The Coding Project Your Kid Wants Isn't the One They Named

Last updated: August 2026

When you ask a child what they want to build, you usually get a sensible answer rather than an honest one. Across 14 tracked sessions with three teenage students in July 2026, every single one of them had a gap between the project we were officially working on and the project they actually cared about. In one case it took six sessions out of nine before the real one surfaced.

That gap is not dishonesty. A kid asked what they want to learn will offer something respectable, something that sounds like an answer an adult would approve of. The project they would work on at eleven at night without being asked is a different thing, and it is the one that determines how much they actually learn.

Key Takeaways

• All three of my July students had a stated project and a different real project underneath it.
• One student described the site we had spent five sessions on as "practice," and her genuine project did not surface until session six.
• The real project reveals itself in behaviour, not in answers: unrequested work, and what they want to show you at the end of a session.
• Once the true project surfaced, it pulled advanced material forward faster than any curriculum plan would have.
• The question that works is not "what do you want to learn," it is "what would you build if nobody was marking it."

Table of Contents

• Three Students, Three Hidden Projects
• Why Kids Give You the Respectable Answer
• The Signals That Give It Away
• What Changes When You Find It
• How to Ask Without Killing It
• Frequently Asked Questions
• The Bottom Line

Three Students, Three Hidden Projects

Alice, 15. We spent five sessions building a portfolio website. Good work, real progress, a live site with custom fonts, animated sections, and an interactive reveal she designed herself. In session five she mentioned, almost in passing, that she was treating the site as practice to learn how to build with AI tools, and that she had something she actually wanted to do but would not say what.

It came out in session six. She keeps a large collection of written notes about her own thinking and how her opinions change over time, and she wants to connect it to an AI system that can model how she thinks and argue back at her. That is a considerably more ambitious project than a portfolio, and it is the one she cares about.

Umar, 16. He came to me to finish a study and revision app he had started building himself. That was the whole basis of the engagement. Partway through, the work redirected toward something else entirely: a practical tool his family had a genuine use for. He raised it himself, unprompted, at the end of a session, having clearly been thinking about it.

Cameron, 15. On paper we are working through Python fundamentals, and we are, properly. But every session for months has ended the same way: he shows me the 2D fighting game he is building, and we talk about the new animations, the combo system, the health bars. That game is the reason he turns up. The Python is the route, not the destination.

Three students, three different shapes of the same thing.

Why Kids Give You the Respectable Answer

I do not think any of them were hiding anything deliberately. There are three ordinary reasons the true answer stays down.

The first is that the real project often sounds unreasonable when said out loud. "I want an AI that argues with me about my own opinions" is a strange sentence for a 15 year old to say to an adult they have known for a few weeks. A portfolio website is a safe, legible thing to want.

The second is that kids correctly read what is being asked. "What do you want to learn?" is heard as "what should you want to learn?" They answer the question they think you are asking.

The third is that they may not have words for it yet. Cameron never told me the game was the motivation. He did not have to. He showed me at the end of every session for months.

The Signals That Give It Away

Behaviour is far more reliable than answers here, and three signals did all the work in July.

What they do without being asked. Umar arrived at one session having independently added difficulty levels, a two player mode, and a moving background to his project. Nobody requested any of it. Unrequested work is the clearest signal there is, because nothing external is producing it.

What they want to show you. Every session with Cameron ends with the game. Not the Python files we spent an hour on. The game. Where a student steers the last five minutes tells you what they were thinking about during the other fifty five.

Hedging language about the current project. This is the one I nearly missed. When a student describes what you are building together as "practice," or "just to learn the tools," or "not the real one," that is a direct statement that something else is the real one. Alice said it plainly and I did not follow it up for a full session.

Signal	What to notice	How reliable
Unrequested work	Features or changes nobody asked for	Very high
End-of-session steering	What they want to show, not what you covered	High
Hedging about the project	"This is just practice," "not the real one"	High, and easy to miss
Direct answer to "what do you want to build"	The respectable version	Low on its own

What Changes When You Find It

This is the part that justifies the effort.

Once Alice's real project surfaced, the material we covered jumped several levels in a fortnight. To make progress on it she needed to understand how AI systems pull information from external sources, what it means to train a model further on a narrow set of examples, why a large context window changes what infrastructure you need, and how to structure a body of documents so a machine can navigate it. None of that was scheduled. Her interest dragged it forward.

That is a much better mechanism than the reverse. When I sequence advanced material because the curriculum says it is time, I have to generate the motivation. When a student's own project requires it, the motivation is already there and my job shrinks to explaining it clearly.

Cameron works the same way in a different register. Abstract logic taught in the abstract is a slog. The same logic framed around a game he is building lands immediately, which is why I anchor conditions and loops to things like frame rates and hit detection rather than to textbook examples. The mechanism behind this is well documented in the research on intrinsic motivation and self-determination, and it matches what I see across more than 200 students: interest that comes from inside the child does more work than any amount of external structure.

There is a second effect worth naming. Connecting new material to something a student already cares about is the learning strategy researchers call elaboration, and a real project supplies those connections for free. I do not have to invent a reason why loops matter to Cameron. His game already contains the reason.

It also changes the honest answer to "how long will this take." A student working on their own real project moves faster than the plan predicts. A student working on a respectable project moves slower.

How to Ask Without Killing It

A few things that work, and one that does not.

Ask what they would build if nobody was marking it. Removing the evaluation removes the reason to give a respectable answer. This is the single most effective question I use.

Ask what they would want it to do, before asking what it is. Kids often have a clear picture of a behaviour and no name for the thing. "It should be able to argue with me" came before any technical description.

Follow the hedge immediately. If they call the current work practice, ask what the non-practice version is, right then, in the same conversation. I lost a session by waiting.

Do not promise to switch to it straight away. The real project is often well beyond current ability, and the stated project may still be the right vehicle for now. Alice needed the five portfolio sessions to be able to attempt the harder thing at all. Finding the real project is about direction, not about abandoning what you are doing.

The one that does not work is asking in front of a parent. Every one of these three revelations happened one to one, in the middle of working on something else, with nobody else listening. That is not a coincidence.

Frequently Asked Questions

How do I find out what my child actually wants to build? Ask what they would make if nobody was grading it, and then watch what they do without being asked. In my sessions, unrequested work was a far more reliable signal than any answer to a direct question.

My child says they want to make games. Is that a real answer or a default one? Usually real, but too broad to act on. Push for the specific game and the specific feature they care about. One of my students is building a fighting game and the detail he cares about is the combo system, which is a much more useful thing to know than "games."

Should I let my child work on something too advanced for them? Direct the learning toward it without promising to build it immediately. One of my students needed five sessions on a simpler project before she could attempt the thing she actually wanted, and those five sessions were not wasted, they were the entry requirement.

Is it a problem if my child loses interest in the project we paid for? It is information, not a problem, and it usually means the real project has surfaced. Raise it with the tutor rather than pushing the child back onto the original plan. In my experience the pivot produces better work than the persistence would have.

Does the project have to be useful to be worth building? No. It has to be wanted. A game nobody plays and a tool a family genuinely needs teach the same fundamentals, and the one the child cares about teaches them faster.

Related Articles

• The Best Kids Coding Tutors Build Lessons Around Your Kid's Obsessions, the teaching approach this finding supports.
• The 30-Second Trick That Makes Kids Care About Code, the fast version of the same idea.
• How to Get Your Child Interested in Coding, what to do when the interest is not showing up at all.

The Bottom Line

All three of my July students were working on one project and thinking about another. The gap is not deception, it is what happens when you ask a child a question they hear as a test. The real project shows up in unrequested work, in what they want to show you at the end of a session, and in the word "practice." Find it and the learning accelerates on its own, because you stop supplying the motivation and start following it.

Want to know what your child would actually build given the chance? Book a free Discovery Call and I will spend the time finding out before we decide on a single lesson plan.