Scope Creep in M&A Carve-Outs: The $5M Problem Nobody Budgets For
10 September 2026 · 11 min read
By the MeridianCogent team — built from working with integration offices, separation leaders, and PE operating partners on M&A execution. This perspective is drawn from published carve-out and TSA research across the industry.
There's a recurring conversation in post-close carve-out reviews: "We thought we had mapped all the shared systems. By month six, we'd added $5 million to the TSA bill." This isn't a diligence failure — it's a structural inevitability of how carve-outs work. A complete entanglement inventory is unknowable until you actually try to separate the business.
Where the $5M actually comes from
Rather than treat that figure as an abstract headline number, here's a representative breakdown of how it accumulates on a mid-sized carve-out (illustrative composite, built from the cost categories that recur across carve-out post-mortems):
| Source | Amount | Why It Wasn't Caught Pre-Close |
|---|---|---|
| Software license co-termination penalties | $1.5M | Enterprise licenses priced per parent entity; separating triggers renegotiation at standalone (unfavorable) rates, plus early-termination penalties on multi-year agreements |
| Legacy ERP dependency rebuild | $2.0M | "Standalone" ERP instance had undocumented integrations into parent consolidation, pricing, and vendor systems — discovered only when the integration was actually cut |
| Stranded IT overhead absorption | $1.5M | Data center capacity, network agreements, and help desk staffing sized for combined headcount don't shrink proportionally when the divested unit leaves |
| Total | $5.0M |
None of these three categories are unusual individually — they show up in some form in nearly every carve-out. What makes the total surprising is that they're rarely modeled together, pre-signing, as a single expected cost.
Why each category hides so well
License co-termination: Enterprise vendors price per entity or organization ID, and don't renegotiate downward. The buyer inherits the parent's pricing leverage as a small division — then negotiates alone, against a deadline, for a fraction of the parent's volume. A €695k/year combined license spend at parent-scale pricing can become €1.2M+/year standalone — a 70%+ uplift that only becomes visible once you actually request standalone pricing.
ERP dependency rebuild: A system can look standalone while quietly depending on a downstream process elsewhere. You discover this only when you try to cut the connection — a "standalone" CRM instance turning out to have dozens of live integrations into parent financial consolidation, pricing engines, or vendor portals is a common pattern, not an edge case.
Stranded overhead: The seller's IT budget doesn't shrink just because a division leaves. Data center space, multi-year enterprise agreements, and help desk staffing are fixed costs in the short term — so the seller absorbs a real cost increase per remaining unit of business, one that's rarely modeled into the divestiture economics upfront.
The scope creep mechanism in the TSA itself
Contract language like "ERP services including payroll, general ledger, accounts payable, and such other services as necessary for continuity" is a blank check both sides sign knowingly. At month two, a new service gets "discovered" and added at extension pricing. By month six, the TSA can cover twice the originally scoped services.
The defense that actually works
1. Divide the TSA into named services plus a bounded discovery window.
| Service Category | Priced | Duration | Discovery Window |
|---|---|---|---|
| Core (named services) | €135k/month | 12 months | N/A |
| Discovery (unscoped) | €40k/month cap | 3 months | 90 days post-close |
A discovery window with a hard expiry (commonly 90 days) forces new-service pricing and cutover decisions quickly, rather than letting undefined scope accumulate indefinitely.
2. Run a discovery audit during pre-signing, not after. The single highest-leverage defense is a dedicated entanglement-mapping exercise before the deal signs — tracing upstream and downstream system dependencies, not just listing named applications. This is exactly the readiness assessment covered in The Separation Readiness Paradox: a 4–8 week effort, done before marketing, that catches a meaningful share of these entanglements while there's still room to price them into the deal.
3. Own the standalone cost model. Build the real bottoms-up cost of running each function independently — commonly 40–60% above parent allocations — and reconcile it before close, not mid-TSA.
4. Price extensions with escalation. As covered in The Real Cost of a Late TSA Exit, a 15–25% monthly step-up on extensions is what keeps "temporary" services temporary.
5. Freeze scope 30 days before close, after running at least two full mock cutovers. Most Day 1 failures trace back to untested dependencies and last-minute scope changes — see What 'Day 1 Ready' Actually Means for the acceptance test that catches this before go-live.
The math of delayed discovery
A $5M discovery landing at month two of a 12-month TSA is a $5M bill. The same discovery at month 10 adds extension pricing (2 months at 15–25% uplift) and accelerated build costs to compress the same work into less time — pushing the real total closer to $6–6.75M. The cost isn't the discovery itself; it's how much of the TSA term is already consumed when it surfaces.
Lessons learned
- Budget discovery explicitly, not as a vague "10% contingency." A named discovery line (commonly $500k–1.5M for a mid-market deal) is more credible to a buyer than false certainty in the base scope.
- The 90-day discovery window creates real discipline. Nobody defers a hard decision to month 11 if they know it means permanent licensing costs, not temporary TSA support.
- Systems integration is the biggest risk category, not individual systems. Map upstream and downstream dependencies before calling anything "ready to separate."
- Stranded costs land on the seller, not the buyer. Model this explicitly as part of divestiture economics, not as a post-close surprise write-down.
Sources: FTI Consulting (carve-out blind spots and value destruction research); BD Emerson (IT separation cost analysis); Kirkland & Ellis (carve-out entanglement documentation); EY (operational separation guidance); PortMux (carve-out data readiness research).