Talk2UserYou'll always have more feedback than you can act on. The mistake is trying to address all of it. The skill is knowing what to ignore. Here's how to find the signal in the noise.
The Core Idea
Weight feedback by who's asking. Power users and paying customers matter more than casual visitors. And always look for the 90/10 solution — 90% of the value with 10% of the effort.
The 90/10 Rule
This is one of Y Combinator's most repeated pieces of advice: when users ask for something, don't build the full version. Find a way to accomplish 90% of what they want with only 10% of the work.
A 90% solution now beats 100% later
If you can ship something that solves most of the problem today, do that. A perfect solution that takes months is worth less than a good-enough solution available immediately.
Users don't care how hard it was to build
They care if it solves their problem. So don't shy away from problems, but always ask: is there a simpler version that gets most of the way there?
Signs of Real Patterns
Frequency
The same issue comes up 5, 10, 20 times. That's a pattern. One person mentioning something is an anecdote.
Intensity
They didn't just mention it — they're frustrated. They wrote a long email. They almost churned. Intensity signals severity.
Who's asking
Your best customers, power users, and people who pay the most — their feedback gets extra weight. They've invested in your product.
Behavior, not words
Watch what they do, not just what they say. If they say they want feature X but never use features like X, that's a signal.
What to Prioritize
Customers who love you
A small group of customers who love you is better than a large group who kind of like you. Prioritize feedback from your most engaged users.
Burning problems
Recruiting 10 customers who have a burning problem is better than 1000 customers with a passing annoyance. Fix what hurts the most.
Things that unlock growth
If fixing something would let you unlock a new segment or reduce churn significantly, that's high leverage. Prioritize accordingly.
When to Ignore Feedback
Not your target customer
If someone wants you to build something for a use case you don't serve, that's not a signal — that's noise. You can't be everything to everyone.
Edge cases
If only 1% of users would benefit, and it takes significant effort, skip it. Focus on what helps the majority.
Conflicts with vision
Some feedback would take you in a direction you don't want to go. That's okay. You're building a specific product for specific people.
Sometimes You Need to Fire Your Customers
This is real YC advice. Some customers cost way more than they provide in either revenue or learning. Their feedback might be loud, but it's pulling you in the wrong direction.
- •If serving them means neglecting your core users, that's a problem.
- •If they're asking for a different product, let them go find it.
- •Twitch became successful when they stopped serving users who streamed copyrighted content and focused on gamers.
Key Takeaway
Weight feedback by who's asking. Power users and paying customers matter more than casual visitors. Look for the 90/10 solution. And don't be afraid to ignore feedback that pulls you away from your vision.