Making Technology Strategy Subscription Work

Figures at the front

$38 billion: 2025 UK software market revenue
$48.37 billion: 2030 UK software market revenue
52 per cent: 2025 AI adoption among UK firms

The field emerged because the gap between a decision and the evidence that unsettled it closed. A strategy used to be a document that appeared every third year and was defended in between. A vendor pricing change, a model revision and a compliance deadline can now land in the same quarter, so boards began paying for a standing arrangement that keeps the decision set current, and stopped treating the document itself as the deliverable. That is the whole origin. The rest of this page is about running that arrangement on purpose.

A strategy is a set of decisions, each with an owner and an expiry point

Everything that comes after depends on one idea. A strategy is a set of decisions, each with an owner accountable for it and an expiry point set by something outside the organisation. The Technology Code of Practice is the clearest published version of that structure: thirteen points a project is scored against, mandatory points that must be met and the rest worked towards as far as is practical, and an instruction that where legacy technology limits adherence, the reason goes to the GDS assurance team. That last provision bears the weight, because where a project cannot meet a point, the cause is recorded beside it and the record stays honest at the moment of failure. Point 11 of the same document asks for a stated approach to what the organisation will pay for, showing that commercial and technology aspects and contractual limits were weighed together. That is the closest published analogue to what a subscription client owes its own board.

Accept the premise and the review problem becomes visible. The criteria move. The same page was updated on 7 July 2025 to change an assurance contact. When the standard moves, every decision calibrated against it moves with it, and a review that happened once has no way to catch the drift.

Why a decision set is sold by the month

A decision set has no completion point, which is why the market sells the work month by month. The United Kingdom software market is expected to generate $38 billion of revenue in 2025 and to reach $48.37 billion by 2030 at an annual rate of 4.83 per cent, according to the country commercial guide for UK strategic technologies, and the same guide records that AI adoption among UK firms rose from 39 per cent in 2024 to 52 per cent in 2025. Estate churn and vendor churn run on a contract expiry clock. A strategy revisited only when somebody commissions a review gets calibrated against data at least one contract cycle old, and the cost of that lag shows up as decisions taken on stale assumptions between reviews.

KPMG's UK technology strategy practice frames the same shift from the other side: the IT function has to move from service provider to strategic business partner, and leadership has to keep the link between technology decisions and the business plan. That link is a standing job. Nobody finishes it in a quarter and files it.

What the recurring fee has to produce

A technology strategy subscription owes three outputs, and an arrangement that delivers no such output is availability sold by the month. The first is a decision record that sets down the decision, the person accountable for it, the alternatives turned down, and the trigger that would reopen it. The second is a reviewed number the board already watches, whether that is spend under management, contracts approaching expiry, or open risk. The third is a standard you are held to, cited by title and by point.

Government publishes its versions openly, which is why they serve as templates. The Technology Code of Practice sets out the criteria, and the Cabinet Office digital handbook gathers the surrounding policies and strategies and sets a review marker on the collection, most recently 4 November 2024. A commercial client should ask for the same two properties from its own provider: a title on the standard, and a review point on the collection.

Where subscription models break, and what that warns a client about

The subscription form has known failure modes, and they transfer to strategy without much adaptation. Oded Koenigsberg's article on new threats to the subscription model sets out the mechanism: supply chain disruption means a company may be unable to honour today's promise at the same quality next month, and inflation squeezes margins while revenue per customer stays flat until the contract comes up again. He asks the resulting question plainly.

Can we still afford to meet the obligations we've made to our customers?

A strategy subscription bears the same exposure. Where the fee pays for a stated number of days from one senior person, the obligation is a fixed quantity of that person's time, which rising costs erode and which a busy quarter quietly absorbs. Where the fee pays for specified outputs at a stated cadence, the obligation is bounded and can be re-priced at each expiry point without renegotiating the relationship. Koenigsberg also observes, in the same MIT Sloan Management Review article, that subscribers cancel easily, which keeps a provider attentive from one period to the next.

The Forbes Technology Council account of challenges for tech service companies switching to subscriptions adds the client-side consequence. Several contributors flag the move from capital to operating cost models, the need to absorb a revenue dip during the changeover, and the difficulty of balancing customisation against scalability. Applied to strategy, that means telling the finance director in advance that the same work now appears as recurring operating cost with no capital asset at the end, and requiring the provider to standardise the core review so that bespoke work stays exceptional.

Four forms of engagement, each priced on a different base

Advisory subscription. This covers senior assessment at a stated availability, usually one or two days a month, and the price ties to the day rate. Its outputs are the decision log and the calls taken. It ends cheaply because nothing accumulates inside it, and the client keeps whatever has been set down.

Review cycle subscription. This secures a fixed annual cycle of reviews: architecture in the first quarter, vendor and contract exposure in the second, security and privacy posture in the third, roadmap refresh in the fourth. The cycle sets the price. Its outputs are four papers with expiry points plus the changes logged between them. It ends cleanly at the close of a cycle, which makes it the easiest form to stop.

Embedded fractional ownership. This takes on accountability for a domain, typically engineering, data or platform, and the price ties to decision load and attendance at governance. Its outputs take in the earlier ones plus the consequences of the decisions taken. It ends expensively, which is where the provisions in the last section begin to matter.

Portfolio subscription. This delivers control of the subscribed estate itself, priced against spend under management or the number of contracts governed. The case for it strengthens each year: TechTarget reported that IT budgets for third-party services grew by about 6 per cent a year from 2019 through 2024, citing Boston Consulting Group, and that unmanaged growth lets cost rise quietly while overlapping tools and unused capacity stay out of sight. Its outputs are the calendar of contract expiries, the position on usage measured against capacity paid for, and an exit plan drawn before the first payment.