I left a client meeting last week still thinking about a question that used to have an easy answer.
Build or buy?
For most of the last twenty years, "buy" won by default — and for good reason. Building your own software meant hiring a development team, waiting six to twelve months, and then owning that thing forever: the bugs, the updates, the maintenance, the person who wrote it leaving and taking half the knowledge with them. So instead, you rented a seat on a big platform. You paid per user, per month, and you quietly accepted that eighty percent of the features weren't built for you. They were built for the millions of other companies paying for the same seat.
That was the trade. Buy meant compromise, but building was too slow and too expensive to justify. So you compromised.
Here's what changed: the cost and time to build have collapsed.
The enormous software-as-a-service companies you know by name were built the way they were because they had to be. To justify the engineering investment, they needed to serve a million users — which meant the product had to be general enough to fit a million different businesses. That generality is a feature when you're the vendor. It's a tax when you're the customer, because you're paying for a tool shaped for the average of everyone, not for the specific way your business actually works.
AI has quietly removed that constraint. Software can now be developed quickly and cheaply enough to serve exactly one company. One team. Sometimes one person, for one very specific job. The economics that used to force you to buy a generic seat now let you build something that fits like it was tailored — because it was.
We built exactly this for a client recently — a firm managing a portfolio of commercial and residential properties.
On paper it was only a few dozen properties. But behind those properties were thousands of individual data points. Is each one insured? What did they pay for it? Are the taxes current? Who's on the property right now, and what's the use case — commercial lease, a home, a rental unit in a larger building? Every property carried its own web of answers.
And all of it lived where this kind of data usually lives: a Google Drive folder full of Word documents, PDFs, and Excel sheets. Siloed. Unstructured. Technically "stored," but not actually usable without a human sitting down and digging.
No off-the-shelf platform was ever going to model that exact mess. Five years ago, building something custom for it would have been too expensive to justify. So the firm did what everyone does — they lived with the folder.
We built them a single application for that one use case. It aggregates every data point into one visual platform, gives them a clear picture of the whole portfolio at a glance, and — because people are still used to looking at data a certain way — exports straight to Excel for dashboards or deeper analysis.
The time savings for the firm were obvious. But that wasn't the real return.
The real return showed up in everyone else's invoices. Their CPAs, their tax strategists, their attorneys used to spend hours combing through unstructured files to find what they needed — and every one of those hours was billable. Now they open a dashboard or an Excel export and the answer is right there. The custom software didn't just save the firm time. It shrank the expensive human hours that surrounded it. It paid for itself in billable hours not spent.
This is the part I want you to sit with, because it's easy to hear all this as "build good, buy bad." That's not it.
Buying is still the right call for plenty of things. The point is that the decision tree itself has changed. There's a whole new branch now — problems that were too specific to ever be worth a custom build, that are suddenly very worth it. If you've been mentally filing certain messy, one-off, "that's just how we do it here" problems under nothing on the market fits, so we'll live with it — that file needs reopening.
The math you did two or three years ago is out of date. A lot of the things you decided you had to buy, or had to live without, are now buildable.
If there's something in your business held together by a shared folder and a few people who "just know how it works," that's usually the clearest signal it's worth a look. Describe it below and I'll tell you honestly whether it's worth building, smarter to buy, or fine to leave alone — and where the real return is hiding.
I read every one of these myself. Expect a real reply, usually within a couple of business days. — John
The Efficiency Edge delivers one practical AI idea like this every two weeks — free, written for owners, not engineers. And if you'd rather we help you decide what's worth building, that's what we do.
Subscribe free → Talk to us →