The 90-Second Test: Film Your Child's Football, Get Feedback Before They Lose Interest

Your kid comes home from training, boots still on, and doesn't want to talk about what the coach said. So you don't push it. You grab a ball, head to the backyard, and they do a couple of shooting reps on their own. One of the strikes sails wide. You can see it, but you can't quite explain why, and they're already losing interest in repeating it.

That's the window. It's narrow. If you had a private coach on speed dial, you'd film it, send it, and get a breakdown by Thursday. But you don't. What you have is a phone in your pocket and about ninety seconds before your kid decides they'd rather go inside.

Here's the actual question most parents haven't tested: can you film a clip, upload it, and get back real, plain-English coaching feedback with a drill attached - before your child loses interest and walks off?

The answer is yes, and it takes less time than tying both boots. Let me walk you through the loop as it actually happens, because the speed isn't the selling point. The selling point is what you can do with the ninety seconds after the feedback lands.

The three things in the loop (and none of them are the AI)

Most parents assume the AI is the slow part. It isn't. The loop has three steps, and the model is the fastest of the three. Here's where the time goes.

Film (20-30 seconds). You need a clip of one technique - a pass, a shot, a first touch. Ten seconds of footage is plenty. You don't need a tripod, you don't need good lighting, and you don't need to film from a perfect angle. If you've read the filming guide, you already know the basics: hold the phone landscape, get the whole body in frame, and let your kid do the thing they normally do. No setup, no performance. Just film what's already happening.

Upload (10 seconds). You open the site on your phone's browser. No app to download, no account to create before you can try it. You tap upload, select the clip, and it sends. That's it. If you've ever emailed a video to a grandparent, you've done a harder version of this step.

Feedback (30-60 seconds). This is where the analysis runs. The clip is sent to Google Gemini - the same frontier model that enterprise tools use, accessed via the Enterprise API - and it reads the mechanics. Plant foot, body shape, strike point, follow-through. Then it builds a structured coaching report and puts it on the page. Not a chat reply. Not three paragraphs of generic encouragement. A named, sectioned report you can work through with your child.

That's the whole loop. Film, upload, read. Ninety seconds, give or take, and most of that is the model thinking. The bit you do - filming and tapping upload - is about forty seconds.

What comes back (and why the structure matters more than the speed)

Speed is useless if the thing that comes back isn't worth reading. So here's what actually lands on the feedback page - and it's more than you'd expect from a "free AI analysis."

The feedback is broken into six named sections. Not a paragraph. Not a chat bubble. A coaching report, structured the same way a real session plan would be. If you want the full walkthrough with screenshots, the dashboard guide goes through each section using a real clip. But here's the short version:

Overall summary. Two sentences. Plain English. What happened in the clip, without jargon. "Good power, but the shot went over the target" - that level. You read it in five seconds and you know the headline.

Strengths. Before it tells you what's wrong, it tells you what's already right, with a name on it. Shot Power. Clean First Touch. Body Position. This is the part you read out loud to your kid on the walk back inside. Not because you're being nice - because it's accurate, and they need to hear what's working before they hear what isn't.

Areas for improvement and drills. This is the core. Each area names the issue, finds the root cause (the mechanical reason it broke, in plain English - "leaning back at contact caused the foot to strike under the ball"), and prescribes a specific drill. Setup, execution, coaching points. Laid out like a session plan you could run that afternoon. And there's usually more than one - a single clip often surfaces two or three fixable things, each with its own drill.

Tactical and mental insights. A note on the decision around the technique, not just the technique itself. "The keeper was central, so placement would have beaten power." That kind of thing. It comes with a recommendation for next time.

Metrics to track. Two to four measurable signals - accuracy percentage, success rate, consistency - so that when you film the same technique in two weeks and upload again, you can see the numbers move instead of just guessing.

Next steps. One paragraph. One drill to focus on this week. Written to your child, in a coach's voice. The bit you read together.

None of that is generic. A chatbot asked about a shot that went over the bar would tell you to "keep your head down and follow through." That's true of every missed shot in history. What lands here is specific to your child's clip - their plant foot, their lean, their keeper position - because it read the actual frames you uploaded.

The part most parents don't ask about (until after they've uploaded)

You're filming your child. That's a different calculation than filming yourself, and the tool was built by a parent who made that calculation before writing a line of code.

Here's what happens to the video. It goes to Google Gemini via the Enterprise API - not the consumer tier. The difference matters: on the consumer tier, the provider can retain your data and use it to train their models. On the Enterprise tier, the Service Level Agreement explicitly prohibits it. Your child's video is processed, the analysis is generated, and the video is not retained by Google for training. It can't be. It comes in, the model reads it, and it's yours.

The clip auto-deletes after five days. Not a setting you have to find - that's the default. The written feedback (the strengths, the root causes, the drills, the metrics) stays in your profile so you can come back to it. If you want the video gone sooner, there's a one-click delete. No support ticket, no email chase, no thirty-day waiting period.

If you want the full breakdown of what to ask before uploading to any AI sports tool - not just this one - the privacy guide covers the five questions every parent should ask.

The ninety seconds after the feedback (the part that actually develops a player)

Here's the thing about the loop that the speed doesn't capture: the feedback isn't the end. It's the start.

You've got ninety seconds of attention from your kid, and thirty of those went to filming. That leaves about a minute - and in that minute, you read the strengths out loud, pick the first drill from the areas for improvement, and say "let's try that once more." The drill is already on the page. Setup, execution, coaching points. You don't need to invent it. You just run the rep.

That's the loop that actually works: film, feedback, one rep of the drill, done. Then next week, you film the same technique again, upload it, and see whether the root cause the analysis flagged has shifted. The metrics section tells you. The next-steps section tells you which drill to keep running. Over a few clips, you've got a feedback loop - and a feedback loop is what actually develops a player, not a single session.

Most parents never get this loop running because they assume it takes too long or costs too much. The first analysis is free - no credit card, no app, no subscription to cancel. If you've got a phone and a kid willing to do one repetition, you've got everything you need.

Want to see what comes back? Film a ten-second clip of one technique, upload it, and read the page that lands. If the feedback isn't specific to your child's clip - if it reads like something that could apply to any kid - you've lost ninety seconds and nothing else. That's the test.