Software spend is up, and half of it is quietly being cut

The common thread today is money and momentum: businesses are spending more on software than they have in a decade, and yet they are retiring roughly as much of it as they add. What decides which tools survive is rarely the longest feature list, and more often three quieter forces — habit, cost discipline, and how the team is organised around the work.

Hold those three in mind as you read the rest. They explain the numbers, they explain why so many products with real users still lose their place, and they explain why buying a new tool is only ever half of a decision.

Spending is up, and so is the cull

The headline number is easy to misread. Total business software spend is growing at its fastest pace in a decade, and a plain reading of that would suggest companies are buying more of everything. The reality reported today is stranger. Spend is rising and a large share of software is losing ground at the same time [1]. Both facts are true at once, which is the point worth sitting with.

The way to reconcile them is to stop thinking of a budget as a shopping list that only grows. A rising budget can mean concentration rather than accumulation — more money flowing to fewer tools that have earned a permanent place, while a long tail of half-used subscriptions is allowed to lapse. A business spending more is often a business that has finally worked out what it will pay for and what it will not.

That has a practical consequence for anyone choosing tools this quarter. The competitor for your budget is not only the other vendor in the same category. It is the tool you already own that does 70 per cent of the job, and the honest question of whether the new one is worth the switching cost and the extra line on the invoice. This is the difference between a purchase that adds value and one that simply adds spend. If you have not read what a subscription plan is really selling, it is a useful frame here: you are rarely buying features, you are buying an ongoing relationship and a recurring cost that has to keep justifying itself.

The discipline that survives a market like this is boring and unglamorous. Before you add a tool, decide what it replaces. Before you renew one, check whether anyone still opens it. A tool that no one can explain the value of, in one sentence, at renewal time, is a tool you are paying for out of inertia. Our own view on this — that software should keep earning its keep rather than coast on the fact that it is already installed — is set out in choose software worth using, and we hold ourselves to the same test.

The pull of what is already there

Today's roundup of alternatives to a well-known automation product opens not with a feature comparison but with an anecdote about the author's father, who still uses Internet Explorer — not because he has judged it the best, but because it was already there when he bought his first computer [2]. It is a small story that carries the whole weight of the spending numbers above.

Most tool decisions are not made on merit. They are made by default. The software that is already installed, already learned, already wired into someone's morning routine has an advantage no rival can list on a comparison page. That advantage is inertia, and it is the reason a genuinely better option can sit unused for years. When a publisher writes a piece about alternatives to an incumbent, the real subject is not the alternatives at all. It is the cost of leaving, and whether that cost is finally worth paying.

This matters for a business in two directions. Outwards, it means your own tools have a gravity of their own: the longer a team uses something, the harder and more expensive it becomes to move off it, whatever its shortcomings. Inwards, it means you should be suspicious of your own defaults. The question is not "does this tool still work" — most defaults do, well enough. The question is what you would choose if you were starting today with no history, and whether the gap between that and your current setup is large enough to act on.

Automation tools sharpen the point, because switching them means rebuilding the flows that other work now depends on. That is exactly why the decision to automate at all deserves care before you commit. We wrote about the test for that in how to tell whether a task should be automated: automate the task that is stable, repetitive and well understood, and leave alone the one that still changes shape every week. A default you chose deliberately is an asset. A default you drifted into is a liability you have not priced yet.

Organising a team around AI, not just buying it

The third item is a set of takeaways from a session at an industry event on building what its speaker called an AI-native marketing team. The speaker's background is worth noting, because it is the kind of experience that lends the advice weight: she has run marketing at a company later acquired by Adobe for billions, and led marketing for Adobe's enterprise division [3]. This is not a first-time experiment being reported as one.

The useful idea underneath the session title is one that is easy to miss in a year full of AI product launches. Getting value from AI is less about which model or which tool you buy, and more about how the team and its work are arranged around it. A business can license every AI feature on the market and change nothing about its output, because the roles, the hand-offs and the review steps still assume a pre-AI way of working. "AI-native" is a claim about structure, not about a purchase order.

That reframes a decision many businesses are facing right now: when to add a person and when to add automation. The honest answer changes once AI is in the mix, because some of the work a new hire would have done is now work you can restructure a role around instead. We laid out the trade-off in when to hire and when to automate, and today's session is a reminder that the two decisions are the same decision seen from different angles. The team that gets the most from AI is not the one that spent the most on it. It is the one that redesigned who does what, and in what order, so the tools have somewhere to fit.

What ties the three together

Record spending, the pull of defaults, and the shape of the team are not three separate stories. They are one. Money is flowing to software faster than ever, and yet a business gets no return on a tool it uses out of habit, or one it bought without changing how it works. The spend is easy. The discipline — knowing what to cut, what to keep, and how to arrange the work around it — is the part that decides whether any of it pays off.