DS Cohen & Associates
Part Two of Two
Part One argued that a merger’s real risk shows up long after the announcement, when the dust has settled. The disconnect starts to show up in the gap between an employee engagement snapshot and the ongoing daily experience that truly decides whether people stay. In most cases, the promise of blending the best of both cultures is never realistic, because culture is a system of decisions that cannot be split down the middle. We’ve seen this pattern in both the Continental and United integrations and, more recently, in Broadcom’s acquisition of VMware. A new generation of tools now claims to have solved the measurement half of that problem. Here is what they get right, and what they still miss.
Better Instruments, Same Blind Spot
Several AI-powered platforms now claim to make leadership and cultural compatibility measurable before a deal closes, using natural language analysis of executives, standardized culture scoring across dozens of dimensions, and employee emotion tracking conducted in waves during integration rather than once at the start. This is undeniably real progress. It replaces a purely subjective, after-the-fact impression of culture with something closer to a disciplined, repeatable measurement.
There is a more basic problem underlying all three tools, and it exists regardless of how many times they are measured or how confident the final score looks. Before any of these algorithms can produce a number, someone has to decide what counts as evidence of a value in the first place. A programmer or a research team defines which language patterns signal trust, which forced-choice answers indicate strong execution, and which emotional responses count as collaboration. That definition comes from the bias of the people who built the tool. It does not come from either organization going through the deal, and it cannot, because the tool was designed long before this specific pair of companies ever came together.
Two companies can use the identical word for a value and mean two entirely different sets of behaviours by it. One company’s version of accountability might mean a manager stands up in a team meeting and owns a missed target in front of everyone. Another company’s version of accountability might mean the same manager quietly fixes the problem before anyone above them ever hears about it. Both are real, functioning definitions of accountability inside their own organization. A generic model trained to detect one pattern will read the other as weak on accountability, when the truth is only that its version of accountability looks different. Unless the underlying model has been built around the specific behaviours that have evolved inside the two companies being assessed, rather than a general definition of what accountability, trust, or collaboration is supposed to sound like, the output measures the vocabulary of culture and misses its substance. A score built on the wrong definition of a shared word does not just fail to help. It hands leadership a clean, confident number about a culture the algorithm never actually understood, which is a more dangerous position than having no score at all. That blind reliance is dangerous and damaging.
But a step toward is not the same as arriving. Even a tool that checks in with employees a few times during integration is still stringing together a handful of frames. None of these tools reach back far enough to capture the years of accumulated behaviour that actually built the trust, norms, and expectations an organization is carrying into the transaction. That history, not the emotional temperature check taken a few weeks after closing, is what determines whether an organization has the depth to absorb a merger at all.
There is a second risk that comes with better instruments, and it is worth naming plainly. Data borrows authority from its own precision. A culture compatibility scores out of one hundred, or an executive alignment rating expressed to two decimal places, feels more objective than a gut instinct. But that feeling is exactly what makes it dangerous when the number is misread or over trusted. A precise number can just as easily be used to justify a decision that has already been made as to genuinely test that decision. Leadership that treats a score as proof, rather than as one input that still requires judgment, can end up more confident and more wrong than leadership that had no data at all.
None of this is an argument against using these tools. It is an argument for knowing exactly what they measure and what they do not. They can tell you, with more rigour than before, whether two leadership teams are likely to clash and whether emotional strain is building in specific teams during integration. They cannot tell you whether either organization has spent years building an employee experience deep enough to survive the answer to that question, and they cannot tell you whether the definition of trust, accountability, or collaboration built into their model is the definition either company has been living by for years. That deeper measurement still must come from elsewhere, and it must exist long before any of these tools are ever switched on.
The Foundation, Not the Checkpoint
Employee experience cannot be treated as a due diligence checkpoint you complete once and file away. It has to be understood as the ongoing, daily reality that produces engagement in the first place, on both sides of the deal, long before anyone starts talking about a merger.
An organization that has spent years building intentional employee experience practices, where managers are equipped to lead difficult conversations is an organization that can absorb an acquisition. It has the infrastructure to reconcile two different ways of working because it already understands its own way of working at a level deeper than a survey score.
An organization that has never done that work has nothing to reconcile with. It has a number on a slide, and numbers do not tell you what will happen when two very different sets of behaviours meet for the first time in the same building, especially when leadership is telling everyone a comforting story about blending rather than being honest about which culture will prevail.
Do the financial due diligence. Do the leadership assessment. But stop mistaking an engagement survey for an understanding of employee experience and stop promising a blend that the mechanics of organizational decision-making will never actually deliver. One is a photograph. The other is the whole story, and it is the only one that predicts whether your deal is still working in year three.
David S. Cohen is the author of “Selecting the Best: Fostering a Workplace Driven by Values for Lasting Success,” amplifies each of the points of this article using a combination of research and anecdotal stories. The appendix contains sample behavioural interview questions. Selecting the Best is available on Amazon and other online book sellers.
