Key points
- Diligence covers more than code: it looks at security, team and key-person risk, IP and licenses, infrastructure costs, roadmap realism and how you handle data.
- Do your own self-assessment first, so nothing in the review surprises you.
- Name your red flags and show the plan for each. Hiding a problem costs far more credibility than the problem does.
- One senior technical person should own the process, with the CEO closely involved.
- Start well before the process opens. Fixing things takes longer than documenting them.
What does technical due diligence look at?
Whether it comes from an investor, an acquirer or a large customer, technical due diligence asks one question: is this technology an asset or a liability? Reviewers usually work through the same areas. The depth varies with the deal.
- Architecture and scalability: can the system handle the growth in the plan, and what breaks first?
- Code quality: is the codebase maintainable, tested and understood by more than one person?
- Security and compliance: who can reach production and customer data, how secrets are managed, and where you stand on frameworks such as SOC 2, HIPAA or GDPR.
- Team and key-person risk: who holds critical knowledge, and what happens if they leave.
- IP and open-source licenses: do you own your code, including contractor work, and do any licenses create obligations?
- Infrastructure costs: what you spend, how it trends and how it scales per customer.
- Roadmap realism: does the plan match the team, the architecture and your delivery history?
- Data handling: what data you hold, where it goes (including third-party AI tools) and under what terms.
What documents should you prepare for technical due diligence?
Build a simple data room for technology before you are asked. Reviewers move faster, and trust you more, when the basics are ready and consistent with each other.
Keep each document short and current. A two-page architecture overview that matches reality beats a long wiki that does not.
- An architecture overview with a diagram, the main components and the key vendors.
- An engineering org chart, with roles, tenure and who owns which systems.
- Security policies, access lists, the most recent penetration test or assessment, and what you changed afterwards.
- Any compliance reports or a written status for each framework that applies to you.
- An open-source license inventory and confirmation that IP assignment agreements cover employees and contractors.
- Infrastructure and major vendor costs, with trends and contract renewal dates.
- Incident history: significant outages, root causes and what you changed.
- The current roadmap, and an honest account of what shipped against the last one.
How do you run a technical self-assessment before investors do?
Work through the same areas a reviewer would, in the same order, and write down what you find. Start with the things that are slow to fix: IP ownership, license problems, security gaps and single points of failure in the team. Documentation can wait until last because it is the fastest to produce.
Be specific. "Some tech debt" tells a reviewer nothing. "Billing runs on a service only one engineer understands, and we are pairing a second engineer on it this quarter" tells them you are in control.
Our checklist of the 40 questions your board will ask about your technology is a good place to start. If any answer makes you uncomfortable, a reviewer will probably ask it too.
Then turn the findings into a risk register: each risk, its severity, its owner and the plan. That single document often does more for a diligence process than everything else combined.
What are common technical due diligence red flags?
Most red flags are ordinary. Every growing company has some. What hurts a deal is a red flag the reviewer finds before you mention it.
For each one, the answer is the same: acknowledge it, size it and show the plan. Fix what you can before the process starts. For the rest, show that the work is scheduled, staffed and realistic.
- A single engineer who is the only person able to deploy, debug or explain a core system.
- Code written by contractors or a former agency with no clear IP assignment.
- Copyleft licenses in shipped code that nobody has reviewed.
- Broad production access, shared credentials or secrets stored in the codebase.
- Backups that have never been tested by actually restoring them.
- Infrastructure costs that grow faster than revenue, with no clear explanation.
- A roadmap that assumes far more delivery than the team has ever achieved.
Who should run technical due diligence preparation?
One senior technical leader should own it: usually your CTO or VP of Engineering. They need to know the systems well enough to answer follow-up questions without guessing, and enough seniority to speak plainly to investors about trade-offs.
The CEO stays close, because technical findings turn into commercial questions about valuation, terms and timelines. Legal counsel should handle IP and license questions alongside the technical lead.
If you do not have that leader, or your technical lead built most of the system and is too close to it to assess it fairly, bring in an outside senior perspective. A fractional CTO can run the self-assessment, prepare the documents and sit in on reviewer calls.
How long does technical due diligence preparation take?
It depends on what you find. Assembling documents for a well-run company is quick. Fixing a real problem, such as untangling IP ownership, replacing shared credentials or spreading knowledge beyond one engineer, takes much longer.
The practical rule is to start well before the process opens, not when the first request list arrives. Once a deal is underway, you will have time to answer questions, not to fix what they uncover.
If a fundraise or sale is likely in the next year, a self-assessment now is cheap insurance. It also tends to improve how you run engineering, whether or not the deal happens.
Frequently asked questions
Is technical due diligence different for a fundraise and an acquisition?
Yes, mostly in depth. Early-stage investors often focus on the team, architecture and the biggest risks. Acquirers usually go further into code, IP, licenses, security and integration, because they will own the result.
Will reviewers look at our actual source code?
Often, yes, especially in an acquisition. That may mean read-only repository access, a session with your engineers or an automated scan for code quality, security issues and open-source licenses. Agree the scope and confidentiality terms before granting access.
Should we fix every problem before diligence starts?
No. Fix what is quick or serious, such as exposed secrets or missing IP assignments. For the rest, a documented plan with an owner is usually enough. Rushed fixes made during a deal can create new problems.
How is a customer security review different?
It is narrower. A large customer mainly wants evidence about security, data handling, access control and reliability, often through a questionnaire. The same preparation, especially current security policies and a risk register, makes these reviews much faster.
What does an outside pre-diligence review include?
Digital Foundry's 30-day CTO Diagnostic covers an architecture and codebase review, a team and process assessment, a risk register and a roadmap, delivered as a report written for your board and investors.
Want a senior second opinion on your situation?
