5 Due Diligence Questions Every Investor Should Ask About Product & Tech

Most technical due diligence focuses on the wrong things. Code quality, test coverage, deployment frequency — these matter, but they don't tell you whether a company is actually viable. Commercial viability comes down to five questions. If you get clear answers to all five, you have what you need. If you get evasive answers to any of them, you have your answer.

1. What is your 12-month cohort retention?

This is the single most revealing product question you can ask. Not headline churn. Not logo retention. Twelve-month net revenue retention by cohort — the actual expansion and contraction of revenue from customers who signed a year ago.

• Good answer: A specific number above 80% NRR, with cohort data to support it and a clear explanation of what drives expansion versus what drives churn.
• Red flag: Any version of 'we don't track it that way' or an inability to produce cohort data. At Series A and beyond, this is a significant operational gap — and often a signal that the retention picture is worse than the headline metrics suggest.

2. What happens to your system at 10× current load?

Scalability isn't just a technical question — it's a commercial one. If the system can't scale, the growth story can't either. You're looking for evidence that the team has actually thought this through, not just assumed the cloud handles it.

• Good answer: A specific architectural description — where the bottlenecks are today, what's been designed to scale horizontally, what would require re-engineering at the next order of magnitude, and what that work would cost.
• Red flag: 'We're on AWS, so it scales automatically.' This tells you the team doesn't understand their own architecture. Stateless services can auto-scale. Databases, legacy components, and third-party integrations generally cannot — and these are where real scaling failures occur.

3. How long until a new customer gets their first meaningful outcome?

Time-to-value is a leading indicator of retention and expansion. Long onboarding cycles correlate strongly with churn — customers who don't reach value quickly tend not to renew. Short time-to-value creates the conditions for expansion revenue.

• Good answer: Days or weeks, with specific onboarding data to back it up. Ideally: 'Our median customer sees X outcome within Y days, based on cohort analysis.' The existence of the data matters as much as the number.
• Red flag: 'It depends on the customer.' This is almost always true — but a good answer acknowledges that and then gives you the distribution. If every onboarding is genuinely bespoke with no predictable timeline, that's a product problem with real commercial implications.

4. What would a competitor need to replicate your core technical capability?

This is the moat question, asked in a way that requires a specific answer rather than a positioning statement. You're looking for concrete barriers — proprietary data, novel algorithms, years of training data accumulation, regulatory position — not narrative.

• Good answer: A specific description of what makes the capability hard to replicate. 'We have four years of proprietary transaction data that trains our fraud model — a new entrant would need equivalent data to match our accuracy, and that takes time to accumulate.' That's a moat.
• Red flag: 'We move fast and our team is really strong.' Execution speed is not a technical moat. Strong teams can be hired. The only defensible answer involves something structural that can't be easily replicated with capital.

5. Who decides what gets built and how?

Product prioritisation reveals how a company will allocate its most constrained resource — engineering time — as it scales. Founder intuition works at pre-product-market-fit. At Series A and beyond, you want a process.

• Good answer: A clear description of how customer feedback, usage data, and commercial priorities translate into the roadmap. Specific examples of decisions that were made based on data, and decisions that were reversed when the data changed.
• Red flag: 'The founders decide.' Or any version of prioritisation that starts and ends with founder conviction rather than customer and commercial data. This isn't about removing founder judgment — it's about whether that judgment is informed or instinctive.

The meta-point

None of these questions require a technical background to ask or evaluate. They require knowing what good looks like — and being willing to sit with an uncomfortable silence when the answer doesn't come. The companies that can answer all five clearly are worth more than their comparable peers. The ones that can't are telling you something important.