In partnership with
Your First Idea Is Almost Never the Best One. Here's the Science Behind It.
We all love the moment when the answer comes fast.
Someone on the team says it, the room agrees, and you move on. It feels productive. It feels decisive. And with AI, that moment comes even faster now.
But I believe the biggest risk in product development today is falling in love with the first idea.
In today's newsletter:
The 1985 study that explains why your team keeps picking the obvious answer
Why AI is making this habit worse, not better
Why testing three ideas beats perfecting one
Quick tips to generate multiple ideas, with or without AI
Why we pick the obvious answer (and the study that proved it)
In 1985, a psychologist named Gary Klein walked into a fire station with a simple question.
How do fire commanders decide what to do when a building is burning and lives are on the line?
He expected to find careful analysis. Weighing options. Maybe a mental checklist.
He found none of that.
Klein interviewed 26 experienced commanders — people with an average of 23 years on the job. He studied 156 critical decisions they had made. Real fires. Real pressure. Real consequences.
In 88% of those decisions, nobody compared a single alternative. They went with the first option that came to mind.
What were they doing instead? They grabbed the first idea that felt right. They ran a quick mental simulation — will this work? — and if it passed, they went with it. No plan B. No backup option.
For firefighters, this actually makes sense. The building is on fire. You don't have time to compare three strategies. Speed saves lives. Pattern recognition is a superpower when every second counts.
But here's the thing.
Your product team does the exact same thing — in meetings where nobody's life is on the line.
Someone proposes a solution. The room nods. The team starts building. Nobody stops to ask:
What if there's a better option we haven't even considered?
And now, AI is making this worse.
You can ask ChatGPT for a product solution and get a polished, confident answer in ten seconds. It feels like the right answer because it's fast, it's articulate, and it came with bullet points. So we ship it.
But speed was never the problem. The problem is that we stopped exploring alternatives. AI gives us the first option faster than ever — and we are even less likely to question it.
I believe this is one of the most expensive habits in product development. According to Harvard Business School, 95% of new products fail. And 35% of startups shut down specifically because they built something nobody needed.
The problem is rarely a lack of talent or tools. It's a lack of alternatives.
Why I make teams sketch three solutions before we discuss one
In my experience running design sprints, I've seen this pattern again and again.
A team walks in convinced they already know the answer. One person has a strong opinion. Everyone else follows.
Then I ask them to sketch three different approaches — individually, on paper, in silence.
What happens next is almost always the same. The winning idea? It's rarely the one everyone walked in with.
Sometimes it's a combination of two sketches. Sometimes it's something completely unexpected from the quietest person in the room. But it's almost never the first option that felt obvious to everyone.
This is the real value of a design sprint. It doesn't just generate ideas. It protects you from your own brain's shortcut — the one Klein documented in those firefighters.
Quick tips: how to actually generate three ideas
Your instinct to move fast is right. But "fast" doesn't mean "first idea wins."
Here's how to force yourself into multiple directions — with or without AI.
Without AI:
Sketch in silence. Give each team member 10 minutes to draw three different approaches on paper. No discussing first. This is the core of a design sprint — and it works because it removes groupthink before it starts. Once you have that foundation, push it further:
Use Crazy 8s. Fold a piece of paper into eight sections. Sketch eight variations of the same idea in eight minutes. Most will be bad. That's the point — you push past the obvious answers.
Ask "what would a competitor do?" Then ask "what would a completely different industry do?" Two quick reframes that pull your team out of their default thinking.
With AI:
Open three separate chats, not one. This is important. If you keep asking for alternatives in the same thread, AI builds on its own first answer. It converges. Start fresh each time with the same problem — you'll get genuinely different directions.
Change the constraint in each prompt. Same problem, different angle. "Solve this for speed." "Solve this for simplicity." "Solve this for a user who has never seen our product." Each constraint forces a different solution.
Use visual ideation tools. UX Pilot, for example, generates three to four screen variations from a single prompt. You see the options side by side instead of reading paragraphs of text. When you see alternatives, you make better decisions.
Key Takeaway
Science shows we grab the first idea and run with it — even for big product decisions.
AI makes this worse. The first answer comes faster and looks more polished, so we question it less.
Design sprints fix this by forcing multiple directions before anyone commits.
The rule: generate at least three options before you discuss one — whether you use paper sketches or AI prompts.
P.S. Klein's firefighters had 23 years of average experience. Their intuition was trained by thousands of real situations. AI has been trained on the internet, which is impressive, but it doesn't know your users.
Use AI to move faster, absolutely. But never let it replace the step where you test multiple directions with real people. That's the step most teams skip. And it's the one that actually matters.
AI Tools I'm Using This Week
💬 Claude — Anthropic's AI assistant, now outpacing ChatGPT in growth
🎨 UX Pilot — Generate multiple UI variations from a single prompt
🤖 Manus — Build presentations, videos, and prototypes from a prompt
The Architecture Behind AI-Native Revenue Automation
In our new white paper, The Architecture Behind AI-Native Revenue Automation, Tabs CTO Deepak Bapat breaks down what it actually takes to apply AI to revenue workflows without breaking the books.
You’ll learn why probabilistic reasoning isn’t enough for finance, how Tabs pairs LLMs with deterministic logic, and why a unified Commercial Graph is the foundation for scalable, audit-ready automation. From contract interpretation to cash application, this paper goes deep on where AI belongs—and where it absolutely doesn’t.
If you’re evaluating AI for billing, collections, or revenue operations, this is the architecture perspective most vendors won’t show you.