How to Vet a Technical Co-Founder in 2026: 7 Checks Before Equity
Seven checks a non-technical founder can run before granting equity, from a paid trial to an independent code review and the paperwork that protects the code.
Sep 29, 2026
How to Vet a Technical Co-Founder in 2026: 7 Checks Before Equity
Short answer: vet a technical co-founder on evidence you can check without being technical yourself: work they have shipped and can walk you through, a paid trial with a defined deliverable, an independent review of their code or plan by a senior engineer you pay, references from people who saw them under pressure, and signed paperwork before any equity is granted. Do all of it before the equity conversation, not after, because a co-founder split is far more common than founders assume.
The number to keep in mind: Carta found that among startups founded from 2016 to 2021, between 25% and 35% of two-founder teams had parted ways after five years, and more than 40% of two-founder teams founded from 2016 to 2018 had split within eight years (Carta, "Dynamic Duos: Equity Math for Two-Founder Teams", 18 June 2026). Vetting is not about distrust. It is about making the one decision in your company that is hardest to reverse with the same care you would give a hire.
Why vetting is harder when you are not technical
A non-technical founder cannot judge code quality by reading code, and should not pretend to. What you can judge is behaviour: whether someone finishes things, explains trade-offs plainly, admits what went wrong, and does what they said they would do on the date they said. The seven checks below are built around that. Each one produces evidence you can see, and none requires you to evaluate a line of code yourself.
The checks also matter more now than they did a few years ago. AI coding tools have made it easy for almost anyone to produce a working demo quickly, so a demo on its own no longer tells you much. What separates a technical co-founder from someone who can prompt a tool is judgment: what to build, what to leave out, how to keep customer data safe, and what to do when production breaks. Good candidates will be running the same kind of checks on you, and why technical people say no to co-founder offers covers what they look for.
The seven checks, in order
1. Shipped work they can walk you through
Ask for two or three things they built that real people used: a live product, an app store listing, an internal system at a previous job. Then ask them to walk you through one of them for 30 minutes. You are listening for specifics: who used it, what broke, what they would change. Vague answers about "the architecture" with no users attached are a signal.
If they point you to a GitHub profile, know what it can and cannot show. GitHub's own documentation shows that activity from private repositories appears on a profile only if the owner chooses to show it, and then only as anonymized activity, so a quiet profile is not proof of a quiet engineer, and a busy one is not proof of shipped products.
2. The failure conversation
Ask: "Tell me about something you built that failed, and why." Strong candidates answer with a specific product, a specific mistake and what they do differently now. Weak candidates blame a previous co-founder, a client or the market, and learned nothing they can name. This single question often tells you more than an hour of technical discussion.
3. An independent technical review
Pay a senior engineer you trust, who has no stake in the outcome, for a few hours to review the candidate's past code or their technical plan for your product. Ask the reviewer three questions: would you hire this person, what worries you, and what would you ask them next. You are buying a translation from someone with nothing to gain from either answer. Keep it short and paid, so it stays independent.
4. A paid trial with a defined deliverable
Before equity, work together on something real for two to four weeks: one feature, one prototype, one integration. Agree the deliverable in writing, pay for it in cash, and have the code land in a repository your company owns from the first commit. A trial answers the questions interviews cannot: whether you communicate well under a deadline, whether estimates hold, and whether you still want to work together on day twenty.
5. Commitment you can count
Ask directly: full-time or part-time, starting when, with how many months of personal runway, and what are they giving up. A co-founder who keeps a full-time job "until we raise" is a different arrangement from one who has quit, and the equity should reflect it. Technical co-founder equity sets out how timing, sacrifice and cash change the number.
6. References from bad weeks
Ask for two people who worked with them when something went wrong: an outage, a missed launch, a difficult client. Ask those people what the candidate did in the first 48 hours. Anyone can be pleasant in a good week. You are choosing someone for the bad ones.
7. Paperwork before code
Before meaningful work starts, sign three things: an agreement that assigns everything they build for the company to the company, vesting with a cliff so equity is earned over time rather than granted on day one, and a short founder agreement that says who decides what and what happens if someone leaves. If a co-founder later walks away, these documents decide whether the code stays with you. Technical co-founder left: who owns the code shows what happens when they are missing. Use a lawyer for the documents; this page is not legal advice.
Red flags, and what to do about each
| What you notice | What it usually means | What to do |
|---|---|---|
| No shipped work anyone used | They have built demos, not products | Run the paid trial before anything else |
| Resists a paid trial | Wants equity before evidence | Hold the equity conversation until after one |
| Every past failure was someone else's fault | Low ownership under pressure | Weight the reference calls heavily |
| Wants 50% on day one, part-time | Equity priced ahead of commitment | Tie equity to vesting and full-time start |
| Will not sign an IP assignment | The code may not end up belonging to the company | Do not start work until it is signed |
| Cannot explain a trade-off in plain words | Judgment you cannot check | Ask your independent reviewer to probe it |
None of these is automatically disqualifying. Each one is a question that deserves a clear answer before the equity is signed.
A four-week vetting plan
- Week 1: checks 1 and 2. Two conversations, one walkthrough of shipped work.
- Week 2: check 3, the independent review, and check 6, two reference calls.
- Weeks 2 to 4: check 4, the paid trial, with the IP assignment from check 7 signed before the first commit.
- End of week 4: check 5, then the equity and vesting conversation, with everything above on the table.
Four weeks feels slow when you want to start building. It is fast compared with unwinding a co-founder relationship two years in.
If vetting keeps ending in no
Sometimes the checks do their job and the answer is no, more than once. That is information, not failure. How to find a technical co-founder covers the seven routes to a candidate, and the platforms, ranked covers where to look.
The other route is to build first and choose a co-founder later, with a product and users to show them. That is the path set out on Technical Co-Founder Alternative: Build First, Keep Equity. KUMO builds the first production version for a fixed scope against published bands, not equity: a Starter Build is $15,000 to $50,000, and the code is yours from day one, so a future co-founder inherits a working product rather than a blank page. KUMO's founders have done the co-founder job themselves: they built Volopay's first production version as its founding engineers. How the engagement works is on founders partnership.
FAQ
How do I vet a technical co-founder if I am not technical? Judge evidence and behaviour, not code. Look at shipped work they can walk you through, run a paid trial with a defined deliverable, pay an independent senior engineer to review their code or plan, and call references from difficult weeks.
Should I do a trial project with a potential co-founder? Yes. Two to four weeks on one real deliverable, paid in cash, with the code in a repository your company owns, shows how you work together far better than interviews do.
What questions should I ask a potential technical co-founder? Ask them to walk you through something they shipped, to describe a failure and what they changed afterwards, how full-time they can be and from when, and what they are giving up to join.
How often do co-founders split up? Carta's data published 18 June 2026 found that 25% to 35% of two-founder teams founded from 2016 to 2021 had parted ways after five years.
What paperwork do I need before a co-founder starts building? An IP assignment to the company, vesting with a cliff, and a short founder agreement covering decisions and departures. Have a lawyer draft them.
What if I cannot find anyone who passes? Build the first version without a co-founder and recruit one later around a working product. The build-first path explains how.
Next step
If you have a candidate and want a second opinion on their plan, or you want to talk through building first, book a free 30-minute call.
Last verified 28 September 2026. Carta figures from the source linked above; GitHub behaviour from GitHub's own documentation on that date.
