The Feature Factory Disease
You have a product roadmap. It's beautiful. Fifteen features. Six months of work. You feel important.
Your first user leaves after two weeks.
Feature factories masquerade as execution. They're actually avoidance. You're building instead of selling. Shipping instead of learning. This is especially lethal for Indian startups with limited runway—₹30 lakh seed rounds don't survive guessing.
Michael Seibel's core principle: "If you're spending more than 25% of your time building features, you're not talking to users enough." Most Indian founders spend 80% building, 15% defending their choices, 5% actually talking to users.
Why Indian Startups Fall Into This Trap
Constraint misinterpretation. Indian founders often interpret limited resources as "move fast, ship hard." They don't pause to validate. They sprint toward some imaginary finish line.
Investor theater. Series A conversations in India reward breadth. More features sound like more progress. "We built 12 new features this quarter" gets nods. "We deleted 8 features because users didn't need them" gets silence. This kills founders.
Engineering bias. Strong technical teams build things. It's what they do. Without user research discipline, they become feature machines. You end up with a product that's technically beautiful and commercially dead. See: most enterprise SaaS from Bangalore.
The roadmap trap. Once you write down "Build X," it feels real. It feels committed. You stop questioning it. Sam Altman calls this "planning your way to failure." Plans are output. User needs are input. You got it backwards.
The YC Rule: 5 Users Before Code
YC's operating principle is brutal in its simplicity:
Before building ANY feature, talk to 5 users who need it.
Not want it. Need it. If you can't find 5 users screaming for it, kill the idea.
Why five? Because:
- One user = anecdote
- Two users = coincidence
- Three users = maybe
- Five users = pattern
This single filter kills 60% of feature ideas at Indian startups. Sixty percent. Gone. Your runway extends by months.
The Anti-Roadmap: What You Won't Build
Here's the non-obvious move: create an anti-roadmap. List features you explicitly reject.
Why? Because saying "yes" to everything means "no" to your core product. "No" to speed. "No" to clarity.
Example from The Messy Middle: Scott Belsky documents how Box (then not yet Box) rejected 200 feature requests to focus on one thing: sync. Dead simple. Worked. Now they're worth billions.
Your anti-roadmap should include:
- "We won't build multi-language support until we have 100K users" (know your stage)
- "We won't add collaboration features" (test if users actually want them)
- "We won't build mobile apps" (web-first until desktop is solved)
Write these down. Share them in investor updates. Watch investors get uncomfortable. That discomfort means you're thinking clearly.
The Validation Framework (For Indian Founders)
Stage 1: Problem Interview (Week 1)
Talk to 5 prospects. Don't pitch. Ask: "How do you solve this today?" Listen for frustration, workarounds, and what they're currently paying.
Stage 2: Solution Sketch (Week 2)
Sketch your idea. Show it to the same 5 users. Don't code. Use Figma. Ask: "Would you use this?" Measure: do 4 out of 5 say yes?
Stage 3: Minimum Implementation (Week 3-4)
Code the smallest version. Not minimum viable. Minimum real. Does it solve the problem? Does the user reach for it?
Stage 4: Delete or Double Down (Week 5)
If fewer than 3 of 5 users use it, delete it. If 4+ use it, double down. No middle ground.
This takes 5 weeks instead of 3 months. And you save ₹15 lakh in wasted engineering.
The Runaway Cost of Feature Factories
Meet Priya. She's a founder in Bangalore. Built an HR SaaS. Beautiful dashboard. 47 features. Series A. Raised ₹2 crore.
Unused features comprise 30% of her codebase. They slow releases. They confuse users. They make onboarding a 90-minute nightmare instead of 15 minutes.
She's now unfunding feature 15-47. The irony: she could have shipped the core product to 500 customers on the original timeline. Instead, she has 60 customers and a technical debt hangover.
Imagine that runway math. If each feature takes 2 weeks, she burned ₹25 lakh on features that don't matter.
The Anti-Feature Habit
Start today:
For the next feature you want to build:
- Don't build it
- Talk to 5 users for 1 hour each
- Ask: "Do you need this?"
- If 4 say yes, build it
- If fewer say yes, kill it publicly
Tell your team: "We deleted this feature request because nobody actually needs it." Watch team energy shift. You're suddenly about real problems, not imaginary ones.
The Actionable Move
This week: audit your current roadmap. For each feature in the next quarter, can you name 5 specific users who asked for it? Write down their names.
If you can't, delete it from the roadmap. Push the engineering team's free time into sales instead. Talk to users. Kill ideas before they consume runway.
Feature factories don't fail from bad execution. They fail from bad prioritization. YC's rule isn't about being selective. It's about being honest. Are you building for users or for yourself?
The best founders know the difference.