Steam Launch Strategy: The 12-Week Backward Plan (2026)

A Steam launch strategy is a backward plan: pick a release date, then work back 12 weeks and give every deadline a calendar slot. The hard constraints are set by Valve, not by you — your Coming Soon page must be live at least 2 weeks before release per the Steamworks release process, your release date locks 14 days out, your build needs review before you can ship, and if a Next Fest sits inside your window, its registration deadline lands 7 weeks before the fest even starts.

This post is the full 12-week plan: what happens each week, when your store page has to be finished (much earlier than most devs think), how to make the demo decision, when to send creator keys, and which wishlist numbers to check at each milestone. It ends where launch day begins — the day itself has its own launch day checklist, and this plan’s last job is to hand you off to it with nothing left to improvise.

Here is the whole thing in one table. Weeks count down to launch day (week 0).

Every row below gets its own section. If your game is more than 12 weeks out, everything here still applies — you just have slack, and the right use of slack is wishlist building, which our guide to marketing an indie game on Steam covers end to end.

The first decision is the date, because every other deadline derives from it. Three filters, applied in order:

1. 01Avoid Valve’s blackout zones. You cannot launch during a seasonal sale start, and launching in the shadow of one is nearly as bad — shopper attention and front-page real estate are spoken for. Valve publishes the official sales and events schedule for each half-year; check it before anything else.
2. 02Avoid your genre’s giants. A calendar-clear Tuesday means nothing if the sequel to your genre’s biggest game ships that Thursday. Track announced releases in your tags on SteamDB.
3. 03Pick the day of the week deliberately. Our best Steam launch dates breakdown covers the day-of-week evidence; the short version is that mid-week launches give you working press and streamers on day one plus a full weekend of visibility while your review score is still forming.

Run your candidate dates through the launch date planner — it checks sales collisions, fest overlaps, and the 2-week Coming Soon minimum in one pass.

One structural fact should anchor this decision: if you want a Next Fest inside your runway, the fest dictates the calendar. For the October 2026 edition, Steamworks lists registration closing August 31 at 11:59pm PDT, the fest running October 19-26, and eligibility requiring your release date to fall after October 26. A game launching in late November 2026 that wants that fest had to register nearly three months before launch. That is what “backward planning” means in practice.

Most launch plans fail here, quietly. The page work gets scheduled as “polish, ongoing” and then launch week arrives with a placeholder capsule. Give each asset a deadline:

• —Capsule art: done by week 8. It is the single highest-leverage asset for click-through, and every fest, creator video description, and Steam surface between now and launch will show it. Our capsule design guide has the specs and the patterns that work.
• —Screenshots: done by week 8. First four do the selling; they should read as gameplay at thumbnail size.
• —Trailer v1: done by week 8, final by week 4. If you are doing a fest, note that Valve pulls trailers for its own promotional material weeks early — for October 2026 Next Fest, trailer retrieval is September 7, six weeks before the fest.
• —Short description and tags: reviewed by week 8. Tags drive which shoppers Steam shows you to; run yours through the tag optimizer rather than guessing.

Why week 8 and not week 1? Because weeks 8 through 3 are when your page gets its highest-intent traffic of the entire pre-launch period — fest visitors, creator-video click-throughs, press readers. Traffic that lands on an unfinished page is spent. A visitor who bounced off a rough page in week 6 does not come back to check whether you fixed it.

If your Coming Soon page is not live yet at week 12, stop and fix that before anything else on this list. Valve requires the page to be up for at least 2 weeks before release, page review takes days on its own, and a page that has only existed for two weeks has had no time to gather wishlists. The realistic minimum for the page is months before launch; the mechanics are in our Coming Soon page guide.

There is no universal answer on demos, but there is a workable decision rule:

Two Valve rules shape the timing. First, titles participate in one Next Fest, ever, so you spend that shot when your demo represents the finished game — ideally the fest edition closest to launch that you can still hit. Second, the fest deadlines arrive in a cascade: for October 2026, demo build and store page submitted for review by September 21 to be included in the Press Preview, all materials final by October 5, Press Preview from October 8, fest live October 19. Full prep sequence in our Next Fest October 2026 guide.

If you launch a demo outside a fest, treat it as its own mini-launch: announce it, send it to creators, and leave it up long enough to gather feedback. Our demo best practices post covers scope and length — enough game to sell the hook, ending before the content runs thin.

Creator coverage is the main external traffic source you control, and the common mistake is sending everything at once, one week before launch. Split it:

Wave 1 (weeks 6-4): preview keys. Target 20-50 hand-picked creators who already cover your genre — their back catalog is the qualification, not their subscriber count. The pitch is short: one line on what the game is, one line on why their audience specifically, a key, and a press kit link. No embargo for small creators; embargoes are for coordinated day-one coverage, and most indies do not have the leverage to enforce one anyway.

Wave 2 (weeks 2-0): launch keys. A wider list, sent so that videos and streams can go live in launch week, when a purchase is possible. Coverage that lands three weeks early generates wishlists; coverage that lands launch week generates sales and reviews. You want both.

Between waves, watch what wave 1 produces. If a mid-size YouTuber’s video moves your wishlist graph, that is your signal about which creator profile converts for this game — weight wave 2 toward more of the same. Our full playbook, including pitch templates and key-distribution tools like Keymailer, is in how to get streamers to play your game. Curators are a lower-effort supplement in the same window; see the curator guide.

Wishlists are the plan’s instrument panel. The community-documented targets come from Chris Zukowski’s wishlist benchmarks (updated June 2026): games entering their final month with 10,000-20,000 wishlists reliably hold a Popular Upcoming slot, 7,000-10,000 get briefer placement, and under ~6,000 rarely appear at all. His long-standing floor recommendation is 7,000 wishlists before launching. What that list placement is worth is covered in our Popular Upcoming explainer, and the full target math in how many wishlists before launch.

Here is what a 12-week runway aiming for that floor can look like. This is an illustrative model, not survey data — the assumptions are in the source line, and your genre and starting point will move every number.

How to read your own numbers against a model like this:

• —Behind at week 8: you still have the fest or demo push ahead; hold course but tighten the creator list.
• —Behind at week 4 by more than ~1,000: consider slipping the date. Zukowski’s own rule is that if you are within about 1,000 of the 7,000 line, a delay plus a focused push is worth it; if you are stuck near 3,000 after six months of genuine effort, ship anyway — more months rarely fix a demand problem.
• —Paid ads to close a gap: viable at $1-2 per wishlist on well-optimized campaigns; stop or rework anything running above $3.

Two more numbers keep expectations honest. The median day-1 wishlist conversion is ~12%, so 8,000 wishlists models to roughly 960 first-day sales before organic traffic — run your own inputs through the wishlist conversion rates data. And with more than 19,000 games released on Steam in 2025, the wishlist count is the clearest pre-launch signal of which side of the median you are on.

Week 4 is the store page freeze. Final trailer up, final screenshots, final description. From here on, page changes are bug fixes only. If you have been running page A/B tests, conclude them now; a page that changes weekly in the final month confuses returning visitors and your own analytics. Run the frozen page through the store page checklist tool and the analyzer one last time — fixing a weak screenshot at week 4 pays for itself, fixing it at week 1 does not have time to.

Weeks 3-2 are for Valve’s machinery:

• —Submit the release build for review. Build review is a hard gate before release and takes time you do not control; budget at least a week of buffer, per the release process docs.
• —Configure the launch discount. A 10-20% launch discount is standard practice. Set it before the date locks, and remember the downstream constraint: after launch, the 30-day discount cooldown means your launch discount timing determines when you can discount again.
• —Confirm the date, then stop touching it. Inside 14 days, changing the release date means writing to Valve. Treat the 14-day mark as the point of no return and communicate the date everywhere: page, socials, creator emails, press kit.

The last week is deliberately boring. Keys for wave 2 are out, the build is approved, the page is frozen, the date is locked. What remains is a dry run: press the release button in your head, hour by hour, and confirm every step has an owner (even if every owner is you).

The hour-by-hour sequence — when to press the button, first-hour page checks, review-prompt timing, community hub setup, what to monitor and what to ignore — lives in the Steam launch day checklist. This plan’s job was to make that checklist executable: a launch day goes well in the preceding 12 weeks or not at all.

The compressed version of everything above, as a final self-check:

1. 01Date chosen against the sales calendar, genre calendar, and day-of-week evidence — verified in the launch date planner.
2. 02Store page final at week 4, not week 0.
3. 03Demo decision made on the decision rule, and the one-per-title Next Fest shot spent on the right edition.
4. 04Creators contacted in two waves: previews at weeks 6-4, launch keys at weeks 2-0.
5. 05Wishlist checkpoints tracked weekly, with the 7,000 floor and the delay-or-ship rule applied at week 4.
6. 06Build reviewed, discount set, date locked at 14 days.
7. 07Launch morning: open the runbook, execute.

If your page is the weak link in that list, start there — run it through the free Steam Page Analyzer today and you have 12 weeks to fix what it finds.