Jordan Lee
Most first-time game budgets are wrong by a factor of two. Not 10-20% off. Double. I stopped being surprised by it a while ago, but I still think it’s worth understanding why, because it’s rarely what people assume.
It’s not that developers are naive or bad at math. Budgeting a game is just a genuinely weird problem: you’re pricing a creative product that doesn’t fully exist yet, built by a team whose velocity you can’t predict, on a timeline that will absolutely change. And most of the advice out there treats it like it’s just a spreadsheet exercise. It’s not.
There’s no universal correct way to do this. What I can share is the framework that’s worked across indie studios and what I learned producing projects from $40K to well over $2M.
Start with Scope, Not Numbers
Opening a spreadsheet before you have a concrete scope document is the single most common mistake. People want to know “how much will this cost” before they’ve seriously answered “what are we actually building.”
Your scope document doesn’t need to be elaborate. But it has to answer: How many hours of gameplay? How many unique environments? What’s the art style (this is a massive cost driver)? Voice acting? Multiplayer? What platforms? Every answer meaningfully changes the number.
Pixel art costs less than 3D. Significantly less. A game with 10 procedurally generated environments has a totally different production cost than one with 40 hand-crafted levels. Voice acting for a 20-hour RPG can run $30,000-$80,000 in talent fees alone before recording and editing. These aren’t details to figure out later.
Once you have a real scope document, you break it into feature categories: core systems, content, audio, UI, QA, marketing, platform certification. Then you estimate each. And here’s where most budgets start lying to themselves.
How to Actually Estimate Labor Costs
Labor is going to be 60-80% of your budget if you’re paying people fairly. Getting this right is everything.
The correct approach isn’t saying “this feature should take two weeks.” Break that feature into tasks, estimate each task, then apply a multiplier for reality. 1.5x is common for experienced teams. For new teams or genuinely novel technical problems, I’d go 2x without hesitation.
Research on software estimation is pretty consistent: humans are optimistic by nature. A 2019 analysis in IEEE Transactions on Software Engineering found that professional developers underestimate task duration by an average of 27% even when they’re trying to account for uncertainty. Game development adds creative iteration on top of that, which makes it worse.
Contractors work differently than full-time employees. A contractor might charge $75-$150/hour for character art, but that’s their actual rate. For full-time employees, the real cost in the US is typically 1.25x-1.4x their salary when you factor in payroll taxes, benefits, and equipment. Someone on a $70,000 salary is actually costing you closer to $90,000-$98,000 per year.
For tiny indie teams where founders work unpaid or deferred, still track what that labor is worth. Not because you’ll pay it now, but because if you don’t, you’ll misunderstand whether your project is actually viable and you won’t know what you need to earn at launch to make it worthwhile.
I’ve used Notion, Airtable, and Jira for task-based estimation. For most indie budgets under $500K, a well-structured Airtable base does the job better than Jira, which adds overhead without much benefit at that scale. Airtable costs around $20/seat/month and handles task breakdown, time estimates, assignments, and status tracking in one place.
The Budget Categories Most People Forget
Labor covered, there are categories that consistently get underestimated or missed entirely:
Software and tooling: Unity’s revenue share model changed significantly in 2024, and Unreal still takes royalties above $1M gross revenue. Beyond the engine, you might need middleware licenses. Wwise audio middleware is free up to a revenue threshold, but commercial tiers get real. FMOD is cheaper. Physics-heavy gameplay might need specialized tools. These costs add up fast if you don’t list them.
QA: Most indie studios budget almost nothing for QA, then spend the last three months doing QA themselves, which is the most expensive approach because your developers are your highest-paid testers. A proper QA budget should be 15-20% of your total development cost. If that shocks you, you probably haven’t shipped a game with online components or platform certification yet.
Platform certification: Sony, Microsoft, and Nintendo all have certification processes. Nintendo’s dev kit access has fees. Certification failures mean resubmission, which means delays, which means costs. Budget time for at least two submission cycles on every platform.
Marketing: This is uncomfortable. The conventional advice is 20-30% of your total budget, but indie developers I’ve talked to got traction on under $10,000 in paid marketing while others spent $100,000 and barely moved the needle. The research is genuinely mixed. If you budget $0 for marketing because “we’ll just do social media,” you’re probably in trouble. Even grassroots approaches need time and money.
Localization: Planning to release in Asian markets? Japan, China, and South Korea show up consistently in Steam data as massive revenue sources for indie games. Localization isn’t cheap. Japanese localization of a text-heavy game can run $0.08-$0.15 per word. A game with 100,000 words of dialogue is $8,000-$15,000 in Japanese translation alone, and that’s before testing.
Building In Contingency (Without Fooling Yourself)
| Category | Percentage of Budget | Notes |
|---|---|---|
| Labor | 60-80% | Highest cost component; includes salary multiplier of 1.25x-1.4x for full-time employees |
| QA | 15-20% | Often underestimated by indie studios |
| Marketing | 20-30% | Conventional advice; actual effectiveness varies widely |
| Contingency | 15-20% | 25% recommended for first-time teams or high technical risk |
| Other (Tools, Certification, Localization) | Varies | Software, platform certification, and localization costs vary by scope and platforms |
Every production budget I’ve seen that shipped on target had a contingency. Every one that didn’t had a crisis.
Standard advice is 15-20% contingency on your total budget. I’d push to 25% for first-time teams or projects with significant technical risk, like a new engine, a new genre for your studio, or complex online infrastructure.
Contingency isn’t a slush fund. It’s a specific reserve for scope changes and unexpected problems, and it needs separate tracking. The moment you treat it as part of regular budget, you lose the buffer you need and you’ll spend it on feature creep instead of emergencies.
A practice that helped me on a couple of projects was keeping a risk register. For each major risk (key developer leaves, platform rejection, engine update breaks something), you write down the probability, potential cost impact, and response plan. Not a fun document, but it forces you to think about where your contingency will actually go.
Tools like Runn or Float handle resource and budget forecasting at $6-$10/person/month and are genuinely useful once you’re managing five or more people with complex schedules.
The Cash Flow Problem Nobody Talks About Enough
Sources
- what are we actually building. Your
- on software estimation is pretty consistent: humans are optimistic by nature
- in IEEE Transactions on Software Engineering found that professional developer
- is genuinely mixed
- as massive revenue sources for indie games
Here’s something that kills projects even when the total budget is right: running out of cash before you run out of project.
Total budget and cash flow are different things. You might correctly estimate that your game costs $400,000 to finish, but if a publisher deal pays in milestones, you need to know when each payment hits, what you need to spend in the gaps, and whether you have enough runway to hit each milestone.
A $400,000 game where you receive $100,000 at greenlight, $150,000 at vertical slice, and $150,000 at content complete is actually a game where you need to survive three separate funding gaps. Miss a milestone and you might not make payroll.
Map your expected income (grants, publisher advances, pre-orders, whatever) against your projected monthly burn rate. Monthly burn rate is: (total labor cost + fixed costs) divided by the number of months in production. If that number scares you, it should. Most small indie studios operate on 2-3 months of runway, which means one bad month can be fatal.
In Canada, the UK, and several other countries, tax credits and government grants are genuinely material. Ontario’s OCASE tax credit, the UK Games Fund, and Australia’s Interactive Games Fund can change your cash flow picture significantly. These aren’t hypothetical: I’ve seen small studios recover from near-death situations because they finally applied for funding they qualified for all along.
