# Technology strategy is mostly what you won’t build

- URL: https://onepointdelta.com/articles/technology-strategy-is-what-you-will-not-build
- Category: Technology strategy
- Published: 2026-08-23
- Reading time: 4 min (719 words)
- Source: OnePointDelta

_A good technology strategy is short. It names the few systems worth owning, the many worth renting, and the line between them nobody crosses without a reason._

Ask a business for its technology strategy and, if it has one, it is usually a list of things to acquire: a new platform this year, a data initiative next year, AI somewhere on the horizon. Lists like that are not strategies. They are shopping. A strategy is a set of decisions about what the business will and will not do, and in technology the most important decisions are the negative ones.

Every business can build or buy almost anything now. The constraint is no longer capability. It is attention, maintenance and the compounding cost of every system that has to be kept alive. The businesses that use technology well are not the ones with the most of it. They are the ones that decided, deliberately, where it matters.

> A technology strategy is a short list of what you will own, a longer list of what you will rent, and a clear line between them.

## The one question

For each part of the operation, the strategic question is the same: does the way we do this make us different, or does it just need to work?

If it makes the business different, the way customers are handled, the way work is scheduled, the specific logic that decides what happens next, that is worth owning. A system built around it becomes an asset nobody else has, because nobody else works that way.

If it just needs to work, sending email, taking payment, hosting files, running a calendar, rent it. The best providers do it better than any business could, and the business gains nothing by owning it.

> Own what makes you different. Rent what merely has to work. Most technology decisions are one of those two, misfiled.

## Where the line usually goes

| Part of the operation | Usual answer | Why |
|---|---|---|
| Customer record and history | Own | It is the business's memory and every process reads it |
| Workflow between stages | Own | This is how the business actually works |
| Decision logic and rules | Own | It encodes judgment specific to the business |
| Payments, email, calendar, storage | Rent | Commodities, done better elsewhere |
| Accounting and payroll | Rent | Regulated, standard, and not where differentiation lives |
| Reading and drafting at scale | Rent the model, own the wiring | The model is a commodity; what it is connected to is not |

The line moves for each business. A company whose product is scheduling will own its calendar. Most will not. The discipline is in drawing the line on purpose rather than letting it be drawn by whichever vendor called first.

It also moves over time. Something rented at the start, because the business was small and the process was ordinary, can become worth owning once the process has become the thing customers notice. The strategy should say when that review happens, so the decision is revisited on schedule rather than in a crisis.

## What the strategy protects against

**Accumulation.** Without a line, tools accumulate. Each one was reasonable on its own; together they are an operation nobody can see whole, connected by people copying between them.

**Building the wrong thing.** Engineering effort spent re-creating a commodity is effort not spent on the system that would actually differentiate the business.

**Renting the wrong thing.** A core process run on a generic tool bends the business to fit the tool. The business slowly becomes the average of the vendor's customers.

**Perpetual migration.** Every rented core process is eventually migrated, at cost, when the vendor changes, raises prices or shuts down. Owned systems change when the business decides.

## Writing it down

A technology strategy for a business of ordinary size fits on a page. It names the handful of systems to own and why. It names the categories to rent and the current choices. It states the rule for the line, so that the next decision can be made without a meeting. And it says what will not be built this year, which is the part that stops the list from turning back into shopping.

That page is more useful than any roadmap, because it survives contact with the next vendor pitch. The question is already answered: does this make us different, or does it just need to work? Everything the business builds after that is a consequence of the answer, and everything it declines to build is the strategy doing its job.