I’m Joe Schiavone. We’ve spent a few posts on inheriting a strategy that wasn’t built to bend, and on a deal landing on a plane that’s still in the air. Today I want to flip the camera to the side of M&A nobody writes about, one I’ve lived. 

I’m talking about the sell-side here, the carve-out, that seat where you’re gripping the wheel while the business is getting split in two around you. That view changes how you read the whole deal. 

Most people talk about carve-outs from the buyer’s side. From that angle, it looks clean: the unit exits, the costs go with it, margins tick up. Easy peasy. The slide builds itself, and I can confidently say the slide is flat-out misleading. 

A Carve-Out Is Never Clean Subtraction 

The mental model almost everyone brings to a separation is subtraction. Pull out a piece, keep the rest humming along like it did yesterday, minus the part you sold. That is not how any of this goes. 

The unit you’re carving out was never sitting in a corner by itself. It was wired into your operation, sharing systems, data, a network, and contracts priced on the two of you being one company, quietly feeding value back to the parent in ways no transition checklist ever captured. Unfortunately, you do not always find those wires during separation planning. In my experience, you tend to find them a week later, when something on your side stops working and the trail runs straight back to the business you just sold. Shocker! 

Separation planning is good at mapping what you hand over. It fails on two points: what was flowing back into you from the unit you’re losing, and what your own deal team already promised in your name. Both are about to become your problem. 

The Void You Didn’t Budget For 

When you divest, you’re removing something that used to contribute to how the business actually runs which needs to be carried forward and built into your post‑divestiture model. Instead, the model stays clean while the reality doesn’t. Your enterprise agreements were priced on combined volume, ERP licenses, cloud commits, telecom, support tiers; all discounted because of your scale together. The unit leaves, your volume drops below the thresholds you negotiated, and pricing quietly resets upward on everything you kept. What looked like a clean cost savings turns into a higher run rate, which you only see when the true‑up hits months later. 

That’s the hole you’re dealing with. The value didn’t walk out with a clean price tag, so now you’re standing there carrying a cost the model never picked up, trying to explain how a deal that was supposed to lift margins just created brand-new work for a team that was already maxed out. 

Picture This (The Commitment I Didn’t Make) 

Now the other half, and this one still makes me laugh, mostly to hide the tears. 

The deal team needed to get to signature. Somewhere in the negotiation the acquirer realized they had nowhere to run the business they were buying. No ERP ready to go, no plan to stand one up, no time. So, a promise got made to keep the deal moving…“they can stay on our ERP for ninety days”, says the deal team.  

OUR ERP! The system running our company, offered up as a transition courtesy. 

Zero ask of me, zero ask of my teams. Nobody walked down the hall to check whether letting the company we’d just sold keep operating inside our crown-jewel for a quarter was technically sane, or survivable on a ninety-day clock. The logic in the room was simple: they’re willing to pay us for baggage we don’t need anymore, so we’ll say yes to whatever they want, and IT will just walk it off. 

IT will be fine. Four words that have funded more lost weekends than any other sentence in the English language, and nobody who says them is ever the one working the weekend. 

So, I owned a problem I was never consulted on: keep an external company, one we’d just sold, running live inside the system that runs us, same instance, same data plane, with a wall down the middle real enough to pass audit and invisible enough not to break either side. The meter running the whole time was a TSA written by people who assumed everything would go to plan.  Anyone stand up and integrate an ERP in 90 days? 

The Wall Hits On Day Thirty, Because It Always Does 

The ninety-day number says ninety because ninety sounded clean in the deal room. What actually happens is the acquirer hits a wall in the first month and the scope creeps. A cutover that was supposed to be done isn’t. A dependency nobody mapped surfaces. And don’t get me started on the circuits. The ones their new environment needed to stand on its own didn’t get ordered until day zero, and carrier lead times being what they are, the wait ate most of the window. They physically could not leave our network on the TSA timeline, so they stayed while everyone pretended this was always the plan. 

Now you’re in the real bind at the strategy level. You could hold the line on the TSA and let them fail. But if they fail, the deal you just closed starts to wobble, value erodes, and the thing booked as a win turns into a dispute. So, you don’t hold the line. You extend, you absorb, you keep providing the exact thing you spent ninety days engineering your way out of. 

That’s the trap. The CIO becomes the last line of defense for a deal everyone else already declared closed, celebrated, and moved on from. You’re gap-filling on your side and propping up theirs on the other. 

Why This Keeps Happening 

None of this is bad luck. It repeats because separation gets treated as acquisition run backwards. Acquisition is additive; bring two things together, find the overlaps, consolidate. Separation untangles dependencies that took years to form. Most enterprises acquire more than they divest, so when a carve-out lands they reach for the integration playbook, run it in reverse, and act surprised when it’s not one size fits all. 

The deal model makes it worse. It counts what leaves and rarely counts what breaks, or what you’re forced to keep providing after. The number stays clean because the messy parts don’t have a line. 

What Actually Has To Change 

The real fix sits earlier in the deal and it’s strategic. 

Map the reverse dependencies before the cut. Separation plans always document what flows out of the parent to the unit. Build the other map too, everything flowing from the unit back into the parent: the shared services it runs, the contracts priced on its volume, the feeds the rest of the business quietly consumes. Your gaps live on that second map, and you want to find it during planning, not during a Tuesday outage with the CFO already on the phone. 

Then, get yourself into the deal room before the commitments are made. The ERP promise, the network arrangement, the ninety-day clock are all technical commitments with deal-level consequences, made by people who can’t see the technical reality. Bring the CIO in as a core input to how the deal is shaped from day one and most of this goes away. If the first time you hear about a TSA commitment is in the signed agreement, the deal is already mispriced. 

Lastly, put the gap-fill and the creep into the deal economics. In reality, TSA extensions happen, and deal-preservation pressure wins every time it meets the TSA’s end date. Price it in while you can. 

The Part Worth Sitting With 

Separation is a leadership call about what you’re actually willing to own and not a modeling exercise. The model draws clean lines because it can’t see the dependencies pulling back into you, doesn’t price the gaps they leave, and has no memory of the commitments made in your name to get the deal across the line. Those show up in your backlog, your cost base, and your timelines. 

That’s how a margin-improvement story turns into eighteen months of stranded cost and a TSA nobody can exit. Same deal, same slide, completely different outcome. The difference is who was in the room before anything was promised, not the asset or the market. Get that right and you capture the value you sold. Miss it, and you spend the next year unwinding decisions you didn’t make but still have to live with. 

I Want To Hear From You 

If a carve-out landed on your desk tomorrow, would you know what flows back into you from the unit you’re about to lose, and what’s already been promised in your name to get the deal done? 

I welcome your perspective. Drop a comment below or connect with me via email: jschiavone@forrester.com.

Share