Indie Game Publisher vs Self-Publishing: The Real Math

Every month I watch another developer agonize over the same decision: pitch publishers or go it alone on Steam. The pitch-deck version of this debate compares a 50/50 split against keeping everything, and that framing hides where the money actually moves: recoup clauses, markup rates, and cost deductions that can mean earning $0 per copy for a year or more. So I pulled the best public contract data that exists, added what the 2026 funding market actually looks like, and ran the per-copy math both ways.

A good publisher brings four things: money, production support, platform access, and marketing muscle. Greg Lobanov (Wandersong, Chicory) puts concrete numbers on the support side: publishers budget roughly $10,000 per person per full-time month of remaining development, around $50,000 per platform for console ports, and about $2,500 per language to localize a 25,000-word game (the store-page side of that work is covered in our localization guide). If you’re mid-development and burning savings, the funding line is usually the real reason to sign. Compare those figures against your own burn rate with our development cost breakdown.

What a publisher can’t do is make an unmarketable game sell, and the good ones don’t pretend otherwise. Raw Fury receives over 1,500 pitches a year and signs under 1% of them (fewer than 7 titles in 2021). Publishers sign games that already look like winners: a sharp hook, a strong Steam page, visible traction. If your wishlist velocity is flat, your pitch will land in the pass pile, which is worth knowing before you spend three months on a deck instead of building wishlists.

My take: treat a publisher as an amplifier and a bank, never as a rescue plan.

First, Steam takes its 30% baseline cut off the top before any publisher split is calculated (the 25% and 20% tiers above $10M and $50M are covered in our Steam revenue share explainer). Everything below is about dividing what’s left, the “net revenue.”

A typical game publishing deal revenue split has two phases.

The recoup phase. The advance a publisher pays you is not income; it’s a loan repaid from your game’s earnings. Until the publisher has recouped that money, your share of each sale shrinks or disappears entirely. Lawyer Kellen Voyer reviewed 30 real indie publishing contracts for a GDC Summer 2020 talk, and it’s still the best public dataset we have: 42% of those deals paid the developer nothing until the publisher fully recouped, while 58% used a split recoup where both sides earn during repayment.

The post-recoup phase. Once the advance is paid back, the split shifts to its long-term rate. Rami Ismail’s Levelling The Playing Field documents the ranges: pre-recoup developer shares run from 0/100 (increasingly common) to 20/80 (the 2022 standard) to 30/70 (what he calls the old standard). Post-recoup, deals most commonly land around 50/50; developers who come in with real traction can push that to roughly 60/40 in their favor.

Across Voyer’s 30 contracts, developers averaged a 60% share of net revenue. The average hides a clear pattern, though: the more money a publisher puts in, the smaller your slice.

One Rami Ismail point I wish more developers internalized: “a higher Pre-Recoup percentage does not translate to higher earnings, only to faster earnings.” His negotiation priority is upfront payment first, then post-recoup share, then pre-recoup share. The post-recoup number is the one you’ll live with for the rest of the game’s life.

In Voyer’s dataset, the average indie game publishing advance was $318,000 including the no-advance deals, and $460,000 among deals that paid one, with a range of $100,000 to $2,000,000. And 18% of agreements had no advance at all.

That data is from 2020, and the market has moved — downward. Devolver Digital’s 2025 investor presentation showed average spend per game falling from $3M in 2022 to $2M in 2024, with plans to reach roughly $1M by 2026. Publishers are writing smaller checks, and fewer of them. GDC’s State of the Game Industry 2026 survey of 2,300+ developers found 35% are self-funded while only 20% rely on publishing deals or project-based financing, and the report notes traditional deals have become much harder to secure.

The contraction is structural, not a blip. Ziff Davis laid off all 36 employees of Humble Games in July 2024, part of a year in which industry job losses hit an estimated 11,400, above 2023’s roughly 10,500. Rebekah Saltsman of Finji, Tunic’s publisher, said in 2025 that an indie asking for $250,000 is “a drop in the bucket” yet companies consider that “isn’t profitable enough to fund,” calling this a “survival era” for indies.

So calibrate: the $460k-average era is the historical baseline, not the 2026 offer to expect. The one upside of a smaller advance is a shorter recoup tunnel.

Here’s the comparison that actually matters, using a $20 game. Steam’s 30% cut leaves $14.00 of net revenue per copy. Self publishing on Steam means you keep all $14. Under a publishing deal, what you keep depends on which phase you’re in.

Now layer recoup on top. Raw Fury open-sourced its publishing agreement in December 2020, which makes it a rare worked example: 50/50 net revenue after recoup, recoup of funding plus a 15% markup at a 100% recoup rate, and external costs (paid marketing, localization, porting) deducted from net revenue. GameDiscoverCo’s analysis ran the numbers: a $500,000 advance implies roughly $800,000 in breakeven costs, which means about $1.28M in gross Steam sales before the developer sees a royalty check. At $20 a copy, that’s around 64,000 copies sold before your per-copy earnings move off zero.

Not everyone thinks those terms are fine. Jan Willem Nijman (Vlambeer; Disc Room, Minit) called the fixed 50% post-recoup publisher share “absolutely WILD” and said 70/30 in the developer’s favor is industry standard for advances up to $100k. The spread between those two positions is the negotiation range you’re actually working in.

On the surface, decisively yes. VG Insights data shows third-party-published Steam games are twice as likely to earn at least $5,000 and four times as likely to earn at least $200,000 as self-published games. An Over Powered Game Marketing analysis of 2024 Steam releases found published games’ mean revenue was 2x and median revenue 5x that of self-published games.

But you can’t read those numbers as pure publisher effect. Publishers cherry-pick. The games they sign already have strong hooks, good pages, and traction, so that cohort would outperform even with zero publisher contribution. Remember the Raw Fury filter: under 1% of pitches signed. Selection bias does a lot of the work in those stats, and VG Insights itself notes the proportion of Steam games with a publisher has been declining for years.

What the gap does tell you: the bar keeps rising. More than 19,000 games released on Steam in 2025, a record, up from 18,477 in 2024 (Chris Zukowski’s January 2026 tally puts the 2025 total at 20,282), and almost half finished with fewer than 10 reviews. A publisher is one answer to that discoverability problem. If you self-publish, the marketing work still has to happen, and you’re the one doing it.

Signing makes sense when

• —You need funding to finish the game and have no cheaper capital. This is the strongest reason.
• —Console ports matter for your genre and you can’t fund the roughly $50k per platform yourself.
• —You have real traction (wishlist velocity, demo numbers, festival attention), which is exactly when your negotiating position is strongest.
• —The offered terms beat Voyer’s averages, or at least don’t trail them badly.

Self-publishing is the better call when

• —The game is nearly finished and funded. Giving up around half of net revenue forever to solve a problem you’ve already solved is a bad trade.
• —Your audience is niche and reachable: you know the subreddits, the Discords, the streamers. Publishers add the least value where the developer already owns the channel to the audience.
• —The only offers on the table are 0/100 full recoup with a small advance and a 50/50 tail. That deal structure only pays off if the publisher genuinely multiplies your sales.
• —You’d be signing mainly out of fear of marketing. A lot of what publishers do is teachable, and half of net revenue forever is expensive tuition.

These terms should end the conversation, not open a negotiation:

If a publisher reacts badly to you questioning any of these, that reaction is also data.

The choice stopped being binary years ago. A few models sit between a full deal and going it alone.

Lighter-touch publishers exist. Whitethorn Games takes a 10-25% cut of gross revenue with console porting included, on that 2-year renewable term. That’s a services-style deal, not a life-of-the-game commitment.

So do à la carte service shops. popagenda positions itself between an indie publisher and a PR agency: you pick specific publishing services (PR, release management, platform relations) on short contracts, with no revenue share on the whole game’s lifetime. You pay cash instead of equity in your game’s future.

Then there are funding-first structures. Epic Games Publishing’s model covers up to 100% of development costs with the developer keeping IP and at least 50% of profits after recoup. Worth knowing as a benchmark for what funding-heavy terms can look like, whatever you think of the storefront question (we compare platforms in Steam vs Epic for indie developers).

If you’re weighing paying for services out of pocket, our indie game marketing budget guide covers what those services cost and which ones matter at each stage.

1. 01Talk to their developers, especially the quiet ones. Not the breakout hit on the publisher’s homepage — the three games that shipped to modest numbers. Ask whether the publisher delivered after launch week.
2. 02Audit their recent catalog yourself. Pull up their last ten releases. Review counts, momentum, store page quality. If their pages are weaker than yours, what exactly are you buying?
3. 03Ask for the recoup ledger format up front. How are costs reported, how often, and what approval rights do you have over spend that lands in your recoup pool?
4. 04Check the company’s health. Humble Games’ developers learned in July 2024 that a publisher’s corporate parent can erase the whole operation overnight. Ask who owns the publisher and how the publishing arm makes money.
5. 05Demand a 90-day plan for your game. A real publisher will tell you specific beats: festival targets, demo timing, creator outreach. “We have a big audience” is not a plan.

What percentage does an indie game publisher take?

After Steam’s cut, expect the publisher to take 40-50% of net revenue post-recoup in a typical full-service deal. Across Voyer Law’s 30-contract dataset, developers averaged 60% of net overall: 71% in no-advance deals, 55% when an advance was paid, and 53% when the advance topped $500k. During recoup, the publisher’s share is larger, sometimes 100%.

Should I get a publisher for my indie game?

If you need money to finish the game, probably yes, if you can get acceptable terms in a market where GDC’s 2026 survey shows only 20% of developers are publisher- or project-financed. If the game is funded and nearly done, probably no: you’d be paying roughly half of net revenue forever for services you can buy or learn.

What is recoup in a publishing deal?

Recoup means the publisher recovers its advance and costs from the game’s revenue before you earn your full share. In 42% of the contracts Voyer reviewed, developers earned nothing until recoup completed. Watch the recoup rate (what percentage of revenue services the debt), the markup (Raw Fury charges 15% on top of funding), and which costs get added to the pool after signing.

Before you pitch anyone, get your numbers straight. Run your price point through the revenue calculator twice, once with 100% of net and once with the split you’ve been offered, and see how many copies it takes for the publisher path to come out ahead. That single comparison settles more publisher decisions than any pitch meeting. And if you land on going solo, start with the self-publishing walkthrough.