After judging 30+ startup pitches across accelerator programs in Lisbon, Latin America, Africa, and Asia-Pacific, I’ve noticed something surprising.
It wasn’t the best products that won. It was the teams with the best relationship with data.
Specifically: the ones who had done user interviews, changed their ICP as a result, and could attribute their traction signals clearly.
Everything else was noise.
The Pattern
I’ve sat in a lot of pitch rooms. Demo days. Accelerator graduations. Investor war rooms. Different geographies, different sectors, different stages — some B2C consumer apps, some deep tech B2B, some fintech, some climate.
The diversity is real. But one pattern cuts through everything.
The teams that win aren’t the ones with the most polished decks or the biggest TAM slides.
They’re the teams who have evidence they’ve been wrong about something important — and changed direction because of it.
Let me be more specific.
Every winning pitch I’ve seen shares three characteristics:
- Real user interviews conducted in the last 30 days. Not six months ago during customer discovery. Not “we talked to some people once.” Recent, documented conversations with actual users or buyers.
- An ICP that was changed because of what those interviews revealed. They thought their customer was X. The data said Y. They pivoted. And they can walk you through why.
- Attribution clarity. They know exactly what caused each traction signal. When signups jumped 339%, they can tell you which specific action drove it. When engagement dropped, they know why. They’re honest about what’s working and what’s noise.
This isn’t about having perfect data. It’s about having a real relationship with your data.
The teams that lose — even with strong products — are the ones still operating on conviction rather than evidence.
The Winning Habits
Let me break down what separates the teams that advance from the ones that don’t.
Habit 1: They Interview Users Constantly
Winning teams talk to users every week. Not just early on. Not just when something breaks. Continuously.
They have a system. They schedule 3-5 user conversations per week. They document what they learn. They share insights with the team. They update their ICP based on patterns.
Losing teams stopped interviewing once they had a product. They assume they know their customer now. They’re building features based on internal conviction, not external feedback.
The question investors and judges ask — explicitly or implicitly — is always the same: “How do you know?”
Winning teams answer with: “We spoke to 15 customers last month and here’s what we learned.”
Losing teams answer with: “We believe…” or “The market wants…” or “Our research shows…”
Belief doesn’t close deals. Evidence does.
Habit 2: They Change Their ICP Based on Data
Here’s the uncomfortable truth: your first ICP is almost always wrong.
You launch thinking your customer is mid-market SaaS companies. Then you realize your best users are actually enterprise logistics firms. Or you think you’re B2B, but your traction is coming from prosumer freelancers.
Winning teams notice this and pivot.
Losing teams ignore the signal because changing ICP feels like admitting failure. So they keep trying to sell to the customer they thought they had, even when the data is screaming that someone else wants the product more.
In one recent cohort, a team pivoted their ICP three times in four months. Each time, they had clear evidence: user interviews, usage data, churn analysis. They showed the evolution transparently in their pitch.
They won. Not despite the pivots — because of them.
Investors and judges aren’t looking for teams who got it right the first time. They’re looking for teams who recognize when they got it wrong and fix it fast.
Habit 3: They Attribute Everything
This is the hardest one. And the most important.
Winning teams can trace every traction signal to a specific action.
- Signups jumped 339% last month? They can tell you it was because they launched a new onboarding flow, and here’s the A/B test data.
- Churn dropped from 8% to 3%? They identified the exact feature that was causing friction and removed it.
- Viral coefficient increased? They know which referral mechanic drove it.
They’re honest about what they don’t know. When something worked, but they’re not sure why, they say so. When a metric looks good but might be vanity, they flag it.
Losing teams present traction without attribution. “We grew 200% month-over-month.” Okay — why? What did you do? Was it repeatable? Was it seasonal? Was it one big customer who might churn?
If you can’t explain what caused your growth, you can’t replicate it. And if you can’t replicate it, you don’t have a business — you have a lucky break.
The Common Failure Modes
Let me show you what losing looks like, because it’s easier to avoid than you think.
Failure Mode 1: Conviction-Based ICP
“We know who the customer is. We just haven’t spoken to many of them yet.”
This is the most common failure mode in early-stage pitches. The team has strong conviction about their target customer — but that conviction is based on assumptions, not evidence.
They’ve built features their imagined customer would want. They’ve priced for their imagined customer’s budget. They’ve written messaging for their imagined customer’s pain points.
And then they’re surprised when traction is weak.
The fix is simple but uncomfortable: talk to 20 potential customers in the next two weeks. Ask them about their problems, not your solution. Listen for patterns. Update your ICP based on what you hear, not what you hoped to hear.
Failure Mode 2: Vanity Metric Confidence
“We have 10,000 signups and 50,000 social media followers.”
Okay. How many of those signups are active? How many became paying customers? How many of those followers convert to users?
Vanity metrics feel good. They look impressive on pitch decks. But they don’t predict revenue.
Winning teams know the difference between a signal and noise. They track:
- Activation rate (signups who actually use the product)
- Retention (users who come back)
- Revenue per customer
- CAC payback period
Losing teams track:
- Total signups
- Social following
- Press mentions
- App downloads
If your traction metrics don’t connect directly to revenue or product-market fit, you’re measuring the wrong things.
Failure Mode 3: The Deck Is Better Than the Business
This one hurts to watch.
Beautiful slides. Polished design. Clear narrative. Great delivery.
And then the Q&A starts. And it becomes obvious the deck is telling a story the business hasn’t earned yet.
The team hasn’t talked to enough customers. They can’t explain their unit economics. They don’t know why their last growth spike happened. They’re pitching a vision, not a reality.
Judges and investors can tell. They’ve seen hundreds of pitches. They know what real traction looks like versus what a good storyteller can make traction look like.
The deck should reflect the business, not replace it.