In July we stopped selling several things we were genuinely good at. Hardware and firmware work, connected sensor projects, 3D scanning and printing, industrial design. We had built real capability in all of it and we had clients who valued it.
Deciding to stop was harder than any technical problem we have solved, and I think about it every time we ask a client to retire something. Here is what it actually involved, because we now ask businesses to do the same thing and it would be dishonest to pretend it is straightforward.
Nothing was failing
That is the part that makes it difficult. If a service line is losing money the decision makes itself. Ours were not. The work was interesting, the clients were happy, and we were better at it than most.
What was wrong was subtler. That work pulled in a different direction from where the rest of the business was going. Every hardware project needed different suppliers, different lead times, different expertise, and a different kind of risk. It sat beside our software work rather than compounding with it, so we were carrying two businesses and getting the benefits of neither.
Meanwhile the work that did compound — connecting systems, automating processes, making operations legible — kept getting better every time we did it, because each engagement made the next one sharper.
The two arguments that nearly stopped us
The first was the capability we had already paid for. Years of learning, equipment, relationships. Walking away from it felt like writing off an investment, and that instinct is powerful.
It is also the sunk cost fallacy wearing a suit. What we spent building that capability was gone either way. The only real question was what it would earn from here compared with the alternative, and once we asked it in those terms the answer stopped being close.
The second was the clients. Real people who came to us for that work, who we would have to tell we were no longer the right partner for it. That one was not a fallacy, it was a genuine cost, and the only decent way through it was to be early and direct rather than quietly deprioritising them until they noticed.
What we did not anticipate
How much faster everything got afterwards. Not immediately, but within a couple of quarters.
Fewer things to be good at meant we could actually be good at them. Decisions got quicker because there were fewer competing directions to weigh. And the reason for the direction became easy to explain internally, which matters more than I expected — when everyone can state what the business is for, you stop having the same debate every month in a different form.
We also found out which capability was load-bearing. Some of what we thought we were keeping for the hardware work turned out to be valuable in the software work, once it was no longer being pulled two ways.
Why I am telling you this
Because it is the same decision we now sit across the table from clients and recommend.
When we map a business and find a product line that does not pay for the complexity it introduces, or a system nobody has the appetite to retire, or a service kept because a good customer once asked for it, we are not describing something abstract. We have made that call about our own business, with our own people, and it was uncomfortable.
It also changed what we think the work is. A partner whose job is to deliver whatever is requested will happily build software for the thing you should be stopping. Being useful over years means occasionally arguing against work, including our own, and you cannot credibly do that if you have never done it to yourself.
What it says about us as a partner
Three things, and you can hold us to all of them.
We will tell you when something is not worth doing, even when it is quotable work for us. We understand that stopping is harder than starting, because the hard part is not the analysis, it is the people, the sunk cost and the customer you have to disappoint. And when we say focus compounds, we are describing something we tested on ourselves at real expense rather than a line from a strategy deck.
An Apex partnership is built on being able to have those conversations for years. It only works if we have earned the right to start them.
