Technical Due Diligence: Four Findings That Moved the Deal Price

The fund was in exclusivity on a healthtech platform serving outpatient clinics. Commercial and financial diligence had gone well: strong recurring revenue, 93% gross retention, credible growth plan. The investment thesis depended on one thing — that the platform could support a doubling of customers over three years and the addition of two new product lines without a major rebuild.

The fund had no way to test that claim. The target's engineering leadership said the platform was modern and scalable. That is what every target says, and it is neither dishonest nor informative.

Envion was engaged for technical due diligence with an explicit brief: assess whether the investment thesis is technically deliverable, and quantify what it would cost if it isn't.

Envion reported eleven findings. Four were material to the deal.

1. Tenancy model blocked the growth plan. The platform used a database-per-customer model — 340 separate PostgreSQL databases, each with its own schema version. Migrations were applied per customer, and 61 databases were behind on schema. This worked at 340 customers with heroic effort from two people. At the 700 customers the thesis assumed, it would not. Migration time already ran nine days per release. Business impact: re-platforming to a shared multi-tenant model, estimated at €2.1M and 14 months, and it had to happen before growth, not after — meaning it would consume the first third of the hold period.

2. Key-person concentration at a level the fund had not seen. One engineer had authored 71% of the billing and claims-adjudication module — the most business-critical and most regulated code in the product. No one else had committed to it meaningfully in eighteen months. He had been with the company eleven years and, in interview, was candid that he had stayed for reasons connected to the founder, who would be exiting at close. Business impact: a departure risk of severe operational consequence, occurring at precisely the moment the deal made it most likely.

3. Delivery capability was weaker than the roadmap required. The growth plan assumed two new product lines in 24 months. The historical evidence: median cycle time 21 days, change failure rate 19%, and 40% of engineering capacity spent on unplanned work over the trailing year. A team spending two days in five firefighting does not ship two new product lines. Business impact: either the roadmap slips substantially, or roughly €1.4M of additional engineering investment is required to buy back the capacity first.

4. A compliance gap with real exposure. The platform handled PHI. Audit logging was implemented in the main application path but absent in three admin tools and one batch process, meaning certain access to patient records left no trace. Not currently breached, and not something the target had concealed — nobody had checked. Business impact: remediation was modest at roughly €120K, but the exposure window was historical and unquantifiable, which is a materially different conversation.

The fund negotiated a €2.4M purchase price adjustment against €3.6M of quantified remediation, agreed a retention package for the key engineer before close, and completed the deal on revised terms. Re-platforming started in month 2 and was delivered in month 15; by month 18, cycle time had fallen from 21 days to 8. The fund's operating partner has since made technical due diligence standard on software deals above a threshold.

The advice that generalizes: delivery history is the most honest evidence available — cycle time, change failure rate and unplanned-work share are measured facts sitting in version control and ticket systems, and they predict what the team will actually deliver post-close. Look for concentration, not just quality: code can be excellent and still represent severe risk if one person holds it, and commit distribution across critical modules takes an hour to compute. And findings are only useful if they're costed and timed — "technical debt in the tenancy layer" is not actionable; "€2.1M, 14 months, must precede growth" is a deal term.