When a carve-out lands on your desk, the technology question is not academic. The asset you are buying runs on systems it does not own. Its CRM sits inside the parent’s Salesforce org, its financials run through a shared ERP, its email and identity live in the seller’s tenant, and its data warehouse mixes three business units in one schema. Every one of those dependencies becomes a stand-up cost, a Transition Service Agreement line item, and a risk to your first-year EBITDA. Carve-out technology due diligence is the work that tells you, before close, how much it will cost to make this business run on its own, how long that will take, and where the forecast can break.
This guide is written for the operating partner or portfolio executive who has budget and a signed LOI, and who needs to decide what to fund, what to defer, and what to price into the deal. It does not define standard terms. It walks the sequence of decisions and the evidence that supports each one.
1. Frame the carve-out as a standalone cost, not an audit
The trap in early diligence is treating the technology review as a health check on the seller’s stack. That is the wrong question. The seller’s IT is fine for the seller. Your question is narrower and more expensive to answer: what does it take to run this asset without the parent, and what does that number do to the model?
Every carve-out carries a stranded-cost problem and a rebuild-cost problem at the same time. The parent absorbed shared services the standalone entity now has to fund. Independent research from firms like Bain’s annual private equity report and BCG’s principal investors practice has repeatedly flagged separation complexity as a driver of value leakage in carve-outs. The diligence output you actually need is a costed separation plan, not a risk register with no dollar signs attached.
Scope the review around three deliverables: a standalone technology run-rate, a one-time separation cost, and a dated timeline for exiting the TSA. If a vendor or internal team gives you findings without those three numbers, the work is incomplete.
2. Map every dependency on the parent
Start with a full inventory of systems the target uses and, for each one, answer a single question: does it come with the asset, or does it stay with the seller? Most of the important answers are “it stays.”
The dependency categories that actually cost money
- Systems of record. CRM, ERP, HRIS, and the data warehouse. If any of these is a shared parent instance, you are looking at a migration, not a login change.
- Identity and email. Directory, single sign-on, and mailboxes almost always live in the seller’s tenant. This is usually the first Day 1 blocker.
- Contracts and licensing. Enterprise software agreements signed by the parent do not transfer automatically. Expect to re-license at standalone (higher) pricing.
- Custom code and integrations. Internal tools, reports, and integrations built by the parent’s team walk out the door with that team.
Rank each dependency by two axes: how hard it is to replace, and how much it blocks Day 1 operations. That grid is the spine of your technology due diligence and the input to every downstream number. Legal treatment of separated assets and disclosures is worth watching alongside this work, and the Harvard Law School Forum on Corporate Governance is a useful source for how deal governance and transition arrangements are being structured.
3. Turn the TSA into a priced, dated exit plan
The Transition Service Agreement is where good intentions go to become recurring cost. The seller provides systems and services for a period, you pay a markup, and every month you stay on the TSA is a month you are funding someone else’s infrastructure and running on someone else’s roadmap.
Diligence should price the TSA two ways: the cost of the services as written, and the cost of exiting each service. Cheap TSA line items are often the ones with the most expensive exits, because the underlying system is the hardest to rebuild. Read the TSA term against your separation timeline, not the other way around. If the CRM migration takes nine months and the TSA covers it for six, you have a gap that turns into an emergency spend.
Set explicit exit dates for each service and assign an owner. The mechanics of scoping, pricing, and exiting these arrangements deserve their own workstream, which is why Transition Service Agreement consulting is treated as a distinct capability. Get the exit sequence wrong and the whole first 100 days gets consumed by dependency firefighting instead of value creation.
4. Cost the standalone stack and the migration
Now build the two numbers that drive the decision. The first is the standalone run-rate: what it costs annually to run this business on its own systems once the TSA is gone. That includes standalone licensing (usually more expensive than the parent’s negotiated rate), an owned data and identity layer, and the people to run it.
The second is the one-time separation cost: migrations, re-licensing, integration rebuilds, and the temporary staffing to do the work. This is the number most likely to be optimistic in a seller-provided model. Deal and data providers such as PitchBook and S&P Global Market Intelligence track carve-out activity and pricing pressure, and the pattern is consistent: separation costs are systematically underestimated at signing. Build in a contingency and label it.
Classify every cost by certainty
Not every number in the model deserves the same confidence. Separate what is realized (already contracted), run-rate (steady-state standalone), forecast (your best estimate of migration cost), and risk-avoided (spend you dodge by acting early). A model that blends these into one line hides the exact places the forecast can break. McKinsey’s private capital research and Harvard Business Review’s work on mergers and acquisitions both point to disciplined value tracking as the differentiator between carve-outs that hit plan and those that quietly erode it.
5. Judge the team doing the diligence
Whether the work is internal or bought, the standard is the same. A credible carve-out technology diligence output ties every finding to a cost, a date, and an owner. Judge the provider on their deliverables, not their deck.
- Do they hand you a costed separation plan, or a list of observations? Observations without dollars are not diligence.
- Do they read the TSA line by line against the migration timeline, or treat the TSA as a legal problem someone else owns?
- Do they name Day 1 blockers explicitly, with a plan to clear each one before close where possible?
- Can they staff the execution, not just the assessment? The gap between a good diligence report and a working standalone stack is people who can do the migration.
The same buyer discipline applies across the revenue stack. The frameworks in how to judge a marketing operations consultant in private equity and what to decide and judge in outsourced RevOps for a portfolio company carry over directly: define the decision, demand the evidence, and refuse activity dressed up as outcome. If the go-to-market systems are part of the carve-out, buying outsourced RevOps without inheriting a mess becomes part of the same separation plan.
6. The pre-close checklist
Before the TSA is signed, the technology diligence should have answered all of the following. If any answer is missing, the number in your model is a guess.
- Every system the target uses is inventoried and tagged owned or seller-retained.
- Each seller-retained system has a replacement path, a cost, and a target exit date.
- Day 1 blockers (identity, email, access) have a plan that clears before or at close.
- The TSA term is longer than the longest critical migration, with margin.
- Standalone run-rate is modeled and compared to the parent’s absorbed cost.
- One-time separation cost carries a labeled contingency.
- Each workstream has a named owner accountable for the exit date.
- Execution capacity is lined up, not assumed.
Carve-out technology due diligence done this way turns a source of forecast risk into a schedule you can run. The alternative is discovering the real separation cost after close, when it is a variance instead of a line item. Industry trackers like Buyouts, Private Equity International, and Preqin continue to report carve-outs as a large and growing share of deal flow, which means the teams that separate assets cleanly compound the advantage across a portfolio.
For the reporting and disclosure standards that sit around this work, AICPA & CIMA and the U.S. Securities and Exchange Commission are the primary references, though the technology separation plan itself is an operating and commercial exercise, not an accounting one.
The software and specialist help that carries an IT separation
Once the dependency map is built, the next question is practical: what tooling and outside help the separation actually requires, and who pays for it before close. The honest answer is that no single platform runs a carve-out. The work splits across discovery, migration, and the program management that keeps the TSA exit on schedule, and each of those draws on a different kind of tool.
For discovery, the target’s existing asset and configuration data is usually incomplete, because the parent managed it centrally and never tracked the business unit as a separate entity. Discovery tooling that scans the environment and produces a defensible inventory of applications, servers, identities, and data stores is worth funding early, because the dependency map and the standalone run-rate both depend on it being right. A spreadsheet built from interviews will miss the dependencies that cost the most to find after close.
For migration, the choice is driven by what is moving. Moving a CRM out of the parent’s Salesforce org is a data extraction and re-provisioning job, not the same work as standing up a separate identity tenant or splitting a shared ERP. Each system tends to need its own migration approach and often its own vendor, so an operating partner should resist any pitch that promises one platform for the whole separation.
The part that is routinely underbought is program management. The TSA exit plan is a dated set of dependencies, and the value of specialist carve-out IT services is mostly in keeping that plan priced, sequenced, and actually exited on time rather than extended at the seller’s rates. Judge a provider on whether it has run separations to a hard TSA deadline, not on the length of its tool list.
If you are pricing a carve-out now and need the separation costed and the TSA exit dated before you sign, route the work through the DevriX private equity practice to move from diligence findings into an executable standalone plan.