I’ve been working on several software development projects recently, and one thing that occurs regularly is a regression.

You introduce new functionality, but in doing so, something that was already working stops working properly. The new functionality can work exactly as intended while the problem appears somewhere else.

It got me thinking that this is not exclusive to software. The same thing can happen in consultancies whenever we introduce something new.

Where regressions occur

Imagine you win a substantial new client.

It is exactly the kind of work you wanted, so naturally it gets your attention.

The time that previously went into marketing begins to move towards delivery. For a while, this seems entirely reasonable. You are busy doing valuable work for a good client.

The effect only becomes visible later, when the project begins to settle and you realise your pipeline is thinner than you expected.

Winning the client created the result you wanted, but it also changed the conditions that had been helping you win the next one.

Why regressions are difficult to see

The intended effect of a change is usually the part we watch most closely. Our attention stays with the result we hoped to create, while any effect elsewhere in the consultancy is easier to miss.

The regression often appears later and in a different part of the business. By the time something starts to feel wrong, the change that set it in motion may no longer seem connected to it.

Protecting what already works

Software developers expect regressions, so when they change one part of a system, they check that the important things which worked before still work afterwards.

Perhaps the useful question for a consultancy is simply:

What might become less reliable while you are making the new thing work?