How many wishlists do you actually need?

Updated September 2026 · ShipYourGame

Two numbers get quoted constantly and neither is useful on its own: "you need 7,000 wishlists" and "wishlists convert at 10%". Both are roughly true and both hide the parts that decide whether your launch pays for itself.

What a wishlist is worth

A wishlist is a person who saw your page, liked it enough to save it, and was not willing to buy right then. On release, Steam emails them. Some fraction buys.

The commonly cited band is that 10-20% of your wishlists convert in the first month, higher if you launch with a discount, lower if your page has drifted away from what those people originally wanted. The long tail matters too: wishlists keep converting for years, particularly during sales.

The trap is applying that percentage to the sticker price. Between the shopper clicking buy and money reaching you, four things happen:

• The launch discount. A 10-20% cut is close to standard and does lift conversion, but it comes off the top.
• Regional pricing. Steam sells worldwide at locally adjusted prices. The average paid across all regions is typically well below the US price. Modelling 70-80% of the US price is more honest than modelling 100%.
• Refunds. Under two hours played and within two weeks, the customer gets their money back, no reason needed. A rate around 5-10% is normal; a slow opening pushes it higher.
• Tax, then Valve's share. Sales tax or VAT comes out before your split in most territories, and then the 70/30 split applies.

Multiply those together and the sticker-price estimate is off by a factor of three to five. Our wishlist revenue estimate shows each deduction separately so you can see which one is eating you.

The Popular Upcoming number

Steam has a front-page list of games about to release. Getting on it is meaningful free visibility in the days when it matters most, and it is widely reported to depend on wishlist count and recent wishlist velocity relative to other games launching the same week.

The figure that circulates is that you want somewhere around 7,000 to 10,000 wishlists to have a real chance. Treat it as a moving target rather than a threshold: it is a ranking against whoever else is launching that week, not a fixed bar. A quiet week needs fewer; December needs many more.

The practical reading is not "hit 7,000 or fail". It is that wishlist velocity in the final weeks counts, so a beat that produces 2,000 wishlists in the week before release is worth more than the same 2,000 spread over six months.

Which number to work backwards from

Do not start from wishlists. Start from what you need.

1. Decide what the game must earn to have been worth making, or to fund the next one.
2. Work out what reaches your bank per copy after discount, regional pricing, refunds, tax and the platform share. On a $14.99 game this is often in the $6-8 range.
3. Divide. That is your first-month unit target.
4. Divide by your conversion assumption, 10-20%. That is your wishlist target.

Doing it in that order tends to produce an uncomfortable number, which is the point. It is better to be uncomfortable four months before release than two weeks after.

What actually moves the count

In rough order of what the post-mortems keep reporting:

• A Coming Soon page that exists early. Wishlists accrue from the day the page is live. A page published three months before release collects for three months; one published three weeks before does not. Steam requires it live at least two weeks ahead, which is the floor, not the plan.
• Steam Next Fest. One per game, ever. Handled well it is usually the single largest wishlist event before launch.
• The capsule and the first three seconds of the trailer. Traffic that arrives and bounces is not a marketing problem, it is a page problem.
• Asking at the right moment. A prompt at the end of a demo run, when the player has just enjoyed something, outperforms a permanent button in the main menu.

What does not move it

Posting the same screenshot in more places. Buying followers. Launching the page and then going quiet for two months, which is the most common failure and the hardest to fix afterwards.