# The Operating Model for Digital Sourcing: Governance Beyond the Award > No single function is qualified to judge a digital deal alone. The operating model that follows — distributed decision rights, staged gates, a standing vendor-management home — and the strategy for a greenfield or major transformation. - Author: Peter Rogov (https://peterrogov.com/about) - Published: 2026-07-12 - Topics: IT Governance, Procurement, Digital Transformation - Canonical URL: https://peterrogov.com/blog/operating-model-for-digital-sourcing-governance-beyond-the-award --- Sourcing digital capability well requires judging architecture fit, data models, methodology as practiced, team quality, and roadmap credibility. Which raises a structural question: who in the traditional process is qualified to judge them? ## The single-owner relay, and why it strains The traditional model is a sequential relay: the business or IT raises a demand, perhaps drafts the RFP, and hands it to procurement; procurement runs the tender, owns the evaluation, and awards; IT reviews the technical annexes somewhere in between. For commodities and standard equipment this works, because evaluation reduces to comparing offer attributes a trained buyer can score. For digital it strains, because the decisive evidence lies outside any single profession. A generalist procurement team should not be asked to assess a sprint cadence, a data model, or an integration architecture. When it is, the technical dimensions are either reduced to checkbox compliance — reproducing "compliant but not fit for purpose" — or the award is made on the dimensions procurement can score — offer attributes that are weakly connected to lifetime value. Mature organizations already mitigate this with technical scoring panels; the argument here is to make that mitigation the explicit design rather than an accommodation. This is not a criticism of procurement professionals; it is an operating-model design question, and the analyst community has been explicit about it. Gartner's sourcing, procurement and vendor management (SPVM) research argues that IT procurement's traditional value proposition is eroding, and prescribes engagement models focused on enabling business outcomes rather than running processes, with procurement repositioned as the organization's essential commercial resource for IT. Birmingham's auditors reached a compatible conclusion from the other side: inadequate program governance and a shortage of in-house expertise sat at the top of their findings. The naive fix — letting IT or the business buy on their own — fails in the opposite direction: ungoverned business-led buying is how organizations accumulate SaaS sprawl, duplicate contracts, and unmanaged risk, and commercial rigor, negotiation leverage, and contract discipline are procurement's genuine and irreplaceable expertise. The answer is not to move the decision from one insufficient owner to another. It is to stop having a single owner at all. ## Four shifts define the target sourcing operating model ### 1. Procurement moves from process owner to commercial partner It orchestrates the sourcing process and owns the commercial dimension outright — market and pricing intelligence, negotiation strategy, competitive tension, contract discipline, process integrity. What it gives up is sole authority over evaluation and award. Procurement's authority shifts from owning the decision to owning the deal. ### 2. Technology becomes a specialized category with specialized people Gartner treats IT SPVM as a distinct leadership discipline, not a branch of general purchasing — Dovgalenko's handbook exists precisely because technology buying is a craft of its own. Build or hire technology-literate sourcing professionals: enough commercial depth to negotiate a SaaS agreement, enough technical fluency to know what questions the architects should be answering. ### 3. Evaluation authority is distributed to where the qualification sits Replace the relay with a standing, cross-functional deal team operating under an explicit decision-rights model: the business owns outcomes and requirements; enterprise architecture owns platform fit and standards, with a veto on architectural grounds; IT owns the technical and methodology evaluation; security and legal own their risk domains; procurement owns the commercial evaluation and the integrity of the process. The award is a joint gate decision under executive sponsorship — no single function signs alone. Distributed rights need a tie-break: specify an escalation path (normally the executive sponsor or steering body) with authority to resolve a deadlock — say, between an architecture veto and a business imperative — within a defined time, or the design trades single-owner error for multi-owner stalemate. ### 4. Governance shifts from the award to the lifecycle Because commitment is staged, governance must be too: investment gates at business case, pilot, scale-up, and each renewal, owned by an executive steering committee, with the governance mechanics written into the contract itself — terms Gartner reports SPVM leaders regularly omit, with poor decision-making as the direct result. Post-award, a standing vendor-management function becomes the organizational home of the relationship, and a named business owner remains accountable for the benefits case, not just the system. The test is simple to state. For every judgment the sourcing process requires, ask who is professionally qualified to make it — and whether they hold decision rights at that stage. Where the answers diverge, the process carries avoidable risk. ## The digital sourcing strategy for a greenfield or major transformation ### 1. Architecture before procurement Define the target digital architecture — operational backbone versus digital platform layer — before any tender, and give every purchase a defined place in it, with integration and data standards non-negotiable. MIT research (Ross, Beath & Mocker, *Designed for Digital*) argues that established companies fail at digital for lack of these foundations, not lack of tools. A greenfield operation can get the architecture right before legacy exists — its single greatest advantage, squandered by buying opportunistically. ### 2. Match ambition to organizational maturity A greenfield brings clean architecture but also immature processes, thin staffing, and no vendor-management muscle — and an understaffed function running a sophisticated buyer's process produces the ceremony of rigor without its substance. An immature buyer de-risks by reducing what each purchase demands of the organization: standard configurations over ambitious ones, discretionary purchases sequenced behind the backbone, independent expertise rented for the gaps (advisors who are not the system integrator grading its own work), and a minimum viable governance core built ahead of the commitment curve — one accountable executive owner, a short set of architecture principles, and a gate with genuine authority to stop. ### 3. Buy outcomes, not specifications — proportionately Express requirements as business scenarios and measurable outcomes, run [an evidence-based selection funnel](/blog/selecting-technology-vendor-evidence-over-offers) — scenario demonstrations, paid pilots, staged elimination — for strategic platforms and a lighter version for lower-stakes purchases, adopt platform standards by default, and reserve customization for genuine differentiators — the central lesson of the Lidl case. (The 2026 HBR argument that AI increasingly lets software adapt to the company softens this trade-off at the margins; it does not yet repeal it for core platforms, where customization debt remains the dominant risk.) ### 4. Evaluate expected lifetime value, not the offer Multi-year total cost of ownership (TCO) including exit; vendor trajectory and ecosystem; data model, migration, and interoperability weighted heavily. Then negotiate price — in that order. ### 5. Contract for change; protect the exits Stage commitments with gates instead of big-bang fixed scope — the MillerCoors–HCL dispute shows where the alternative can lead. Spend negotiating capital on the clauses that govern a decade of dependency: data ownership and portability, exit assistance, escrow where warranted, renewal caps, enforceable service levels, security and audit rights. ### 6. Resource adoption and benefits as a workstream, not a footnote If value materializes through adoption, budget and staff change management (process redesign, training, super-users, communications) on the same footing as the technical build; assign a business owner to the benefits case with reporting obligations that survive go-live; and treat data migration and data quality as a governed workstream from day one. Birmingham's auditors identified under-investment in exactly these areas as central to the failure. ### 7. Stand up the governance model and its permanent home Implement the distributed decision-rights model above before the first major tender, and give the post-award relationship a standing owner: a vendor-management function handling renewals, subscription optimization, and performance and risk reviews for the life of each platform. Integrators execute; the organization owns architecture, governance, and change. ### 8. Keep build-vs-buy live AI is making composed and bespoke solutions viable for non-tech enterprises. You need not become a software company — but re-ask the question for each capability rather than assuming a packaged purchase, and apply the same lifecycle-value lens to what you build as to what you buy. ## The bottom line Your procurement discipline isn't wrong, and it was never merely about price. It is a sophisticated machine for optimizing specified value — comparing offers across known parameters and selecting the best combination. The difficulty is that digital value largely does not sit in the offer: it is emergent, co-created, and dynamic — it materializes through adoption and architecture, compounds through roadmaps and data, and erodes through vendor lock-in. [Lidl, Birmingham, MillerCoors, and Hertz](/blog/buying-technology-like-steel) each selected an offer that scored well on the day; what failed was everything the score could not see — fit, governance, change, and the decade after signature. The organizations that get this right do not abandon procurement rigor. They redirect it — from evaluating offers to maximizing expected lifetime value — and they pair it with the governance and change management that every post-mortem shows carry equal weight, with process, criteria, and contracts rebuilt around that objective and scaled to the stakes of each purchase. For a greenfield operation or a digital transformation program, this is not a procurement footnote. It is a strategic choice that will shape cost base, agility, and data position for twenty years. Buy accordingly. :::references{title="Sources — case references"} 1. Lidl/SAP: *Handelsblatt* (2018) on Lidl's eLWIS program. 2. Birmingham/Oracle: Grant Thornton, report on Birmingham City Council's Oracle implementation (February 2025). 3. *MillerCoors LLC v. HCL Technologies Ltd.*, U.S. District Court, N.D. Illinois (filed 2017; settled 2018). ::: :::references{title="Further reading — books & research"} 1. Gartner, IT Sourcing, Procurement and Vendor Management (SPVM) research, incl. *Leadership Vision for 2025*, the annual *SPVM Operating Model Primers* (2023–2025), and *Governance Framework to Manage IT Service Engagements* (2023) 2. Jeanne Ross, Cynthia Beath & Martin Mocker, *Designed for Digital* (MIT Press, 2019) 3. Sergii Dovgalenko, *The Technology Procurement Handbook* (Kogan Page, 2020) 4. Deep Nishar & Nitin Nohria, "The End of One-Size-Fits-All Enterprise Software," *HBR*, April 2026 :::