Make Versus Buy Must Follow Portfolio Strategy — Not Define It
The ability to buy more work does not reduce the need to know what matters.
The Decision That Started at the Wrong End
A technology portfolio review began with a seemingly practical question: which services could be outsourced within twelve months?
The team classified 420 applications, two data centres, desktop support, networks, the service desk and application maintenance. Market availability was scored, supplier interest tested and indicative savings estimated. The final proposal suggested transferring 62 per cent of the annual technology cost to external providers under two large contracts.
Only after the recommendation reached the executive committee did a more important question surface: which technology capabilities would determine the organisation’s future operating model?
The review had produced a sophisticated answer to what can be bought. It had not first decided what must remain strategically controlled. Outsourceability had become the organising principle of the portfolio. The tail was wagging the dog.
This white paper argues for a different decision order. Make-versus-buy choices should follow portfolio strategy, not substitute for it. The organisation should first identify where technology creates strategic value, where integration creates enterprise risk and where internal knowledge is necessary to govern future change. Only then should it decide which services the market can perform more efficiently.
Why the Question Matters Now
The present outsourcing wave is not accidental. Technology estates expanded rapidly through the 1990s. Year 2000 remediation exposed duplication and poor documentation, while enterprise system programmes left many organisations with expensive mixed estates: mainframes, distributed servers, packaged applications and hundreds of local systems. Boards now expect lower cost and more predictable service.
Suppliers have also become more capable. Data-centre operation, networks, desktop support and application maintenance can be delivered at scale. Offshore development and support models promise substantial labour savings. Multi-year contracts offer a route from fragmented internal provision to measured service levels and a single commercial relationship.
The case for buying is therefore strong:
- access to scale and specialist expertise;
- conversion of internal activity into contracted service;
- clearer unit costs and service measures;
- accelerated standardisation of infrastructure and support;
- reduced dependence on hard-to-recruit technical staff.
Yet these benefits arrive through a contractual boundary. That boundary changes how priorities are set, how knowledge moves and how quickly the portfolio can be reshaped. A service may be cheaper to operate and more expensive to change. A supplier may meet every service level while the enterprise loses the ability to judge whether the service still supports its strategy.
The make-versus-buy decision is therefore not merely a sourcing choice. It is a portfolio design decision with consequences for future freedom of action.
The Failure in the Conventional Business Case
The conventional outsourcing business case usually compares an internal cost baseline with supplier charges, transition cost and retained overhead. This is necessary, but it contains three recurring weaknesses.
The internal baseline is incomplete
Internal costs are dispersed across technology, property, procurement, finance and business units. Depreciation, contractor spend and local support are treated inconsistently. The resulting baseline may understate or overstate the true cost, making the apparent saving unreliable.
More importantly, internal cost contains activities that will not disappear after transfer. Architecture, demand management, information security, contract management, service integration and business relationship roles remain necessary. When these retained capabilities are excluded, the saving is overstated precisely where governance need will increase.
The contract prices stability, not uncertainty
Suppliers can offer attractive prices where volumes, service definitions and responsibilities are clear. Technology portfolios are rarely that stable. Application demand changes, acquisitions introduce new systems, regulation creates work and business programmes alter priorities.
A low initial price can therefore depend on narrow assumptions. Cost reappears through change requests, volume bands, project rates and disputed responsibilities. The organisation has not removed variability; it has moved variability outside the base price.
Strategic value is reduced to service criticality
Most assessments distinguish critical from non-critical systems. That is insufficient. A payroll system may be operationally critical but strategically standard. A modest customer-pricing engine may have lower current volume but contain knowledge central to competitive choice. An integration layer may be invisible to customers yet determine how quickly the organisation can combine products or reorganise operations.
Operational criticality asks what must not fail today. Strategic value asks what the organisation must be able to change intelligently tomorrow.
A portfolio should not retain work because it is difficult, nor outsource work because it is possible; it should place each capability where strategic control and economic advantage can both be sustained.
A Portfolio-First Decision Model
Every technology capability should be assessed across five dimensions before a sourcing route is chosen.
| Dimension | Question | What high exposure implies |
|---|---|---|
| Strategic differentiation | Does this capability shape products, pricing, customer treatment or operating advantage? | Retain strong internal ownership and design knowledge |
| Integration dependency | Does change here affect many processes, applications or suppliers? | Preserve architecture and integration control |
| Market maturity | Can several credible suppliers deliver a genuinely comparable service? | Buying becomes more viable and contestable |
| Change intensity | Are demand, volumes and requirements likely to change materially? | Avoid rigid commitments and narrow charging assumptions |
| Transition feasibility | Are assets, processes, data and responsibilities understood well enough to transfer? | Delay or phase sourcing until ambiguity is reduced |
These dimensions do not produce a mechanical answer. They expose the trade-off that the financial case alone conceals.
A mature, standard service with stable demand and several capable suppliers is a strong candidate to buy, even if it is operationally important. A capability central to product design or enterprise integration should remain strategically owned, even where external specialists perform parts of the work. A poorly documented service may eventually be outsourced, but ambiguity should not be exported and then paid for through transition disputes.
The key distinction is between execution and control. An organisation may buy execution while retaining control of architecture, priorities, data, service design and supplier integration. Conversely, it may keep staff on the payroll yet have little real control because knowledge is fragmented among contractors and local teams.
The Options That Should Be Weighed
Four broad options deserve honest consideration.
Continue internal provision
This protects direct authority and accumulated knowledge. It is appropriate where capability is differentiating, demand changes rapidly or supplier markets are immature.
Its weakness is that internal provision can preserve inefficient scale, inconsistent processes and hidden cost. Retaining work without a credible improvement plan is not strategy; it is avoidance.
Outsource an integrated tower
A broad infrastructure or application-services contract can simplify accountability, accelerate standardisation and secure scale economies. It may be suitable where services are mature and interfaces can be clearly defined.
The risk is concentration. A long-term supplier gains commercial leverage and deep operational knowledge. If the organisation weakens its retained architecture and commercial capability, future choices become dependent on the incumbent’s interpretation, price and willingness.
Source separate service components
Separating data-centre operation, networks, desktop, service desk and application maintenance can preserve competition and select stronger specialists.
The weakness is integration. Incidents and changes cross service boundaries. Without an effective retained integration function, each supplier can meet its own obligation while the end-to-end service fails. The apparent avoidance of dependency may create dependency on internal coordination that the organisation has not funded.
Use a selective hybrid
A selective model retains strategically differentiating and integration-intensive capabilities while buying mature execution services. It can combine market efficiency with strategic control.
Its difficulty is management complexity. The boundary must be designed deliberately, internal roles must be credible and suppliers must work within enterprise standards they do not own.
The strongest argument for a comprehensive transfer is simplicity. One contract, one accountable provider and a substantial cost commitment can force decisions that internal governance has repeatedly deferred. Selective sourcing can become an excuse for keeping every politically sensitive activity in-house while outsourcing only the easy margins.
That objection is valid. The answer is not indiscriminate retention. It is a disciplined portfolio test with explicit evidence and executive ownership. Selectivity must be based on strategic and integration logic, not departmental preference.
Recommendation: Selective Sourcing Around a Strong Retained Core
For most diversified organisations, the defensible position is a selective sourcing model governed by a small but capable retained core.
The retained core should own:
- technology portfolio priorities and investment allocation;
- enterprise architecture and integration standards;
- information security policy and risk acceptance;
- demand shaping with business units;
- service design and end-to-end performance;
- commercial strategy, contract management and benchmarking;
- knowledge required to change suppliers or bring work back.
External providers may execute infrastructure operation, desktop support, networks, service desk activity, maintenance and defined project work where the market offers scale and comparable performance. Strategic applications can still use external development capacity, but product direction, architecture, data judgement and acceptance of design trade-offs should remain under enterprise control.
This is not a compromise between making and buying. It is a recognition that the boundary need not place every element of a capability on the same side.
A customer decision system, for example, may retain business rules, data definitions, architecture and release authority internally while using an external team for coded enhancements and testing support. The organisation buys capacity without outsourcing the judgement that determines what the system must become.
The Economics Must Include Freedom to Change
A credible business case should compare more than five-year operating cost.
Consider a representative estate with annual technology spending of £160 million. An initial supplier proposal reduces the visible run cost by 14 per cent, or £22.4 million. After adding £8 million of retained capability, £6 million of transition cost spread over the term and realistic change demand, the expected saving falls to approximately £10 million a year.
That may still be attractive. But the decision should also test:
- cost under high and low volume scenarios;
- project and change rates outside the base service;
- responsibility for asset refresh and software licences;
- the cost of retained integration and governance;
- termination assistance, data return and knowledge transfer;
- the time and cost required to compete or separate services later.
A business case that models only the preferred volume and a stable portfolio gives false precision. Scenario ranges are more honest. If the saving disappears under plausible change demand, the contract is not creating efficiency; it is betting that the portfolio will remain still.
The most important commercial terms are therefore not merely the initial unit rates. They are the mechanisms that preserve adaptability: benchmarking, transparent charges, usable performance data, change control, partial termination rights, transition assistance and continued access to documentation.
How to Make the Decision
The executive portfolio forum should make the sourcing decision in a deliberate sequence.
- Define strategic outcomes. State how technology must support the operating model, product plans, cost position and regulatory obligations over the planning period.
- Map capabilities, not departments. Group work into coherent capabilities and services, including the architecture, data and decision rights that govern them.
- Assess the five dimensions. Rate strategic differentiation, integration dependency, market maturity, change intensity and transition feasibility using evidence.
- Design the retained core. Identify the minimum capabilities needed to direct, integrate, assure and later change the sourcing model.
- Test options and scenarios. Compare internal improvement, integrated outsourcing, component sourcing and selective hybrid arrangements under realistic demand variation.
- Sequence the portfolio. Source mature, well-understood services first; stabilise or document ambiguous capabilities before transfer.
- Approve the whole operating model. Decide suppliers, retained roles, governance, information flows and exit mechanisms as one design, not as separate procurement activities.
The decision gate should reject any proposal that cannot name the retained decision owners, explain end-to-end integration or show how the organisation could change course after contract signature.
The Case for Change
Outsourcing should reshape the technology portfolio, but it should do so by clarifying strategic control, not by allowing supplier capability to define enterprise priorities.
The present market creates a genuine opportunity. Many organisations no longer need to own every data-centre task, desktop team or maintenance function. External scale can improve service and release capital. But the ability to buy more work does not reduce the need to know what matters. It increases it.
If the portfolio begins with outsourceability, the organisation will optimise the services suppliers are ready to sell. If it begins with strategic value, integration and future change, it can use the market without allowing the market to design the portfolio.
The recommendation is therefore clear: build the sourcing case from the portfolio outward. Retain the authority, knowledge and integration capability that determine future choice. Buy execution where the service is mature, contestable and economically superior. Price uncertainty honestly. Preserve the ability to change the boundary.
Make versus buy is not a verdict on whether internal teams or suppliers are better. It is a decision about where the enterprise must keep judgement, where the market can add scale, and how both can work without surrendering strategic freedom.