What does a game designer actually do? 3 overlooked essentials

We've been doing this for years, and today we're revealing all! Here's what happens in almost every first meeting with a new client.

Every time we start working with a client, we find they haven't actually thought through why they want to make a game.

They haven't even asked whether making a game fits their marketing goals.

For a planner, whatever the task, aligning with the goal and getting the job done is what matters most!

From a game planner's point of view, we'll walk you through what a game planner is responsible for in a project, and whether AI can take on, or even replace, this role.

Key 1: A planner's job isn't writing documents

Many people picture a game planner as the person who sorts out requirements, writes gameplay descriptions and draws flowcharts. It sounds like admin work, as if you could have an experienced person do it on the side or let AI handle it.

But a planner's real job is to identify every potential problem and clear it out before development starts.

During development, programmers need to know where the logical boundaries of the gameplay are: what triggers this mechanic, how edge cases are handled, and what happens if a player does something unexpected. If the planner hasn't worked these out properly, engineers can only fill the gaps with their own judgment, and everyone's judgment turns out differently.

Artists need to know the game's visual language: the characters' emotional arc, the rhythm of UI interactions and what each screen should make the player feel. If no one spells it out, artists can only go on instinct. What they make may look great but not match what the game is trying to say.

When problems come up on both sides, the planner has to make trade-offs: which requirement gives way, which one stays, and how to balance budget and quality. These calls aren't in any document. They build up one decision after another during development. So, back to key 1: even when AI is used for planning tasks, the person using AI needs hands-on experience of their own for the plan to be accurate, and for the expected results to match the goals.

Key 2: Integration is the hardest part to see

Some clients leave the proposal discussion full of confidence, then hire engineers, artists and planners separately, assuming the game will come together on its own.

On paper, the project looks workable, but once it's actually running there are a lot of problems.

Halfway through, the programmer finds a logic gap in the design document and goes back to the planner, who says they hadn't considered that case and it needs to be discussed again. After finishing the third character, the artist realizes it doesn't match the style set by the first one and has to go back and redo it. Technology can't solve these problems. They come from a breakdown in integration. This happens often when work is outsourced in pieces, and it can easily delay the project's final schedule and push back the release.

In game development, finding individual skills is actually easy. The ability to bring programming, art and planning together is the hardest part.

The game planner role goes by different names in teams of different sizes. On small projects, the planner usually takes it on as well. Larger teams have a dedicated director or producer. Whatever the title, the core requirement is always the same: understand all three areas at once, and know how to pull things back when any of them goes wrong. That is the most critical point of all.

Across many projects, we've often seen clients with strong software development skills who underestimate the value of game planning, which leads to project delays.

Key 3: Why using AI to push the price down doesn't work

Over the past two years, clients have often asked us at the quoting stage: "Do you use AI? If so, can you lower the price accordingly?"

There are countless ways to answer this, and we tailor a comfortable solution to each client's situation. When the budget is truly limited, we tell clients directly: the art can be AI-generated without manual retouching, which brings the cost down but lowers quality. Visual consistency and control over detail will both suffer, so this trade-off needs careful thought.

If the client understands and accepts this premise, we proceed on that basis. But to be honest, the final product often falls short in quality. So we usually strongly recommend investing a bit more up front, which leaves us room to optimize later if anything still isn't right. For clients struggling with a fixed budget, we also help them revisit why they came to us and whether making a game is truly necessary.

If the idea is to use the existence of AI to push the overall quote down, that logic doesn't hold for game development.

Using AI to generate draft art, scaffold code and produce design documents saves some execution time, but it doesn't save any time on judgment.

A design document generated by AI still needs someone with professional background to review it line by line: whether this requirement and that one can coexist, whether this gameplay truly appeals to the target audience, and whether the flow and pacing are really designed into the experience or just exist on paper. Every one of these calls needs a professional eye to step in and check.

In one of our projects, the design document looked well paced and had a difficulty curve. But in playtesting with real people, players got stuck at one point in level 3. It wasn't that they found it hard. They found no reason to keep going. We went back, adjusted the pacing and changed the timing of one piece of feedback, and only then did the next test work. In other cases, a big gap in the art late in production made it hard for testers to complete their tasks.

None of this denies AI's usefulness. When the budget is limited, AI is an option, and we tell clients so ourselves. But when you choose AI, even if what you gain is lower execution cost, what you give up is stable quality and depth of judgment.

The importance of planning lies in the fact that a planner's judgment comes from experience: taking on real projects, watching players test and refining through repeated adjustments. AI can generate documents, but it has never sat beside a player and watched the look on their face as they gave up at level 3. That accumulated experience is something AI can't replace yet.

When the planner role is doing its job, you don't notice it's there. Often only after something goes wrong do you realize how important planning is, and that AI can't replace it.

In game production, engineers write logic, artists produce designs and everyone seems to be pushing forward at full speed. But is anyone looking at the big picture, judging whether these pieces will come together in the end and whether the product is heading in the right direction? That is what a planner is truly responsible for.

A planner's value has never been in producing a document that looks complete. It's in making, throughout production, the judgments and decisions no one notices but that cause problems the moment they're missing. Conflicting requirements need someone to mediate. Problems found in testing need someone to assess how to fix them. When the flow and pacing drift, someone needs to know where to start correcting it. These judgments, added up bit by bit, are the key difference in whether a game is truly fun. And this part is still something AI can't fully take on.

Talk with me about your direction before you make a game

FAQ

Q1: Are the game planner and the game producer the same person?

Not necessarily. In small marketing games, both roles are usually taken on by one person. The planner works out what to make and how, and the producer makes it actually happen. On small projects these two are hard to separate. Only large projects have a clear division of labor.

Q2: Can planning be outsourced?

Yes, as long as the other side truly understands your brand, target audience and marketing goals, rather than just filling in a template to write a generic document. The real risk in outsourcing is understanding: whether they have really figured out what this project is about. No matter how complete the design document is, if the writer doesn't truly understand the core of the project, judgment calls during development will still go wrong.

Q3: Can AI replace a game planner?

In our years of making games, what AI produces still needs someone to sit down, look at it and decide: this is right, that is wrong. Without that layer, what you make easily becomes a disposable game people forget after two minutes of play. AI helps you move a bit faster, but sitting down to look at that output and deciding what's right and what's wrong is something you can't skip.

Q4: How do you tell whether a planner is good enough?

The questions they ask tell you more than the documents they write. A good planner, early in a project, asks you many things that make you think, "I hadn't thought of that": in what situations the target audience will play, what success looks like to you, and if you could keep only one core feature, which one it would be. Being able to ask these shows they're really thinking about the project. Someone just filling in forms can't ask questions like these.

Q5: What is the most overlooked part of a planner's work on a project?

Stopping problems before they happen. This kind of work is completely invisible, because the problems never happen. Clients don't feel its value until they switch to a team that doesn't do it, and only then do they understand where all those problems that never happened would have come from.