When the Integrator Owns the Strategy, Transformation Loses Its Spine
Strategy cannot be outsourced without also outsourcing the organisation's power to choose.
The Programme That Looked Decisive
The first sign of trouble is often a thick binder on the boardroom table. It has the right vocabulary: business process redesign, enterprise architecture, change management, benefits tracking. It carries a credible timetable and a resource model with impressive precision. The steering committee can point to workstreams, milestones and named deliverables. After several hesitant months, the organisation finally appears to have movement.
Yet the decisive document has usually been written somewhere else.
A large integrator has shaped the case, defined the programme, selected the sequence, and supplied most of the people who will execute it. The executive team has approved the plan, but has not truly authored it. The organisation has bought not only implementation capacity, but strategic momentum. In the climate of 2001, with the excesses of the e-business boom now being marked down sharply, this distinction matters more than it did two years ago.
The pattern that recurs across technology-enabled transformation programmes is not that vendors are too powerful by nature. It is that clients too often invite vendors to fill a vacuum that should never have existed: the vacuum where business judgement, executive appetite, and operational sequencing belong.
Strategy cannot be outsourced without also outsourcing the organisation’s power to choose.
This paper argues for a more disciplined settlement. Large integrators have an important place: they can supply method, scarce technical labour, programme discipline and experience from comparable implementations. But they should not own the strategy. When they do, the programme becomes coherent on paper and brittle in practice.
Why Vendor Ownership Became So Attractive
The late 1990s created a particular kind of organisational dependency. Many executive teams had just come through Y2K remediation, enterprise resource planning decisions, early customer relationship management programmes, shared services reviews and the rush to put credible propositions on the web. The language of business change had become inseparable from the language of systems. Boards were told, often correctly, that fragmented processes and ageing applications were strategic constraints.
At the same time, many organisations had hollowed out the internal muscles needed to direct large change. Business units knew their operations, but not how to translate them into enterprise-wide design choices. IT knew the estate, but was too often treated as a service department rather than as a participant in strategy. Finance could challenge the investment case, but not the implementation logic. The programme office, where it existed, was commonly administrative rather than directive.
The integrator arrived with what the organisation lacked:
- a named method, which gave executives comfort that complexity had been domesticated;
- experienced programme managers, who could create pace where internal consensus was slow;
- software package knowledge, particularly around ERP and CRM suites;
- a commercial proposal that converted uncertainty into a sequence of phases, deliverables and rates;
- external authority, which helped senior teams say difficult things to their own organisation.
This was not irrational buying. In many cases it was the only way to move at all. The serious objection to retaining strategy internally is that many organisations were simply not equipped to do it. They had neither the methods nor the people nor the scars. A vendor-led approach appeared to reduce risk by placing responsibility with the party that had done it before.
That objection deserves respect. But it mistakes experience for ownership. A contractor can know the route through the mountain without owning the reason for crossing it.
The Hidden Cost of Borrowed Certainty
The danger is not that vendors deliberately produce bad strategies. The danger is subtler: the strategy becomes shaped by what the vendor can confidently deliver, sell and govern. The result is a programme that is internally consistent but externally under-owned by the business.
Three failure mechanisms are especially common.
- The answer is framed before the question is settled. Once an integrator has anchored the work around a package implementation, a shared services design, or an enterprise architecture programme, alternatives become harder to see. The question shifts from what must the business become? to how do we implement the chosen design?
- The sequencing follows delivery logic rather than business appetite. Vendors naturally prefer sequences that reduce technical risk and mobilise teams efficiently. The business may need a different order: fixing pricing discipline before customer reporting, clarifying product ownership before workflow redesign, or resolving regional operating model tensions before common processes are imposed.
- The organisation learns to review rather than decide. Steering committees become adept at approving papers, escalating issues and monitoring amber milestones. They become less adept at making uncomfortable trade-offs about scope, benefits, service levels and managerial accountability.
Consider a composite manufacturing group in 2001, still digesting two acquisitions and operating three different order-to-cash processes across Europe. The integrator’s proposal recommends a single ERP template, a phased rollout by country, and a central programme office staffed largely by consultants. The business case promises a 6 percent reduction in working capital and a 12 percent reduction in administrative headcount over three years.
The numbers are plausible. The programme is not foolish. But the first decision that matters is not the software template. It is whether the group is prepared to standardise credit policy, customer master data ownership, discount approval and month-end discipline across operating companies that have historically treated these as local rights. In the first steering committee, the regional directors agree to the programme in principle while reserving local exceptions. Nobody calls this a strategic contradiction. It is recorded as an implementation issue.
Eighteen months later, the technical build is 80 percent complete and the benefits case is already decaying. The local exceptions have multiplied. The common process design is common only at the level of charts. The central team is now negotiating with countries one by one, and the integrator has become the visible owner of compromises that the executive team should have made at the outset.
The mechanism is plain: when strategic choices are deferred into implementation, the people with the largest delivery footprint become the people who broker the strategy by default.
What Should Remain Inside the Organisation
A better model does not require executives to become systems specialists. It requires them to retain ownership of five decisions that cannot safely be delegated.
| Decision | Why it must stay internal | Vendor contribution |
|---|---|---|
| Strategic intent | Only the organisation can decide what advantage, discipline or capability the programme is meant to create. | Test whether intent is implementable and expose contradictions. |
| Operating model choices | These allocate power between functions, regions and business units. | Provide design options and consequences from comparable situations. |
| Benefit ownership | Savings and performance gains appear only when line managers change behaviour. | Help build the measurement discipline and dependency map. |
| Sequence and pace | The right order depends on appetite, politics, operational risk and cash constraints. | Model delivery feasibility and resource implications. |
| Governance rights | Decision rights determine whether the programme can resolve conflict. | Supply programme control, issue discipline and escalation mechanisms. |
These are not ceremonial decisions. They are the substance of transformation. If they are not made by the organisation, they will still be made; they will simply be made indirectly through design assumptions, timetable pressures, workshop outputs and change requests.
The most effective organisations I have observed treat vendor expertise as an input into executive judgement, not a substitute for it. The distinction shows up in small behaviours. The sponsor does not ask, “What does the method say?” before asking, “What are we unwilling to standardise?” The finance director does not accept a benefits schedule until each benefit has an accountable line owner. The operations director does not let process design workshops become forums for preserving local custom by another name.
A More Effective Commercial and Governance Settlement
The recommended response is not adversarial procurement. It is sharper client-side authorship.
Before appointing the main integrator, the executive team should complete a short but demanding strategy ownership exercise. It need not take months. In a well-run organisation it can be done in four to six weeks. The output is not a full design; it is a set of non-negotiable choices that shape the vendor brief.
- Define the business outcome in operational terms.
- Avoid phrases such as “modernise the enterprise” or “enable e-business” unless they are translated into measurable changes: reduced order errors, faster close, lower inventory, improved customer profitability, fewer manual reconciliations.
- State what will be different in management behaviour, not only what will be different in systems.
- Name the decisions that are already made.
- For example: one chart of accounts, one customer master, one credit policy, one procurement approval route above a stated threshold.
- If such decisions are not yet made, do not pretend implementation planning can proceed cleanly.
- Name the decisions that remain open.
- Invite vendors to advise where there are genuine options.
- Separate advisory uncertainty from delivery uncertainty so the commercial proposal does not hide one inside the other.
- Assign benefit owners before assigning workstream leads.
- A workstream lead can deliver a process design.
- A benefit owner must change the operating rhythm that makes the benefit real.
- Establish the client-side design authority.
- This body should be small, senior and cross-functional.
- It should decide on exceptions, process standards and unresolved trade-offs; it should not merely receive progress reports.
Only after this work should the integrator be asked to propose delivery. The request should make clear where the vendor is being asked to advise, where it is being asked to implement, and where the client has already decided.
Commercially, this means resisting the comforting fiction that a single large contract can transfer transformation risk. It can transfer certain delivery obligations. It cannot transfer managerial accountability. Contracts should reward delivery discipline, transparency and capability transfer, not the expansion of vendor dependency.
A practical contract will therefore include:
- a defined client-side decision calendar, with named executive decision owners;
- explicit assumptions behind the delivery sequence, so business-driven changes are visible rather than disguised;
- joint quality gates where business readiness carries equal weight with technical completion;
- staged commitment, allowing scope to be refined after strategic decisions without converting every clarification into a commercial dispute;
- planned reduction of vendor roles in selected areas as internal capability matures.
The last point is often neglected. A programme that still requires the vendor to interpret the organisation’s strategy after two years has not built sufficient client capability. It may have installed systems, but it has not strengthened the organisation’s power to change itself.
Evidence of a Healthier Pattern
In healthier programmes, the integrator is still prominent. The difference is that prominence is not confused with ownership.
A recognisable pattern emerges. The first executive workshops are not vendor showpieces; they are decision sessions. The design authority keeps a visible log of policy choices, exceptions and unresolved tensions. Benefits are stated in operational measures that line managers already understand: days sales outstanding, stock turns, invoice rework, close timetable, procurement compliance, call-handling transfer rates. The programme office reports not only milestone completion but the age of unresolved decisions.
In one composite services business, the initial plan assumed a nine-month rollout of a common finance process across six countries. The client-side design authority forced a pause after discovering that 40 percent of invoice rework came from contract terms agreed upstream by sales teams, not from finance processing itself. A vendor-owned strategy would likely have treated this as a finance workflow issue. The revised programme put contract approval discipline before finance system rollout, delayed two country deployments by one quarter, and reduced expected first-year savings from 8 percent to 5 percent.
That looked like a retreat. It was in fact a better strategy. By lowering the declared saving and correcting the sequence, the organisation made the benefit more likely to appear. The integrator still delivered the process and system work. The client owned the judgement that made delivery meaningful.
Recommendation: Keep the Spine, Buy the Muscle
The evidence points to a clear recommendation: organisations should use large integrators for delivery strength, method and experience, but retain internal ownership of strategic intent, operating model choices, benefit ownership, sequencing and governance rights.
This requires a stronger client than many organisations currently have. It also requires humility. Executive teams must admit when they lack the capability to direct transformation, then build that capability deliberately rather than conceal the gap behind a branded method.
The practical test is simple. At any point in the programme, ask: if the integrator left the room, could the executive team still explain why this sequence, this scope and these benefits are right? If the answer is no, the programme has become vendor-led in the dangerous sense. It may still be busy. It may still be professionally managed. It may even be on budget. But it has lost its strategic spine.
In the post-boom environment of 2001, capital is less forgiving and executive patience is thinner. The organisations that will benefit from technology-enabled transformation are not those that buy the most complete external method. They are those that recover the discipline of making their own choices before asking others to implement them.