The Software Hypocrite

StoryChapter 1

I'm a software hypocrite.

30 years telling companies to buy, not build. AI changed my mind.

For most of my career, I gave clients the same advice: buy software, don't build it.

I was not trying to be memorable. Building was expensive. It was slow. It created systems that were hard to maintain and easy to regret. Buying a proven product, and configuring it carefully, was the responsible recommendation. I gave that advice as a consultant, as a CIO, and as someone who had founded and led a software consulting firm. I still believe a lot of it.

I also think I was incomplete. And I have a further admission, which is harder than changing a piece of advice.

I didn't just tell companies to buy the software. I helped put it in.

I founded a firm that implemented packaged systems. I spent years as the person companies called when they needed a CRM, a platform, or a large product installed and made to work. In that sense I helped set the fire. A lot of businesses now run on stacked, rented layers that are hard to change and expensive to keep — a CRM that never quite fit, a commerce platform that grew heavier than the operation, a managed website that costs a great deal and still does not do what leadership needs. I know those fires because I spent a long time lighting some of them, with the best advice I had at the time.

That is also why I can help put the right ones out. Taking a layer out of a running business is not a demonstration. If you do it without knowing what fails, you can take the house with it.

What changed is not my view of expertise. Experienced people are still the point. What changed is the leverage those people can now have. AI can make a strong team 10 to 100 times more effective. Work that used to mean a large budget, a long project, and a vendor relationship can now be done in weeks, by people who already understand the business. That leverage is useful. It is not the product. I am not asking anyone to trust a tool. I am asking them to trust an experienced team that finally has enough leverage to build what the business actually needs.

That is why I have become willing to say something I would not have said for the first 30 years of my career: throw away your software. Build, don't buy.

I need to be precise, because that line is easy to misuse. I am not telling anyone to rip out every system they have. I am not interested in standing up a new site or a new tool as cheaply and as quickly as possible. Plenty of teams can now do that in a weekend. That is not the same as something you can run a business on.

The useful question is more disciplined than "buy or build." For each system, I ask: should we keep it, simplify it, build around it, or retire it? Some software should stay. The common processes often should still be bought. The parts that make a company different, and the layers that have become expensive to rent and hard to change, are where building belongs — carefully, and with a plan for failure.

If you change a system without knowing what happens when it fails, you are taking a risk the business did not ask for. Every serious change should start by finding the points of failure and defining a fallback for each one. I call that a Resilience Check. The business should keep running while the work is done.

If this sounds like a conversation you have already been having internally, start small. Bring me one workflow your team keeps working around. That is usually enough to see whether the next step is a plan, a teardown, or simply leaving a working system alone.

The advice did not get sloppy. The math changed.

Video placeholder

[PLACEHOLDER: 60–90 second video, once recorded]

In short

  1. I spent 30 years telling clients to buy software, not build it. I meant it.
  2. I also helped set the fire. I built a career implementing packaged software.
  3. AI changed the math. It makes an experienced team 10 to 100 times more effective. It does not replace them.
  4. Throw away your software. Build, don't buy. That does not mean ripping out everything you have.
  5. For each system: keep it, simplify it, build around it, or retire it.
  6. If you change a system without a fallback, you are taking a risk the business did not ask for.
  7. Bring me one workflow your team keeps working around.

Bring me one workflow your team keeps working around.

Start a conversation