Supplier Continuity Risk
also called Vendor Viability Risk, Supplier Longevity Test
The chance that a supplier is acquired, repriced or shut down before you are ready to leave - and the design work that turns that event into a migration you schedule rather than one that schedules you.
A team buys a small vendor's product for a capability on the critical path. Integration takes three weeks, the price is a tenth of building, and the record correctly says buying wins. Eighteen months later the vendor is acquired. The product is not switched off; it stops getting the feature the team was waiting for, the renewal price doubles, and their plan is sunsetted with 90 days' notice.
Nothing here is a vendor failure. It is the normal life cycle of a small company, and the part of build-versus-buy that feature matrices never score. Exit cost says what leaving costs; continuity risk is how likely you are to be made to leave, and when.
Why it matters
The decision this changes is not "build or buy" but which suppliers may hold which capabilities. A young vendor is a good choice off the critical path, and a bad one for the thing your product cannot serve a request without.
The failure mode is rarely bankruptcy. It is a roadmap change after an acquisition: the acquirer has a different customer in mind, and the product keeps working while quietly becoming the wrong product - harder to notice than an outage.
Implementation patterns
- Score viability before signing: is this product the company's main line of business, how many customers like us does it have, and what funding stage is it at? A side-project product inside a big company carries startup risk.
- Keep your own copy of the data the vendor holds. A nightly bulk export in a loadable format turns a shutdown into a migration. Check it is genuinely bulk: an API paging at 100 records a second needs about 4.6 days for 40 million records.
- Confine the integration to one module with call sites in single digits, because an exit you can grep for is a rewrite rather than a hunt.
- Put the protections in the contract: notice period for sunsetting, renewal price protection, an assignment clause surviving acquisition, and a documented export right. Then re-measure exit cost annually in engineer-weeks - that number is your negotiating position.
Industry example
The regulator wrote this down for financial firms. Under the EU's Digital Operational Resilience Act, Regulation (EU) 2022/2554, Article 28(8) obliges financial entities to maintain exit strategies for ICT services supporting critical or important functions, explicitly accounting for the provider failing or its service deteriorating, with alternative solutions and transition plans for moving data elsewhere or back in-house. Article 28(3) requires a register of ICT contractual arrangements flagging the critical ones, and the technical standards call for a documented exit plan per critical arrangement, tested periodically.
Without the compliance framing it is an engineering checklist: know which suppliers hold critical capabilities, keep a tested plan for leaving each, and treat deterioration as planned-for.
Failure scenarios
- Acquisition followed by a roadmap change, so the exit happens anyway - at the acquirer's moment.
- Renewal repricing, where a supplier with a 9-month exit cost raises the price sharply and paying is the honest answer, the alternative being a quarter of unplanned engineering.
- An export that does not scale: the format exists, the bulk mode does not, and a rate limit nobody tested sets the migration's pace - alongside a sunset notice shorter than the migration.
Trade-offs
| Choose | Gains | Pays |
|---|---|---|
| Young supplier on the critical path | Price and speed now | A real chance of an unscheduled migration within two to three years |
| Established supplier | Predictable life cycle and a boring renewal | Higher price and slower features |
| Build it | No supplier risk | Permanent ownership - the cost always underestimated |
Continuity risk is not a reason to build. Building is the most expensive hedge available and replaces a supplier risk with a staffing risk nobody prices.
When not to use it
Skip it for commodity capabilities with interchangeable suppliers - object storage, email delivery, SMS, DNS - where interfaces are near-identical and switching is days. Skip it where exit cost is small: one module and your own data means a supplier disappearing is a sprint. The assessment earns its place only on the critical path with an exit measured in months.
Interview question
Q: A team wants to buy a six-month-old startup's product for the core of a new feature. Price and integration effort are both attractive. What do you require before approving it, and what makes you say no?
What a strong answer covers: estimating exit cost and continuity risk separately; whether the capability is on the critical path; a bulk export actually run with its throughput measured; one integration module with counted call sites; contract terms on notice, assignment and renewal; and the refusal case - a months-long exit on the critical path with no export path.
Quick check
Quiz: Which figure decides whether a supplier's youth matters? Probability times exit cost. A two-week exit at a 30% chance is days of expected work; nine months at 5% is weeks, arriving at a moment you do not choose.
Flashcard: What does supplier continuity risk add to exit cost? - Probability and timing. Exit cost says what leaving costs; continuity risk says how likely you are to be made to leave - which is why a young supplier belongs off the critical path.