← Back to home
Technology Strategy

Technology Strategy Is the Discipline of Trade-offs

A strategy is not a list of technologies to adopt. It is a coherent set of choices about where to invest, what to simplify, and what not to do.

I once sat through an annual technology strategy offsite that ended with fourteen strategic priorities.

Every one of them was reasonable. Cloud adoption. Data platform modernization. API enablement. Zero trust security. Developer experience. Automation. AI. Observability. Customer 360. And several more.

Each priority had an owner, a slide, and a target date. The final presentation looked comprehensive and confident.

A year later, I looked at what had actually changed.

The organization now had three integration platforms instead of two. Two migrations were roughly half complete, which meant both the old and new environments had to be funded and supported. Teams were stretched across so many initiatives that the most important ones had moved slowest of all.

Nobody had made a bad decision. The problem was that nobody had made the hard ones.

To me, a strategy that only says yes isn't a strategy. It's a wish list with a timeline.

A roadmap is not a strategy

Technology strategy is often presented as a roadmap: platforms to adopt, systems to migrate, capabilities to build, arranged neatly across quarters.

A roadmap is useful. It shows intent and sequence. But it doesn't explain why those choices, and it rarely shows what was given up to make room for them.

Consider two ways of presenting the same cloud program.

The first says: "Migrate 80% of workloads to the cloud by the end of next year." It is clear and measurable. It is also silent on why, and it invites teams to lift and shift whatever is easiest to hit the number, including systems that gain nothing from moving.

The second says: "Our product teams currently wait weeks for environments, and our peak-season capacity costs us heavily for capacity we use a few weeks a year. We will move the customer-facing and seasonal workloads first, because that's where elasticity and speed pay off. Stable back-office systems will stay where they are until there's a reason to move them."

Same technology. Very different strategy. The second version makes a trade-off visible, and it gives teams a way to make good decisions without escalating every one of them.

Choose the outcomes first

I find it helpful to start every strategy conversation with a simple question: what needs to be different for the business in two years, and how will we know?

Faster product change. Lower operational risk. Better customer experience. Stronger resilience. Lower cost to serve. These are outcomes. A new platform is only ever one possible means of achieving them.

When strategy starts with technologies, it tends to measure adoption: how many workloads migrated, how many teams on the new platform. When it starts with outcomes, it measures change: lead time, incident rates, cost per transaction, customer effort.

Adoption is easy to report. Change is what the business is paying for.

Make the constraints explicit

Every technology decision has costs, and the most dangerous ones are the costs nobody wrote down.

Migration effort. Skills the organization doesn't yet have. Vendor dependency. Additional operating complexity. And, most often overlooked, opportunity cost: everything the same people could have been doing instead.

A strong strategy puts these costs on the table rather than hiding them behind a target-state diagram. It is uncomfortable, because it makes it obvious that not everything can be done. That discomfort is the point.

I've found that a simple one-page view for each major initiative, covering the outcome, the investment, the people it needs, what it will displace, and the risk if we don't do it, leads to far better conversations than any capability map.

Three priorities, not fourteen

When everything is strategic, nothing is.

In my experience, very few organizations can genuinely execute more than three transformational priorities at the same time. The constraint usually isn't money. It's the limited number of people who understand the business, the systems, and how to lead change, and the leadership attention needed to unblock the hard parts.

A credible strategy identifies the few changes that will create disproportionate value and protects the capacity to deliver them. Everything else is either business as usual, deliberately deferred, or stopped.

That means some good ideas won't get funded this year. I believe that's not a failure of the strategy. It's evidence that the strategy is working.

Define what you will stop doing

This is the part most strategies leave out, and I think it's the most valuable.

Organizations are very good at adding technology and remarkably poor at removing it. Every acquisition, every urgent project, and every enthusiastic team leaves something behind. Over time, the estate fills with overlapping tools, bespoke integrations, and systems that nobody wants to own but everybody is afraid to switch off.

I've seen an environment running four separate monitoring tools, each introduced for a sensible reason at the time. Incidents routinely involved four teams looking at four dashboards, each with part of the picture. Consolidating them wasn't exciting work, and it never appeared on a strategy slide. But it reduced incident resolution time more than any new tool would have.

A rule I'd like to see in more strategies: every new platform should come with a named retirement. If we are introducing something, we should be explicit about what it replaces, by when, and who is accountable for switching the old one off.

Simplification doesn't happen as a side effect. It has to be funded, tracked, and celebrated like any other delivery.

Build where you differentiate, buy where you don't

Build-versus-buy decisions are where strategy often becomes emotional.

Building feels like control. Engineers enjoy it, and custom solutions fit the organization's processes precisely. But I've seen a team spend two years building an internal ticketing and workflow tool that, by the end, required six engineers just to keep running. A commercial product would have done 90% of what was needed from the start.

My default is straightforward. If a capability is not the reason customers choose you, such as identity, HR, service management, or most internal tooling, buy it, configure it lightly, and adapt your processes where you can. Save your scarce engineering talent for the few things that genuinely make the business different.

There are exceptions, of course. But they should be argued for, not assumed.

My perspective: strategy is the courage to choose

I have come to believe that the real test of a technology strategy is not how ambitious it sounds, but how clearly it says no.

A good strategy helps an organization say yes with conviction to the few things that matter most. Just as importantly, it helps people say no, or not yet, without feeling that they are blocking progress, because the reasons were agreed in advance.

There are contexts where this needs nuance. Early-stage and fast-growing organizations sometimes benefit from keeping options open rather than committing too hard, too early. Some regulated industries genuinely need redundancy that would look wasteful elsewhere. Those are real considerations. But they are specific situations with specific reasons, not a general excuse to avoid choosing.

My view: technology strategy is the discipline of trade-offs. Choose the outcomes, make the costs visible, limit the priorities, and be explicit about what you will stop. If a strategy can't do that, it isn't finished yet.


A roadmap tells people what you'll do. A strategy tells them what you won't, and why.

END OF ARTICLEExplore more thinking ↗
DISCUSSION

Reactions & comments

Sign in with GitHub to react or join the conversation.

Discussion is not configured yet.

Add your giscus repoId and categoryId to site.config.ts.