What technology due diligence does not tell you

Technology due diligence is good at the questions it was designed for. Is the architecture sound. Are the licences in order. Was there an incident nobody disclosed. Is the code base an asset or a liability. A competent provider answers all of those, and the answers go into the model.

None of them tell you what the estate costs to run.

That is not a criticism of diligence. It is a different question, asked at a different time, by people who will not be there afterwards. Diligence describes the thing you are buying. Operating it is about who turns up on Monday, what they know, and how much of it is written down.

Gap one: the person nobody mapped

Most mid-market businesses have one individual who understands the part of the estate everything else depends on. The historian. The integration between the order system and the plant floor. The reporting finance closes the month on. That person is rarely the IT director and often is not in IT at all. Diligence meets the management team. It does not meet them.

You find out who they are the week they resign.

Gap two: the work that was deferred rather than done

Every estate carries a queue of things postponed for a good reason at the time.

• A version upgrade held back because the vendor's next release was due.
• A firewall rule that was meant to be temporary.
• A server left on an operating system that stopped receiving patches, because the application on it was never certified for anything newer.

Individually each was a sensible call. Collectively they are why a project that should take a quarter takes three. Diligence sees the estate as it stands. It does not see the queue behind it, because the queue is not a document. It lives in the heads of the people who made those calls.

Gap three: the TSA on a carve-out

A transitional services agreement is priced as an interim arrangement and treated as a bridge. It behaves like a meter. The seller has no reason to make the exit easy and every reason to keep the clock running, and the exit is invariably harder than the schedule assumed, because the systems were never designed to come apart.

A TSA that overruns is not a line item. It is the most common single way a technology plan loses its first year.

Three questions before the model is signed

• Who would have to be in the room to restore the business after a bad Friday? If the answer is one name, you have found a risk diligence did not price.
• What has been deferred? Ask the engineers rather than the executives: what would they fix first if given a month and nobody complained.
• On a carve-out, what does the exit look like? The cost is knowable. The exit is the part that overruns.

None of this needs a longer diligence. It needs a different conversation, held with different people, about the year after the deal rather than the day of it. The businesses that get it right are the ones with somebody who had operated an estate like this before, asking the questions an operator asks.