Build & Pitch: Clear the Runway
Apply the PM, UX, and debug lenses to a real build — an aviation-themed mini-game — then pitch it in one sentence.
❓ Answer Your Two Questions First
Before you open Gemini, write down — even just in your head — your answers to lesson one’s two questions:
- Who is this for?
- What’s the ONE thing it has to do well?
If you’re building today’s suggested project, here’s a starting point:
Build a fast-paced clicking game called "Clear the Runway" — a
grid of [6-9] runway slots, where a small airplane icon pops
into a random empty slot every [half a second], for a bored
first-time flyer killing time at a gate who wants a 30-second
distraction. The ONE thing it must do well: clicking an
airplane before it disappears scores a point, and missing it
costs nothing — just keeps the game moving. Show a running
score and a 30-second countdown timer. When the timer hits
zero, show my final score and a "Play Again" button.That’s your PM lens already baked into the prompt: a specific audience (“a bored first-time flyer killing time at a gate”) and a specific single job (“click the plane before it disappears, track the score”).
If you chose to iterate yesterday’s boarding-pass project instead, apply the same lens to it: name who it’s for, name the one thing it should do best, and prompt Gemini for one change that serves that job.
Run the loop — Describe, Build, Test, Refine, Repeat — just like yesterday. This time, as you test, check specifically for the three UX basics:
- Is there one clear main action on the page? (Clarity — it should be obvious you’re supposed to click the planes, not the empty slots)
- Does something visibly happen when you click? (Feedback — a color flash, a point counting up, a small “nice!”)
- Does the most important thing look the most important? (Hierarchy — your score and timer should be the biggest, boldest things on screen while the game runs)
It’s tempting to ask for sound effects or animations before you’ve confirmed the core loop actually works end to end. Get the ONE thing — a plane appearing, a click registering, the score going up — solid first. Polish is a refine step for after the core job works, not before.
🐛 When (Not If) It Breaks
Timers and random pop-ins are exactly the kind of feature that breaks in small, specific ways: a plane that never disappears, a score that counts twice, two planes appearing in the same slot. When it happens, don’t panic and don’t retype the whole prompt — run the debug loop from last lesson:
Notice something’s off → Isolate the exact click or moment → Describe expected vs. actual → Ask Gemini for that one fix → Retest.
For example: “When I click a plane, the score goes up by 2 instead of 1 — it should only count once per click.” That sentence alone is usually enough for Gemini to find and fix it in one pass.
☕ Break (10 min)
🛠️ Keep Building
Continue looping. If you finish your core feature early, revisit the UX triangle from lesson two — is there one place where clarity, feedback, or hierarchy is weak? Fix that before adding anything new.
🎤 Submit & PM Pitch
Save your file using the same convention as yesterday: name_day2_v1.html if this is a new project, or the next version number if you iterated. Save it into the Day 2 subfolder of the shared Drive (or commit it, if you’re using GitHub this week).
Then, in one sentence, tell the room:
“This is for [who], and it does [the one thing] well.”
That’s the whole pitch. No slides, no long explanation — just the PM lens, spoken out loud.
Two students both built Clear the Runway today. One says: 'It's a clicking game.' The other says: 'It's for someone bored at a gate, and it gives them a 30-second distraction where every click feels satisfying.' Which pitch shows stronger PM thinking, and why?
Reveal Answer & Explanation
The second one — it names a specific audience and a specific, single job the build does well.
“It’s a clicking game” describes a mechanic, not a purpose. The second pitch answers both of lesson one’s core questions in one sentence: who it’s for, and what it does well. That’s the difference between describing what something is and describing what it’s for — and it’s the same gap between a feature and a value.
🚀 Stretch Challenge (optional)
Finished early? Try one:
- Make planes appear faster as the timer counts down, so the game ramps up in difficulty
- Add a sound or animation when you hit vs. when you miss
- Save your high score with
localStorageso it survives a page refresh - Automate the boring part: ask Gemini to write a small script that auto-plays 20 rounds (clicking at random times) and reports your average score — a taste of using AI to automate a repetitive task instead of doing it by hand
⏭️ Up Next
Tomorrow, someone else uses your project — cold, with no explanation from you. What they say next changes everything about how you finish it.