Rescue vs Rebuild, and Hire vs Partner: Salvaging a Stalled €7M Programme
A European clinical laboratory network (34 labs, 2,800 staff) was three years and ~€7M into replacing its laboratory information management system — a platform that had been "85% complete" for eleven months, with two unstable pilot labs and legacy support ending in fourteen months. Envion’s five-week assessment found the truth between the two factions: the domain model and data layer were genuinely good (~€2.6M of embedded value), the analyser integration layer was not viable, and €2.9M was genuinely sunk. The recommendation: partial rescue, not rebuild — and hire an internal engineering core as a precondition. Go-live across all 34 labs landed in 11 months, ahead of the legacy end-of-life, for €2.4M.
The challenge
Three years and roughly €7M into replacing its laboratory information management system, the client had a platform that was, in the delivery vendor's assessment, "85% complete." It had been 85% complete for eleven months.
Two pilot labs were running on it. Both reported instability — result-reporting delays, intermittent failures in the analyser integration layer, and a regulatory validation package that had twice failed internal audit. The remaining 32 labs were on the legacy system, which the original vendor had end-of-lifed with support ending in fourteen months.
The board was split. The COO wanted to write it off and restart with a different partner. The CFO could not countenance a €7M write-down. Neither had an independent technical view — the only assessment available was the delivery vendor's own, and it said everything was nearly fine. And the client had no internal engineering function at all: every line of code had been written by an outsourced team.
Decision path
Envion ran five weeks of technical due diligence with a brief to answer three questions: what actually exists, what is salvageable, and what would each path cost from here. The situation was neither of the two stories being told — the system was not 85% complete, and it was not worthless.
The domain model and data layer were genuinely good: specimen lifecycle, chain of custody, result versioning, audit trail — modelled carefully by someone who understood laboratory work, the hardest and most expensive part of a LIMS to get right, roughly €2.6M of embedded value. The analyser integration layer was not viable: eleven bespoke instrument integrations with no shared abstraction, no retry semantics, and error handling that in three cases silently dropped results — classified as a patient-safety issue and escalated on day nine rather than held for the final report. The frontend was roughly 40% complete, not 85%, built against workflows that two subsequent regulatory changes had invalidated. And the validation package was unrecoverable — assembled retrospectively rather than generated from the development process, which is why it kept failing audit.
Envion contribution
Envion measured progress against acceptance criteria rather than the vendor's own task list — the remaining work had been consistently scoped as "integration and hardening," the category that absorbs unbounded time when the underlying design is wrong.
Envion also assessed the client's ability to own the outcome and found the central structural problem: with zero internal engineering capability, the client had no way to evaluate any vendor's claims — this one or a replacement. Restarting with a new partner would reproduce the same blindness with a fourteen-month deadline instead of a three-year runway.
And Envion told the CFO the thing he did not want to hear: roughly €2.9M of the €7M was genuinely sunk and should be written off — continuing to protect that number was the reason the programme had run eleven months past "85% complete."
Delivery
The recommendation: partial rescue, not rebuild — and build an internal core team as a precondition. Keep the domain model, data layer and audit infrastructure — restarting would spend €2.6M re-solving a solved problem, and the deadline made that impossible regardless. Rebuild the analyser integration layer against a single hardened abstraction, treating the three silent-failure integrations as a priority incident on the pilot labs immediately, ahead of any programme decision. Complete the frontend against re-validated current workflows, scope cut to the 60% of functionality the labs actually used — established from legacy system usage data, not requirements workshops. Regenerate the validation package from the development process going forward, with a computerised-systems-validation specialist embedded from week one rather than consulted at the end.
Hire, then partner — in that order: six permanent engineers including an internal technical lead to own the platform, with the existing vendor retained in a reduced, supervised delivery role rather than fired — the vendor held three years of context, and the failure was one of governance and architecture ownership, not individual competence. Envion provided interim technical leadership for four months while the internal lead was recruited and onboarded — explicitly time-boxed, with the exit date in the engagement contract.
Outcome and evidence
Go-live across all 34 labs landed 11 months after the assessment — ahead of the legacy end-of-life. Cost to complete came to €2.4M, within the revised budget; the modelled full-rebuild alternative was €6.1M and 26 months, past the EOL date. The write-off was named at €2.9M of the €7M. The three silent result-drop defects in production were resolved in week 3. The validation audit passed on the first attempt. Internal engineering capability went from zero to six permanent engineers owning the platform in-house. Analyser integrations went from 11 unstable bespoke connections to 24 on a shared abstraction.
The lessons generalize: assess by layer, not by percentage — "85% complete" is not a measurement. Sunk cost is the mechanism, not the excuse — protecting the €7M was actively producing the overrun. If you can't evaluate your vendor, that's the first problem to solve. Firing the vendor is rarely the fix — context is expensive. And insist your interim leadership has an exit date — anyone providing technical leadership during a rescue should be contractually committed to making themselves unnecessary.
| Metric | Position at assessment | Outcome |
|---|---|---|
| Go-live across 34 labs | No credible date | 11 months, ahead of legacy EOL |
| Cost to complete | Unknown (vendor: "3 months") | €2.4M, within revised budget |
| Cost of full rebuild (modelled alternative) | — | €6.1M, 26 months — past EOL |
| Value written off | — | €2.9M (of €7M) |
| Silent result-drop defects | 3 in production | Resolved in week 3 |
| Validation audit | Failed twice | Passed first attempt |
| Internal engineering capability | 0 | 6 permanent, platform owned in-house |
| Analyser integrations live | 11 (unstable) | 24 (shared abstraction) |
Client feedback
What the client says about this engagement
“We had two factions and no facts. Our vendor said three months, our COO said burn it down, and nobody in the building could referee because we didn't employ a single engineer. Envion's assessment was the first document any of us trusted, largely because it disagreed with both camps.
The hardest conversation was the write-off — I had been protecting seven million euros and Envion showed me that protecting it was exactly what had cost us eleven months. The advice I'd underline for anyone in our position is the hiring recommendation. They told us not to outsource the ownership again, even to them. They put an exit date in their own contract.”
Evidence gate. This page publishes only what Envion's project records and client disclosure permissions support. Outcomes are added once verified against a baseline, a measurement period, and an approved source.
FAQ
Questions about this case
Facing a similar challenge?
Facing a rescue-versus-rebuild decision with two factions and no facts? Envion delivers an independent technical assessment in weeks — while the decision is still cheap.
Discuss a Similar ChallengeKeep exploring
Similar case studies
Executive Technology Leadership
Support for high-stakes product and AI decisions
Bring senior technology leadership into the business when the roadmap is unclear, delivery is at risk, an AI initiative needs stronger ownership, or the company needs an experienced technical voice before hiring a permanent CTO.
Discuss Interim CTO SupportCore responsibilities
- Align product and technology priorities with business goals and measurable outcomes.
- Review architecture, delivery risks, data foundations, security needs, and AI readiness.
- Lead internal teams and external partners through a practical execution plan.
- Clarify team structure, ownership, decision rights, and delivery cadence.
- Support investor, board, partner, and due-diligence conversations with credible technical judgment.
New experience
Prompt-to-Page — try it right here
Describe the landing page you want, in your own words. We turn it into a finished page and email you a private link in 5–10 minutes — no briefs, no calls, $0 to see the result.
- Describe what you want to create.
- We structure, write, and compose the page.
- You receive a private link when it is ready.
Start with a sentence — the interactive builder takes it from there.
Generate My PageSafe, respectful content only. No obligation.