The Diligence Gap: Why Nearly Half of Dead Deals Now Die in the Data Room
What kills M&A deals has shifted from financing to what buyers find in the data room – and earnings quality is the fastest-rising killer.

Financing used to be what killed a deal. Not anymore.
The clearest read on why deals die comes from the Axial Dead Deal Report, which broke down a set of transactions that collapsed after a signed letter of intent. Two causes now dominate: non-QoE diligence findings at 25.3% and quality-of-earnings EBITDA discrepancies at 21.3%. Together that's roughly 47% of dead deals – nearly half – dying not because the money fell through, but because the buyer looked closely and didn't like what they found.
That's a real shift in where deals break, and if you're doing diligence support for advisors or buyers, it should change where you point your attention.
The crossover: from money problems to scrutiny problems
The trend underneath the headline is the interesting part. In the same data, quality-of-earnings discrepancies roughly doubled over two years – from about 10.6% of failures to 21.3%. Over the same stretch, financing-driven failures fell by half, from 21.3% down to 10.7%. The two lines crossed.
When capital was scarce and expensive, deals died on the financing. As money got easier, buyers spent their attention on scrutiny instead – and the deals that broke increasingly broke because the scrutiny turned something up. The report's own framing captures it: most deal killers are found, not random. The kill is sitting in the data room the whole time. Whether the deal survives depends on whether anyone surfaces it before the close.
That reframes diligence from a box-checking ritual into the actual hinge of the transaction. The earnings quality question – is the EBITDA the deal is priced on the real EBITDA – is now one of the single most common reasons a deal dies. Add-backs that quietly recur every year, customer concentration hidden inside an aggregate, the gap between reported earnings and the cash that actually shows up. None of it is exotic. All of it is findable. The difference between a closed deal and a dead one is often just whether the diligence was sharp enough to find it in time.
Longer reviews, not necessarily better ones
You'd expect all this scrutiny to mean diligence has gotten more thorough. The buyers I'd worry about are the ones who assume it has.
SRS Acquiom's 2026 study found 73% of dealmakers expect diligence to grow more complex, and – tellingly – that technology diligence is now the single most burdensome element of the entire review for 51% of them. Attention is finite. When tech and cyber review expand to fill the calendar, something gets crowded out, and it's often the financial workstream where, per the dead-deal data, the deals are actually dying. More diligence hours spent in the wrong place is not the same as more protection.
The cost of getting this wrong is well documented. One 2026 analysis notes that due-diligence shortcomings are cited in roughly 60% of failed integrations – the problems that surface after the close were usually findable before it. The diligence either found them and priced them, or it didn't and the buyer ate them.
Where the attention should go
The takeaway for anyone supporting a transaction isn't "do more diligence." It's "spend the diligence where deals actually die."
That means treating quality of earnings as a primary workstream, not a line item – pressure-testing add-backs, tracing concentration, reconciling earnings to cash. It means making sure the expanding tech and cyber reviews don't quietly starve the financial work of the senior attention it needs. And it means synthesis: the kill is rarely one dramatic finding, it's a pattern across documents that only shows up when someone capable is actually connecting them, not just collecting them.
The data room has always held the reasons a deal should or shouldn't close. What's changed is that the reasons are now, more often than not, the financial truth hiding in plain sight. The deals that die are mostly the ones where nobody surfaced it in time. The job is making sure you're the one who does.
More from Consulting Operations

Thirty Years of Agile and Software Success Is Still Stuck at 31%
After decades of Agile, cloud, and better tools, software still succeeds about 31% of the time. The bottleneck was never the tooling.

Your Team Doesn't Trust You, and the Data Says You Can't See It
Trust in leadership is at multi-year lows, worst in tech – and the leaders who own it are the last to see it. The cost isn't who quits.

The Verification Gap: Your Developers Don't Trust the Code They're Shipping
Developers don't trust AI-generated code – and ship it anyway. The verification gap is a management problem, not a model problem.
Want help running a sharper practice?
Managed Intelligence handles the research and synthesis behind your client work – a living deliverable kept current, so more of your time goes where your name is on the line.
See Managed Intelligence