How I Evaluate Early-Stage Startups: A Framework from 100+ Reviews
Most startup evaluation frameworks are written by people with cheque books. They’re written from 30,000 feet — TAM, SAM, SOM, unit economics, competitive moat. All valid. All things you get to eventually.
But before you get there, you spend a lot of time in rooms with founders who are pitching to you. And the frameworks that actually help you in those rooms are the ones built from repetition — from seeing the same patterns fail and succeed across dozens of pitches, sectors, and founder archetypes.
I’m not a cheque-writing investor. I’m the person in the room building the deal pipeline at Delhi Angels — evaluating startups, taking the notes, asking the follow-up questions, and deciding which founders make it to the next conversation. After reviewing over 100 early-stage startups across fintech, edtech, climate tech, and SaaS, here’s how I actually think about evaluation.
The First 5 Minutes Are the Most Revealing
Before the deck opens, I’m already forming a signal. Not a conclusion — just a signal. I’m watching for three things.
Can the founder explain what they do in one sentence? Not a polished elevator pitch. An honest, specific sentence. “We help D2C brands reduce return rates using post-purchase AI nudges” tells me more than “We’re building the future of commerce.” Clarity of explanation is a proxy for clarity of thinking.
Do they know who they’re talking to? Founders who walk into a Delhi Angels meeting without knowing what Delhi Angels does, who’s in the room, or what stage we typically invest at — that’s a signal. Not a dealbreaker, but a signal about how they do their homework.
What’s their energy around the problem? There’s a specific quality in founders who are genuinely obsessed with the problem they’re solving versus founders who found a market opportunity and decided to build something. The former can talk about the problem for an hour. The latter runs out of depth around slide 4.
The Questions That Do the Most Work
After 100+ evaluations, I’ve narrowed down to a handful of questions that consistently surface the most signal.
“What have you tried that didn’t work?” This is my favourite. Founders who have real, specific answers to this — “we tried X, it failed because Y, so we pivoted to Z” — understand their business at a granular level. Founders who struggle to answer this either haven’t been in the market long enough or aren’t being honest.
“Who said no to you and why?” Every startup has had investors, customers, or partners pass on them. How a founder talks about rejection tells you a lot. Do they understand the objection? Have they addressed it? Or do they dismiss it as the other person not understanding the vision?
“Walk me through your last 30 days of customer conversations.” This separates founders who are building from the market from those building in a vacuum. Strong founders have specific stories. Weak founders have aggregate statistics.
“What does the business look like in 18 months if everything goes right? What if funding falls through?” The first question tests ambition and vision. The second tests resilience and resourcefulness. Both answers matter equally.
What I’m Looking for in the Numbers
At the early stage in India, I’m not expecting perfect unit economics. Most pre-seed and seed companies don’t have them. What I am expecting is that the founder understands their numbers — and knows which ones actually matter for their business model.
The two metrics I focus on: retention and burn rate relative to progress. Retention tells you whether the product is solving a real problem. Burn rate relative to progress tells you how efficiently the team turns capital into learning.
A startup burning ₹15 lakh a month with 80% month-2 retention and three enterprise pilots is a very different conversation from one burning the same amount with 40% retention and no customer validation. Both have the same spend. Only one has earned the right to raise more.
The Red Flags That End Conversations Quickly
These aren’t the obvious ones. Everybody knows to be wary of founders with no domain knowledge or products with no differentiation. The subtler red flags are the ones worth noting.
Vagueness about competition. “We have no direct competitors” is almost never true and always worrying. Founders who say this either haven’t looked hard enough or are unwilling to acknowledge the landscape. Both are problems.
A team that can’t ship without the founder. At the early stage, I want to see that the founding team has shipped something — anything — without a full team in place. Founders who are still “building the MVP” six months in with three co-founders and a technical team raise questions about execution velocity.
Founder who dominates every answer in a co-founder pitch. When two or three co-founders pitch together and one person answers every question, I start wondering about the team dynamic. I’ll deliberately direct questions to the quieter co-founder to see how they respond.
A roadmap that solves every problem at once. Startups that plan to launch in five cities, serve three customer segments, and build four product features in the next six months are usually spreading themselves too thin. Focus is a discipline. Founders who have it stand out.
The Meta-Pattern Across 100+ Evaluations
Looking back across the startups I’ve evaluated, the single strongest predictor of a good outcome isn’t the market size, the product, or even the traction. It’s the founding team’s quality of thinking under pressure.
The pitches that go somewhere are the ones where, when I push back on an assumption, the founder doesn’t fold or deflect — they engage. They think out loud. They say “that’s a fair challenge — here’s how we’ve thought about it” or “honestly, we haven’t figured that out yet, but here’s our hypothesis.”
That intellectual honesty, combined with the energy to actually go build the thing, is what I’m ultimately looking for. Everything else is a proxy for it.
If you’re building in India and want to understand what the evaluation process looks like from the inside, feel free to reach out. I’m always open to honest conversations with founders who are doing the work.
Related reading: Inside Delhi NCR’s Startup Ecosystem — where these pitches actually happen, and what the ecosystem looks like from the ground. And: The AI Tools I Actually Use for Deal Sourcing — how AI supports the research and pipeline work around these evaluations.