Building custom software used to be slow and expensive, so most businesses rented instead. That has shifted. Building is now faster and more affordable, subscription prices keep climbing, and businesses want to own their data rather than depend on vendors. The result is a steady move to replace rented tools that no longer fit with builds that do.
Information current as at 5 July 2026
For a long time the advice was simple: never build what you can buy. That advice was right for its time, because building was slow, expensive and risky. The ground has moved, though, and a growing number of businesses are quietly reversing the default for the tools that matter most to them. This is why, and whether it applies to you.
The old advice made complete sense. Building software meant hiring a team, waiting many months, spending a large sum, and carrying the risk that the result would not work. Against that, a subscription was cheap, instant and someone else's problem to maintain. For all but the largest companies with the most specific needs, renting won almost every time. This history is worth respecting, because it explains why so many businesses have a stack of subscriptions today. It was the rational choice, and for a great many tools it still is. The point is not that the old advice was wrong, but that the conditions it was based on have changed. It is worth holding both thoughts at once: the default to buy was correct for its era and remains correct for most tools today, and yet the specific conditions that made it near-universal have softened for a certain band of software. Respecting the history keeps you from swinging to the opposite error of assuming everything should now be built, which is just as costly a mistake as the old refusal to build anything.
Three things shifted at once. First, building got dramatically faster and cheaper, so the huge upfront cost and long wait that made building unthinkable for small businesses shrank to something a fraction of a legacy build. Second, subscription prices kept climbing, especially per-seat and usage-based pricing that grows with your success, so the running cost of renting became more visible and more resented. Third, businesses grew wary of depending on vendors who can change prices, remove features, or hold data hostage. None of these alone would flip the default, but together they narrowed the gap until, for certain tools, building crossed from unthinkable to sensible.
If you have made something and it needs to become real, send it over. We will tell you honestly what it needs to be live, safe and yours, whether that is a quick fix you can do or a proper build. No obligation.
The businesses making the move are rarely doing it for novelty. They are doing it because a specific tool stopped serving them. The subscription that got more expensive every year while doing the same job. The stack of overlapping tools that never quite fit together. The workflow that is genuinely specific to them, which no off-the-shelf tool serves well because it was built for the average business, not theirs. And the growing wish to own the system that runs their operation rather than rent it from someone who could change the terms. The switch is a response to real friction, tool by tool, not a wholesale rejection of buying.
This shift does not suit everyone or every tool, and it is worth being honest about that. If your subscriptions are cheap, well-fitting and stable, there is no reason to move, and building would cost you more than it saves. The businesses this suits have a different profile: a core workflow that is specific to them, subscription costs that climb faster than the value they deliver, a real need to own their data, or a tangle of tools that one system could replace. If that describes you, the move is worth considering seriously. If it does not, the old advice still holds, and the smart choice is to keep buying and spend your attention elsewhere.
It is worth keeping the movement in proportion. The businesses replacing tools are doing so selectively, for the systems where the case is strongest, while continuing to rent everything else. Almost nobody is abandoning subscriptions wholesale, and the ones who try usually regret rebuilding things that were cheap and well-solved. What has actually changed is that building crossed from unthinkable to reasonable for a certain band of tools, so a decision that used to be automatic, always buy, is now a genuine choice worth thinking through tool by tool. That is the healthiest way to read the trend: not as a signal to build everything, but as a reason to stop treating buying as the only option and to ask, for each tool that frustrates you, whether it has reached the point where owning it would serve you better. For most of your stack the answer will still be to keep renting. For one or two tools, it may not be, and the value of the shift is simply that the question is now worth asking.
If you have built something with AI already, you are closer to this shift than most, because you have seen first-hand that software you shape to your own needs is now within reach in a way it was not a few years ago. That experience is genuinely useful, but it also carries a risk: having built one thing, it is tempting to want to build everything, including tools that are far better rented. The sensible response to the trend is neither to ignore it nor to over-apply it. Look at your own stack with fresh eyes. Find the tools that frustrate you, that cost more each year, that fit your work badly or trap your data, and ask honestly whether those specific tools have reached the point where owning them would serve you better. For the rest, keep renting without a second thought. The businesses that benefit most from this moment are not the ones that build the most software, but the ones that read their own situation clearly and act on the handful of tools where building genuinely wins, while leaving the many where it does not exactly where they are.
If you have made something and it needs to become real, send it over. We will tell you honestly what it needs to be live, safe and yours, whether that is a quick fix you can do or a proper build. No obligation.
Whether you can name exactly what you want built, or you just know something is leaking, the next step is the same conversation.