When to pay for a professional technical review, and what firms say it costs

A professional technical review is the right answer when the hour-long check is not proportionate to the money at risk. This page gives the decision rule, the prices that firms state on their own pages, and how to scope a review so that an angel group gets a page it can act on rather than a report it cannot read. Every price here is a vendor's own statement, attributed to the vendor and its page, and none has been audited.

When is paying proportionate?

The rule is simple: pay when a wrong technical answer would cost more than the review. Working through it takes three numbers.

Your cheque. Wiltbank and Boeker's November 2007 study of 539 group angels found the median investment per venture was $50,000 and the mean was $191,000, including follow-ons. An angel writing a $25,000 cheque who spends $20,000 on a review has spent most of the cheque on the decision, which is out of proportion unless the review is shared.

The group's combined exposure. In an angel group or syndicate, the review protects every member's cheque, and the cost can be shared. The Golden Seeds playbook from 2010 describes a group moving to full diligence only once at least $250,000 of interest is soft-circled. At that level, a $10,000 review is 4 percent of the combined exposure.

The size of the technical question. A pre-seed company with one developer and a simple product raises four technical questions, and the one-hour technical check answers them. A seed company with a team of five, an AI feature at the centre of the pitch, and enterprise customers asking about security raises questions the hour cannot answer.

Put together: a solo angel at pre-seed does the hour. A group or syndicate at seed, with a software product at the centre and a combined cheque in the hundreds of thousands, considers a review. The view from the other direction is in a 2016 Hacker News thread on whether startup technical diligence is worth doing, which reached 287 points; the top-voted reply, from a professional who does the work, put it as "when you put $50M into a company, spending $50k is a pretty good safeguard", while agreeing that at the angel stage the calculation is different.

What do firms say it costs?

These are the vendors' own figures, on their own pages, on the dates given. Each firm sets its own scope, and a "review" at one is not the same work as a "review" at another.

Vendor	Stated price	Stated duration	Where stated	Date
MEV	"$5,000 to $30,000, depending on the size and complexity of the system", "quoted upfront"	"2-4 weeks"	MEV's own technical due diligence guide	Published 16 January 2025, updated 5 August 2026
Specialist technical firms	"$35K-$95K" at "$300-$600/hour"	Not stated	Papermark's Due Diligence Cost in 2026 survey	Published 27 October 2025, updated 19 August 2026
Big Four technology advisory	"$50K-$150K" at "$400-$700/hour"	Not stated	Papermark, same page	Same
madewithlove	Not published	"within 2 weeks, conducting up to 8 interviews and a deep code review"	madewithlove's own guide, Yannick De Pauw	3 May 2024

Two cautions. Papermark is a data-room vendor summarising a market rather than quoting its own service, so its figures describe what it says firms charge rather than what any named firm has agreed to. And MEV's low end is the only figure on this table sized for an angel cheque; the rest are priced for a fund.

The ACA's own guidance has always allowed for this spend. Its July 2007 due diligence guide says investors "should also be ready to shoulder the cost of the due diligence process, which may prove to be considerable", and gives two examples from the time: a $4,900 investor due diligence report from Bioability taking 4 to 6 weeks, and a Forrester market report at $30,000 to $60,000. Those are 2007 figures for a different type of expertise, and they show that paying an outside expert has been standard angel practice for two decades.

How should an angel group scope the review?

The largest cost in a professional review is scope creep, and the cure is to write the scope before the engagement. For an angel group, the scope is the same four questions the group's own reviewer would answer, done properly with code access.

1. Does the product run for real users? With code access, this becomes: is the production system the one in the repository, and does the deployed version match the code the founder showed.
2. Who can change it, and what breaks when the founder is away? The commit history read in full, the deployment pipeline inspected, the written steps tested by having the reviewer follow them.
3. What does the AI claim mean? The test run, if there is one, and a small test built if there is not. The eval suite as a diligence artifact explains what a good one looks like.
4. Has anyone checked security? A scan run by the reviewer, with findings sorted into fix-before-close and fix-later.

Reveneau offers a review scoped to these four questions for angel groups and syndicates, delivered as a one-page finding with a flag against each question and a longer appendix for the member who wants it, with the scope agreed in writing before the work starts. If you want that run, the investor page is where to begin.

What to leave out, at seed: architecture opinions beyond "will this survive the next eighteen months", framework preferences, and cloud cost modelling. Those belong on a Series A review, and what technical due diligence costs covers how the price scales when they are added.

What should the group receive?

One page first, then the rest. The page has four rows, one per question, with what was found, a flag of fine, price it or stop, and one line the deal lead can read aloud. Behind it sits the appendix: the scan results, the test results, the commit analysis. The group's technical member reads the appendix; the lead reads the page.

Ask for the page format before engaging. A firm that will only deliver a forty-page report has priced its work for a fund with an analyst to read it. The technical due diligence report structure shows what the long form looks like, so that you can ask for the short one.

Ask also who does the work. Papermark's hourly ranges imply a senior person's time; check that the person on the call is the person reading the code. And ask what happens when the review finds a stop flag: a good firm tells the founder what to show, and re-checks once, inside the price.

How does this compare with the group's own reviewer?

	Group member	Professional review
Cost	A member's week	$5,000 and up, vendor-stated
Code access	Usually none	Full
Security	"Has a scan been run?"	A scan, with findings
AI claim	Sees the founder's test	Runs a test
Output	One page	One page plus appendix
Best at	Pre-seed, solo angel, small group cheque	Seed, syndicate, combined cheque in the hundreds of thousands

Assigning the technical reviewer in an angel group covers the left column. When a group has both, on the same deal, the two pages should agree on the stop flags and may differ on the terms, and the difference is where the group learns what its own reviewer sees and misses.

When is paying the wrong answer?

Three cases. When the round is pre-seed and the cheque is a solo angel's, because the hour is proportionate and the review is not. When the group already has a member who has shipped software recently and the four questions are all the deal raises. And when the founder has said, in writing, that they will not give code access, because a review without access is the one-hour check at a firm's price. In that last case the refusal is itself the finding, and technical red flags at pre-seed and seed says what to do with it.

For choosing between firms on a larger deal, how to choose a technical due diligence firm is the buyer's page. The pillar page puts this decision at the end of the angel process, after the hour and after the group's own page, because most pre-seed deals never need to reach it.