read as visual noise at thumbnail size. Pick one and let it dominate.
• Strong contrast and brand-distinct color. Tiles that blend into the row's background lose impressions. Tiles with a bold, saturated, brand-distinct color stand out even when they're 80 pixels wide.
• Faces, characters, and clear human subjects outperform abstract or environment-only tiles, almost universally. The exception is brand-recognized environments (a familiar IP location), but unless you're working with a known IP, lead with character.
• Tile text only when essential. If your name communicates the mode clearly, tile text is wasted space. If your name is abstract, a one-word tile label can help. Three-word tile labels almost never help.

For names: short, memorable, and unambiguous wins. A name that immediately tells the player "this is the kind of map I want right now" will out-CTR a more clever or thematic name 9 times out of 10. Save the cleverness for the description.

Editorial featuring: how to actually get it

Editorial featuring is the other half of discoverability and it works on completely different inputs from the algorithm. Editorial decisions are made by humans at Epic looking at the pipeline of upcoming and recent content, weighing things like production polish, freshness, partner relevance, and ecosystem fit. The signals that move editorial are:

• Polish that survives a quick play. An editor's spot-check of your map needs to land. Janky onboarding, broken systems, or visible bugs in the first three minutes will disqualify you regardless of how strong the rest is.
• A clear pitch. Your map should be summarizable in one sentence. "Co-op heist with a heat mechanic and a getaway phase." If the pitch takes a paragraph, the map is harder to feature because it's harder to slot into a row's narrative.
• Timing and freshness. Maps that align with seasonal events, ecosystem moments, or current Fortnite themes get featured more often. Plan launches around events when the calendar lets you.
• A track record. Creators who have shipped consistently and built a body of work are more likely to be featured because the editorial team has more confidence the next thing will land. If this is your first published map, the bar is higher; if it's your fifth competent one, the bar is meaningfully lower.
• Direct outreach, done well. Reaching out to Epic creator-relations channels with a clear pitch, a short clip, and a launch date well in advance puts you on the radar in a way that publishing into the void doesn't. Don't expect a response on every outreach; do expect that the cumulative effect of consistent, professional outreach builds awareness over time.

A common misconception: that featuring is the goal. It isn't. Featuring is a boost, not a foundation. A featured map with weak retention will burn through its boost in 48 hours and end up worse off than an unfeatured map with strong retention, because the boost cohort produces poor signals at scale and the algorithm pulls back hard. Build the foundation first; treat featuring as a multiplier on a foundation that already works.

What creators control vs. what they don't

A useful framing: there are roughly three buckets of things in publishing on UEFN.

Things you fully control: the map, the tile, the name, the description, the gameplay loop, the first-session experience, the catalog depth, the publish timing, your creator audience activation. This is most of the work, and it's where the most leverage is.

Things you influence: the cohort that comes in (through which channels you light up), the early retention numbers (through how good the first session is), the social-share rate (through how shareable your moments are). You can move all of these, but the magnitude depends partly on factors outside your control.

Things you don't control: the algorithm's exact weights, what other maps launch the same day, the editorial team's roadmap, seasonal calendar shifts in player behavior, large-scale Fortnite ecosystem events that change attention. These will affect your numbers and there's not much to do about them other than time launches around the calendar and accept that some weeks are softer than others.

The trap is to let things in the third bucket dominate your mental model of why a map did or didn't perform. They matter at the margin; they don't decide the outcome. The first bucket decides 80% of the outcome and that's where every hour of effort returns the most.

A publishing-day checklist that consistently works

• Tile and name pass a 80-pixel-thumbnail readability test.
• First 60 seconds of gameplay tested with at least three players who haven't seen the map before.
• First-session arc is concrete and completable in under 25 minutes.
• Daily reward, returning-session bonus, and streak mechanic are live on day one.
• Creator audience activation lined up to hit on day one or day two, not spread across the week.
• Social-share moments are visually distinct and clip-friendly.
• Bug-bash session done within 48 hours of publish; only critical fixes shipped during the launch window.
• Retention dashboard set up before launch so you're not building it under launch-day pressure.
• Plan for the post-launch week: what the first content beat is, when it ships, how you'll communicate it to returning players.

If those nine items are clean, you've done the foundational work that most launches skip. The map will get a fair shot at the algorithm, the algorithm will read the early signals correctly, and the rest of the discoverability story will follow from how good the map actually is.

Discoverability isn't a black box. It's a feedback system with knowable inputs and a strong bias toward maps that produce the right signals in their first three days. Build for that, measure what you can, and you'll find that "the algorithm" stops being a mystery and starts being a tool you can actually work with.

Related reading

• About to publish your first map? Make Your First Fortnite Map (Part 3): Ship It and Read the Numbers walks through the publish flow and the metrics worth watching in week one.
• New to UEFN? Start with Day One: From empty editor to a playable island.
• The Mapwright glossary defines Discover, island code, engagement payout, and the rest of the surface vocabulary used here.
• Pair this with the retention engineering guide and the UEFN economy field guide for the deeper systems work that decides whether a launch keeps its lift.