Blog · Reviews
How long does a Steam game keep getting reviews after launch?
For years. A typical Steam game with 100+ reviews gets 22.8% of its three-year reviews in the first 28 days, 62% in the first year and half by day 226. Most never reach a plateau inside three years. Based on 41,854 games released 2016–2025, data as of 31 October 2026.
- 22.8% of three-year reviews arrive in the first 28 days (median)
- Day 226 when a typical game reaches half of its three-year reviews
- 50% of first-year reviews land in month one for 2025 releases, against 37% in 2016
- ×1.4 vs ×3.7 first-year reviews over month one, when week one is 90%+ or 30% or less of the month
How many reviews come in the first week and month?
About an eighth in the first week and under a quarter in the first month. The median game gets 12.5% of its three-year reviews by day 7 and 22.8% by day 28. The spread is wide: a quarter of games are past 35% at day 28, another quarter are under 13%.
Line: median game. Darker band: middle half of games. Lighter band: 10th to 90th percentile. Time axis is compressed so the first days stay readable.
| By | 25th percentile | Median | 75th percentile |
|---|---|---|---|
| Day 1 | 0.0% | 1.0% | 2.9% |
| Day 3 | 3.0% | 6.7% | 12.5% |
| Day 7 | 6.3% | 12.5% | 21.4% |
| Day 14 | 9.5% | 17.7% | 28.4% |
| Day 28 | 13.0% | 22.8% | 35.0% |
| Day 91 | 22.3% | 34.9% | 48.6% |
| Day 181 | 32.1% | 45.4% | 59.0% |
| Year 1 | 49.5% | 62.0% | 73.3% |
| Year 2 | 77.3% | 84.7% | 90.3% |
Share of the reviews a game collected in its first three years. 15,033 games older than three years with 100 or more reviews in that window.
When does a game reach half of its reviews?
Around day 226 for the median game. A quarter of games get there by day 102, another quarter only after day 367. Reaching 80% takes 634 days at the median, close to two years.
| Share of three-year reviews | 25th percentile | Median | 75th percentile |
|---|---|---|---|
| 25% | day 10 | day 37 | day 114 |
| 50% | day 102 | day 226 | day 367 |
| 80% | day 479 | day 634 | day 761 |
| 90% | day 714 | day 831 | day 916 |
Reviews slow down, but they do not stop. In its second year the median game still gets 7.3% of the daily review rate it had in its launch month. The third year runs at 72% of the second. That is why “time to 95% of all reviews” depends on the window you choose: there is no ceiling in the data to measure against.
Are Steam launches getting shorter?
Yes. For games released in 2016–2019, 36–42% of first-year reviews arrived in the first 28 days. For 2025 releases it is 50%. The first year now brings ×2.02 the reviews of the first month, down from ×2.70, and the second year is thinner: 23% of the first for 2024 releases against 36% for 2016.
| Released | Games | First 28 days, % of year one | Year one ÷ first 28 days | Year two, % of year one | Half of three years by |
|---|---|---|---|---|---|
| 2016 | 1,662 | 37.0% | ×2.70 | 36% | day 254 |
| 2017 | 1,581 | 39.0% | ×2.54 | 34% | day 280 |
| 2018 | 1,630 | 42.0% | ×2.39 | 47% | day 365 |
| 2019 | 1,802 | 36.0% | ×2.80 | 37% | day 275 |
| 2020 | 2,242 | 38.0% | ×2.64 | 30% | day 213 |
| 2021 | 2,466 | 42.0% | ×2.39 | 27% | day 182 |
| 2022 | 2,697 | 44.0% | ×2.26 | 27% | day 175 |
| 2023 | 3,085 | 45.0% | ×2.22 | 26% | day 176 |
| 2024 | 4,223 | 48.0% | ×2.10 | 23% | – |
| 2025 | 3,981 | 50.0% | ×2.02 | – | – |
Medians for games with 50 or more reviews in their first year. Dashes mark windows that have not closed yet.
- Observed The share of first-year reviews written in the first 28 days rose from 37% (2016 releases) to 50% (2025 releases). Steam user reviews · 2026-10-31
- Inference More of a game's audience now arrives at launch: wishlists, Next Fest demos and launch discounts concentrate buying into the first weeks.
- Inference Part of the trend can come from the data itself. Reviews on older games are deleted over time with their authors' accounts, which would thin the early years of older games most.
What does the first week say about the first year?
A lot. When nearly all of month one's reviews came in week one, the game has had most of its year already. When week one is a small part of the month, the tail is long.
| Week one, % of month one | Games | Year one ÷ first 28 days |
|---|---|---|
| 30% or less | 1,482 | ×3.71 |
| 30–45% | 3,521 | ×3.17 |
| 45–60% | 7,635 | ×2.53 |
| 60–75% | 8,843 | ×2.07 |
| 75–90% | 3,453 | ×1.67 |
| 90% or more | 411 | ×1.38 |
Median multiplier. Read it as an estimate: a game with 100 reviews in its first 28 days, 65% of them in week one, lands near 207 reviews by the end of year one. A single game can end well above or below its median.
To place your own game: add it on PlayRecon by its Steam link, read its first 28 days and its week-one share off the review chart, and find its row in the table.
What shapes do review curves take?
Six, from a launch burst to a slow build. We grouped curves by the share of three-year reviews reached at fixed days. Steady and Long together cover half of all games. Burst is the rarest.
Click a shape in the legend to hide or show it.
| Shape | Share of games | By day 7 | By day 28 | By year 1 | Half by | Largest examples |
|---|---|---|---|---|---|---|
| Burst | 4.6% | 49% | 66% | 90% | day 8 | Lost Ark, PAYDAY 3, Outriders |
| Strong start | 12.7% | 31% | 47% | 80% | day 38 | Cyberpunk 2077, Hogwarts Legacy, Starfield |
| Start and tail | 21.4% | 20% | 33% | 71% | day 118 | Elden Ring, Fall Guys, Battlefield 2042 |
| Steady | 26.3% | 12% | 22% | 61% | day 237 | Baldur's Gate 3, Hades, Resident Evil 4 |
| Long | 23.2% | 7% | 13% | 49% | day 368 | Phasmophobia, Destiny 2, Red Dead Redemption 2 |
| Slow build | 11.8% | 3% | 6% | 32% | day 570 | PUBG: Battlegrounds, Among Us, Rust |
Shares of three-year reviews, median within each group. 15,033 games older than three years with 100 or more reviews.
Which games keep collecting reviews longest?
Early Access games, multiplayer and co-op games, and the largest games. Early Access titles collect 35% of first-year reviews in month one, against 45% for full releases, and their first year is ×2.86 the first month instead of ×2.22. RPGs and roguelikes sit at the other end: half of their first-year reviews come in month one.
| Tag | Games | First 28 days, % of year one | Year one ÷ first 28 days |
|---|---|---|---|
| Action RPG | 1,056 | 50.8% | ×1.97 |
| JRPG | 1,101 | 50.6% | ×1.98 |
| Roguelite | 1,444 | 49.9% | ×2.00 |
| Roguelike | 1,325 | 49.8% | ×2.01 |
| RPG | 5,148 | 47.9% | ×2.09 |
| Point & Click | 1,378 | 39.5% | ×2.53 |
| Co-op | 1,238 | 39.1% | ×2.56 |
| Relaxing | 1,647 | 38.9% | ×2.57 |
| Multiplayer | 3,369 | 38.7% | ×2.58 |
Five tags with the shortest and four with the longest review life among the 60 most common tags. Medians for games with 50 or more reviews in year one.
How often does a game take off late?
In about one large game in eleven. Among games with 1,000+ reviews in three years, 9.3% (391 games) got less than 5% of those reviews in their first 28 days. 39% of them are Early Access games. The opposite case, more than 60% of three-year reviews in the first week, is rare at 0.4%, and mostly online games that later shut down.
| Game | Released | Reviews in three years | First 28 days | Year one |
|---|---|---|---|---|
| Among Us | 2018 | 521,239 | 0.0% | 0.2% |
| Wallpaper Engine | 2018 | 342,351 | 0.5% | 11.5% |
| Rust | 2018 | 309,958 | 1.9% | 13.6% |
| ARK: Survival Evolved | 2017 | 197,539 | 3.5% | 18.7% |
| Deep Rock Galactic | 2020 | 162,754 | 3.1% | 29.0% |
| Bloons TD 6 | 2018 | 131,787 | 0.2% | 4.8% |
Release year from the Steam store page. For Early Access games it is usually the start of Early Access.
Do review scores drop after launch?
Slightly, on average. The median share of positive reviews is 87.3% in the first 28 days and 84.1% over days 28–361. For 33% of games the share falls by 5 points or more; for 16% it rises by as much. Cheap games lose the most.
| Price | Change in positive share after month one |
|---|---|
| Free | -0.9 points |
| Under $5 | -3.1 points |
| $5–10 | -2.5 points |
| $10–20 | -2.0 points |
| $20–40 | -0.6 points |
| $40+ | -0.1 points |
Median change, 22,486 games with 30 or more reviews over days 28–361.
- Inference The first reviewers are the most committed buyers: people who wishlisted the game or played the demo. Later buyers arrive through discounts and recommendations and expect less specific things.
- Inference Patches can move the score either way, which fits the 16% of games whose share rises after launch.
What this data does not say
- Nothing direct about sales. Reviews are written by a small and changing share of buyers. A long review tail means a long tail of players who write reviews.
- Not why a curve has its shape. Updates, sales, ports and streamers all move reviews. This article measures the shape, not the causes.
- Not the future of one game. The tables are medians over thousands of games. A single launch can sit far from them in either direction.
How we counted
We took every public Steam user review for 41,854games with at least 30 reviews, released between 1 January 2016 and 31 October 2025, and placed each review on the day it was written relative to the release date on the store page. Reviews written before that date, such as during a closed test, are left out of the curve. The count includes reviews from keys activated outside Steam, which Steam leaves out of the score; our totals are a median 1.02 times the number Steam shows. Shares “of three years” use games released by October 2023; first-year figures use games released by October 2025. Data as of 31 October 2026. The review curve of any game, with its launch, updates and sales marked, is on its page in PlayRecon.
