Most data work gets sold as risk reduction. Clean it up so diligence does not blow up the deal. Fix it so the dashboard stops lying. That framing is true, and it is also why data work keeps losing the budget fight. Risk reduction is a cost. It competes with everything else that promises a return.
So here is the case made the other way. As a return. Because the numbers are now good enough that you do not need to argue from fear.
A data foundation pays for itself twice. Once in operating ROIC during the hold, and again in the multiple at exit. The firms skipping it are leaving both on the table, and they are doing it at a moment when buyers have started to price the gap and most sellers have not noticed.
The operating return is not small
Start with the hold period, because that is where the cash actually shows up.
BCG’s work on AI puts the number plainly. Companies pursuing systematic AI deliver roughly double the ROIC of their peers. Not a few points of margin. Double the return on invested capital. That is the kind of spread that changes a fund’s outcome on a single portfolio company.
The word doing the work in that sentence is systematic. Not a pilot. Not a data science hire with a mandate and no plumbing. Systematic means the company can actually feed AI the inputs it needs, repeatedly, across the business, without three people pulling a spreadsheet together every time. Systematic is a data foundation by another name.
The same research separates digital from digital plus AI. Digital transformation on its own returns something on the order of 15 to 20 percent. Digital plus AI runs materially higher, on the order of 30 to 35 percent. The AI layer roughly doubles the return of the digital layer underneath it.
But the AI layer only earns that premium if the layer underneath it holds. You cannot bolt a 30 percent return onto a foundation that produces a different revenue number depending on who you ask. The model is only as good as the inputs, and the inputs are the data foundation.
This is the mechanism I keep coming back to. AI readiness starts with data readiness. The AI does not fail because the AI was wrong. It fails because the data underneath it was never ready, and the return that was supposed to double quietly never arrives.
Why the foundation is the multiplier, not a line item
Look at how those numbers stack and you can see why data sits in a different category from the rest of the value creation plan.
Digital alone returns 15 to 20 percent. Add AI on top and it goes to 30 to 35 percent. The systematic version doubles ROIC against peers. Each of those steps assumes the one below it is solid. The data foundation is the floor every other return stands on.
That is what makes it pay for itself in the hold rather than just cost money. It is not one initiative competing with pricing, procurement, and retention for budget. It is the thing that makes pricing, procurement, and retention actually work, because every one of those levers runs on the company being able to see itself clearly.
A pricing initiative needs unit economics at the customer level. A retention program needs clean customer data joined across CRM, billing, and product. A procurement consolidation needs spend visibility that does not live in miscategorized GL accounts. Each of those returns is real. Each of them is gated by the same foundation. Fix the foundation once and you unlock the whole set at the same time. Skip it and you fund the same initiatives anyway, and they underdeliver one at a time.
That is the operating half of the case. The foundation pays for itself before you ever get to a sale.
The exit return is already being priced
Now the second payment, the one at exit, and this is where the timing has shifted.
Buyers used to treat digital maturity as a nice-to-have. A line in the management presentation. Something to improve after close. That is no longer how the better buyers behave.
BCG’s data shows around 40 percent of buyers have already taken a valuation haircut of 5 percent or more on an acquisition because of low digital maturity. That is not a future risk. It has already happened, on closed deals, at a scale that shows up in the data. Two in five buyers have repriced for this.
Sit with the size of that. A 5 percent haircut on enterprise value is not a rounding error. On a business changing hands at a serious multiple, it is the difference between a top-quartile exit and an average one, and it is being taken specifically because the digital and data foundation was not there. I have watched the mechanism up close. A buyer asks for revenue by customer by product by month, the seller takes three weeks and produces numbers that do not reconcile, and the offer comes back lighter. The discount is not punitive. It is the buyer pricing the work they now have to do, plus the risk that the numbers they were shown were never trustworthy.
This is the bifurcation I have written about before. Record multiples make data your edge, not your debt. The same asset trades at a premium or a discount depending on whether the buyer can trust what it produces. The premium and the haircut are the same coin seen from two sides.
The gap most firms are sitting in
Here is the part that should change how you act, because it is a gap, and gaps are where return lives.
The same BCG data shows that while around 40 percent of buyers have already taken a haircut for low digital maturity, only a minority formally use digital readiness as a go/no-go gate in diligence.
Read those two numbers together. Buyers are punishing the lack of a data foundation roughly twice as often as they are formally checking for it. The discount is being applied more widely than the discipline that would catch it. The market has started pricing the gap before the market has built the process to measure it.
For a seller, that is the cleanest possible case for doing the work early. The haircut is already common. The formal gate is not. So the buyers who do not have a gate are still finding the weakness, just later and more crudely, and pricing it into the offer when you have the least leverage to push back. You do not get to argue your way out of a number that surfaces in the final weeks of diligence.
For a buyer, the gap is an edge. If only a minority of firms formally grade digital readiness at the gate, then grading it well is a source of advantage on both sides of the trade. You avoid overpaying for a foundation that is not there, and you find the underpriced asset where the foundation is better than the seller knew how to show.
Why skipping it loses twice
Put the two halves together and the cost of skipping the foundation is not one missed return. It is two.
In the hold, you forgo the operating ROIC. The AI initiatives stall on data that cannot support them. The pricing and retention and procurement levers underdeliver because they run on a company that cannot see itself. The double-ROIC outcome was available and you funded everything except the floor it stood on.
At exit, you take the haircut. The buyer prices the missing foundation into the offer, and because most buyers do not run a formal gate, you do not even get clean warning. The discount arrives as a softer number rather than a checklist item you could have prepared for.
The firm that does the work captures both. Better operating returns through the hold, and a cleaner number at the sale because the asset produces trustworthy data on demand. The firm that skips it pays twice for the privilege of saving the line item.
None of this requires a multi-year transformation, which is the usual objection. The point is not to boil the ocean. It is to make the company’s core numbers trustworthy and repeatable, so the foundation is in place early enough to compound through the hold and already standing when the buyer arrives. That discipline is what data governance that raises valuation is actually about. Not control for its own sake. The conditions under which the returns above are even possible.
Where to start
The honest first move is to find out which side of the line you are on. Most teams assume their foundation is better than it is, right up until a buyer’s request exposes it.
If you want a fast read on whether your portfolio company can actually support the AI and digital returns its plan assumes, the AI Readiness assessment walks through the foundation questions a buyer’s diligence team will eventually ask anyway. Better to answer them in year two on your terms than in the final weeks on theirs.
The math is not subtle. The operating return is roughly double. The exit haircut is already common. The gate that would catch it mostly does not exist yet. The firms that move while that gap is open will compound the advantage across every deal in the portfolio. The ones that wait will keep paying for the foundation twice and calling it a cost.