From 13% to 48% PMF in 20 Weeks: How BirBir Knew When to Scale
How a classifieds marketplace used the Sean Ellis PMF survey to avoid premature scaling and hit the 40% threshold before ramping up traffic
BirBir
Classifieds Marketplace Platform
"PMF is not magic, it's a method. The survey helped us keep pace with users before the budget burned out."
— BirBir Founder
The Challenge
In September 2024, BirBir launched their MVP classifieds platform with basic listing and search features. The product was live, users were signing up, but the founding team faced a critical question:
When should they scale traffic spend?
They had three options:
- ✗Wait for retention to plateau (day 90-180) — too slow, would miss the market window
- ✗Use DAU/MAU metrics — distorted by paid campaigns, not predictive
- ✓Measure PMF directly — Sean Ellis survey with 40% threshold
The team chose the third option. Analysis of 100+ startups showed that growth accelerates when ≥40% of users answer "very disappointed" to losing the product.
The Method: Three Surveys, Three Learnings
Survey #1: Wrong Question Wording
First attempt included NPS and too many response options. Result: 26% "very disappointed" — but the scale was ineffective.
Insight: Too many options diluted the signal. Needed to simplify to the core Sean Ellis format.
Survey #2: Clean Measurement
Ran a clean 4-question Sean Ellis survey. Result: Only 13% "very disappointed" — the product wasn't critical yet.
Insight: Product wasn't a must-have. Time to stop guessing and start fixing.
What they learned from qualitative feedback:
- • 80% of "very disappointed" users valued speed vs. competitors
- • Ad posting was too complex (too many steps)
- • Communication between buyers/sellers had friction
- • New users didn't understand the value prop immediately
Survey #3: Passed the Threshold
After 20 weeks of improvements, ran the same survey. Result: 48% "very disappointed" — ready to scale!
Insight: Hit the 40% threshold. Now they could confidently invest in traffic acquisition.
What They Changed Between Survey #2 and #3
In-Depth Interviews
Talked to users who answered "somewhat disappointed" to understand what was missing for them to rate higher.
Focus on Must-Have Use Case
80% valued speed vs. competitors. Doubled down on making buying/selling faster, postponed other improvements.
Eliminate Friction Points
Simplified ad posting (reduced steps), improved buyer-seller communication, added push notifications.
Convey Value Upfront
Improved onboarding so new users saw the speed advantage immediately instead of discovering it later.
The Results
Most importantly: BirBir avoided the #1 startup killer — premature scaling. They waited until 48% PMF before investing heavily in traffic acquisition, knowing their product was truly essential to users.
The Checklist: How to Repeat This
BirBir's founder shared their exact process for running the PMF survey:
Gather Your Sample
At least 40 respondents who used the core feature at least twice in the last two weeks.
Send the 4-Question Survey
Use the Sean Ellis format:
- • Q1: How disappointed would you be if you could no longer use [product]? (3 options + "don't use")
- • Q2: What is the main benefit you get from [product]?
- • Q3: What would you use as an alternative?
- • Q4: What can we improve?
Segment by Response
"Very" → core users | "Somewhat" → near-core | "Not" → noise
Extract Insights from Q2-Q4
Find your value narrative (how users articulate value) and blockers list (technical + UX issues).
Prioritize Your Backlog
Use the "minimum effort, maximum gain" rule based on what core users need.
Iterate & Repeat
Fix, ship, measure again. BirBir went from 13% → 48% in ~20 weeks with 2 major releases.
Key Takeaway: The survey turns "gut feeling" into a repeatable process: Measure → Find blockers → Fix → Repeat. BirBir knew exactly when to scale because their benchmark was 48%.
Customer Feedback Platform for SaaS & Startups
Don't just slap a Feedback button on every page. Place targeted in-product Surveys to get contextual feedback WHEN & WHERE it matters
• No Credit card needed • Install in minutes