For twenty years there were two questions a business asked when it hit a wall.

Is there a tool for that? Or who do we hire?

The first question built the SaaS economy. The second built every consulting firm, agency, and in-house team that ever existed. One sold you software you had to learn and manage and rarely did it deliver everything you needed. The other sold you people's time. Both were the only options on the menu.

Both are now the wrong question.

Software stopped being the valuable part

Something (not so) quiet happened to software. It stopped being scarce. The thing that used to take a team a quarter to build, a competent operator can now generate in an afternoon. When anyone can produce the tool, the tool stops being the valuable part.

Aravind Srinivas, the founder of Perplexity, said it best. When people dismissed his company as just a wrapper around someone else's model, he shrugged it off: everything is a wrapper. OpenAI is a wrapper around Nvidia and Azure. Netflix is a wrapper around AWS. Salesforce is a wrapper around an Oracle database, worth hundreds of billions. The wrapper was never the point. What the wrapper does for you is.

Now the cost of wasting time learning a new tool that solves only part of your problem is oftentimes more expensive than building the tool yourself for your specific purpose.

Take data analytics. It used to be one of the most expensive software sectors going, the domain of Tableau, Power BI, and Looker, platforms that cost a fortune and a specialist to run. Now you can build a dashboard from your own data in an hour. More importantly, you can describe how you want to see the data, in plain language, and get the result back.

But what if you know nothing about data analytics? Well you could try to wrangle a tool and be frustrated, or hire someone to manage your analytics. Both leave you carrying something. The tool hands you the work. The hire hands you a salary, a manager, and the hope that the output is good.

A third option on the menu

A new option has appeared between the two old ones, and most people do not have a name for it yet. Some have dubbed it software in a service.

It is easy to mistake this for consulting, so start with what it is not. A consultant or an agency sells you their time. You brief them, you wait, you pay by the hour or the retainer, and what comes back is work done by hand, once, for you. The cost scales with the effort: twice the work means twice the bill. And when the engagement ends, the capability walks out the door with them.

Software in a service is different in one decisive way. You are not paying for the expert's time. You are paying for the systems they have already built, the ones that turn their expertise into a result they can deliver again and again, without doing it by hand each time. The software helps them deliver it, but the software was never the point. The knowledge is the product. The software is just the wrapper around it.

So you are not buying a tool, and you are not hiring a person. You are buying a result. You pay for the expert's knowledge and experience, wrapped in productized deliverables. More than you would pay for software alone, because there is real judgment on the other end. Less than a full custom build or a new hire, because the systems do the heavy lifting and the expert is never starting from zero.

Go back to the analytics problem. You do not buy the tool, and you do not hire the analyst. You subscribe to the outcome: the dashboard built the way you need it, kept current, and the read on what the numbers mean, delivered by someone whose entire operation is built to produce it. It lands every week, done. You never touch the machinery.

The part that matters most is what you stop carrying. You stop learning the tool. You stop managing the tool. You stop paying for someone's hours and hoping the work is good. You stop worrying about whether any of it survives the next model update. You worry about the outcome. The service takes care of the rest.

That last line is the whole thing. The value is someone who can reliably deliver a result, and who has built the machinery to do it again and again without reinventing it each time.

The question companies should ask first

This changes the first move a company makes when it needs something done.

The old reflex was headcount. More work means more people. It is so automatic that most leaders never question it. But headcount is the most expensive, slowest, least reversible way to buy capability. You are not just hiring a skill, you are taking on management, onboarding, tooling, and risk that compounds for years.

The new first question is different. Before you hire, before you buy a tool, ask: what can we have fully managed, and what do we genuinely need to build internally? Some things are truly core, and you should own them. Most things are not. Most things are outcomes you want delivered reliably, not capabilities you want to babysit.

This is not a fringe prediction. IDC expects 70% of software vendors to rebuild their pricing around outcomes, consumption, or capability by 2028, and Gartner projects at least 40% of enterprise SaaS spend to shift to usage, agent, or outcome based pricing by 2030. The market is already repricing itself around results instead of access.

The business is becoming microservices

Software went through exactly this shift a decade ago, and the lesson is worth borrowing.

We used to build monoliths. One enormous application that did everything, where every part was tangled into every other part, and changing one thing meant understanding all of it. Then we learned to break the monolith apart into microservices. Small, independent units, each doing one job well, each replaceable without touching the rest. The system got more resilient, more flexible, and far easier to reason about.

Businesses are about to go through the same decomposition. The monolithic company, the one that hires for everything and owns every function in-house, is the old architecture. The new one looks like microservices. A lean core that you genuinely own around your own expertise and IP, surrounded by a set of outcomes delivered as managed services, each one replaceable, each one accountable for a result rather than a line on the org chart.

You do not staff a department to handle a thing. You subscribe to the outcome of that thing, from someone whose entire operation is built to deliver it.

This is already how it works

I am not forecasting. This is already happening. I run my own business this way and know many others that do the same. I don't learn a new tool or hand my knowledge to a new hire. I wrap my knowledge into a system, then build the software tools around it as I need them. On paper I have built 6 different value chains, but they all feed into the same business managed by the systems underneath. We structure all of our new partnerships on outcomes and objectives, not on time based rates.

The companies that win the next few years will not be the ones with the most software or the most people. They will be the ones who figured out, earliest, that they were never really buying either. They were buying outcomes. And outcomes are a lot easier to buy now.