What to put in a game publisher pitch deck

Raw Fury, the publisher of Kingdom and Sable, publishes its pitch requirements on its own site, and the most telling line is about the build: it cannot promise to play your game more than once, so the first play is the evaluation. That reframes the deck. A publisher reads it to decide whether to open the build, and to know what to look for when they do.

So a pitch deck is not a game design document, a trailer in slide form, or a business plan. Every slide answers one of three questions a publisher has to answer before it spends money: what the game is, why it will sell, and whether this team will finish it. A slide that answers none of them goes.

Ten slides cover it; fifteen is the ceiling. In the order a publisher's questions arrive:

Publishers differ in taste but agree on materials. Raw Fury's page, last updated in December 2024, asks the deck to answer six questions: what you are doing, why, who you are, where you want to go, what you need to get there, and when it will be done. Length and format are up to you.

For the build, Raw Fury asks for a playable PC build that conveys the essence of the game to outsiders, and lists what helps: skippable cutscenes and intros, checkpoints in different parts of the build, English subtitles if the spoken language is not English, and the controls, in a separate document if need be. A video playthrough is optional and welcome, ideally with timestamps.

In GamesBeat's 2018 round-up of publisher interviews, Devolver Digital asked for a playable prototype or a video of someone playing and explaining it; Team17 listed what a first email should hold (the premise, team size and experience, project status, target platforms, release window, footage from a vertical slice, and the support you need); and Raw Fury's then-head of business development said the budget is the item developers most often leave out.

Raw Fury's guidance is unusually specific: a development budget in US dollars for one PC build, covering only what it costs to reach a releasable game. Leave out marketing, PR, ports, events and localisation, which the publisher budgets separately. The simple version is months left multiplied by monthly burn.

Three things make the number credible:

• Lines, not a total. People by role and month, outsourced art and audio, QA, software, and a contingency line of 10% to 20%.
• A schedule that matches it. If the budget says nine months and the schedule says fourteen, the publisher will believe the schedule.
• A size that fits the market. The publisher will set the budget against what comparable games sold. If the ask needs a top-5% result to pay back, either cut scope or say why this game is a top-5% game.

Square Enix Collective's Phil Elliott added a practical point in 2018: plan cash flow around platforms paying roughly a month in arrears.

Every genre has one breakout, and naming it as your comparable tells a publisher you have not looked at the rest. Choose three to five games that are recent (the last three years), close in genre, price and scope, and honest about the range: one that did well, one that did fine, one that struggled.

For each, show public signals: release date, price, Steam review count, followers, and an estimated range of sales. The common shortcut is that a Steam game sells roughly 20 to 60 copies per review, a rough rule that varies by genre and price; our guide to estimating sales from reviews works an example, and the Steam revenue calculator takes copies to what reaches the developer. Then say in one line what your game shares with each and where it differs.

A publisher's real question is risk. Nothing lowers it like evidence that players already want the game:

• A live Steam page collecting wishlists, with the number and its trend. Even a few thousand wishlists with a steady daily rate says the hook works.
• Followers, which publishers can check themselves on any Steam page.
• Demo or playtest results: players, median play time, what they said.
• A clip that travelled: views on a short, a Reddit post, a festival showing.

If you have none yet, say what you will do to get them before launch. A deck that admits the gap and has a plan reads better than one that inflates a Discord count.

State the amount, what it pays for, and the services you want beyond money. Then know what you are likely to give in return. Voyer Law's analysis of more than 100 publishing agreements signed between 2017 and 2025 found that in deals with an advance, developers kept an average of 58.2% of revenue (median 50%), the average advance was about $675,000 and the median $300,000, and developers kept their IP in 96.4% of them. In deals without an advance the developer's average share was 67.9%.

Those are averages across a small sample, not a price list. They do tell you the shape: the more money you ask for, the larger the publisher's share, and the recoup terms matter as much as the split. Our revenue share guide works through the arithmetic.

Paste this into your slide tool and replace each line:

• Concept art where gameplay should be. A publisher funds the game, not the paintings. Every image should be from the build.
• Outlier comparables. One breakout as the only comp reads as wishful.
• A budget with no lines, or an ask with no reason.
• A deck that needs you in the room. It will be forwarded to people you never meet; each slide has to stand alone.
• Twenty-five slides. If it does not fit in fifteen, the game is not clear yet.
• A build that crashes in the first ten minutes. If it might be played once, make that once count.

Flockbay is a free AI game builder on the official Godot engine, and it prepares much of the evidence a deck needs from the game itself. Video studio cuts the core-loop frames, a trailer and devlogs from your real gameplay; the Steam numbers read your followers and an estimated wishlist range for free, and your exact wishlists from your own Steamworks key, which never leaves your machine; and playtest links bring back players, play time and crashes for the evidence slide. Video studio and hosted playtests are free in beta until 1 December 2026. Flockbay takes no share of your sales and no rights to your game, and the pitch, like the deal, is yours.