You hold IT suppliers accountable by deciding what good looks like before you sign, writing it into the contract and statement of work as measurable outcomes, chairing a regular steering meeting, controlling change requests through a clear process and always keeping a realistic exit option. Accountability is not about being tough with suppliers. It is about governance that makes performance visible and gives you real choices.

Key takeaways

  • Most supplier problems start at the contract: vague scope, activity-based billing and no outcome measures.
  • Mix service KPIs with business outcome KPIs, and keep the list short.
  • The client, not the supplier, should chair the steering meeting and own the agenda.
  • Change requests are where budgets leak. Every one needs a business reason, an owner and a decision.
  • Plan the exit on day one: data, documentation, knowledge and notice terms.

Why supplier accountability breaks down

In my experience it is rarely because suppliers are acting in bad faith. More often the client does not have senior capacity to govern the relationship, so the supplier fills the vacuum. They write the roadmap, define the priorities, report on their own performance and propose the change requests that extend their scope. Each step is reasonable. Together they shift control.

I have seen this from both sides. Having founded and run an IT consulting and implementation business since 2013, I know how suppliers think, what they need from a client to do good work, and where the commercial incentives can pull away from the client's interests. Good suppliers welcome clear governance. It makes their work easier to plan and their value easier to prove.

Contracts and statements of work

The contract sets the legal frame. The statement of work (SOW) is where accountability actually lives.

What a good SOW contains

  • The business purpose. One paragraph on why this work exists and what it should change for the business.
  • Scope in and out. What is included, and just as important, what is explicitly excluded.
  • Deliverables with acceptance criteria. Not "configure the CRM" but what must work, for whom, tested how.
  • Roles and responsibilities. Including what the client must provide and by when. Many delays are client-side dependencies nobody wrote down.
  • Named key people. With a right to approve replacements for critical roles.
  • Commercial model. Fixed price, time and materials or a mix, with clear rules for each.
  • Outcome KPIs and reporting. See below.

Watch the commercial model

Pure time and materials rewards effort, not results. Pure fixed price can push a supplier to cut corners or fight every change. A common healthy pattern is fixed price for well-defined phases, time and materials with a cap for exploratory work, and a portion of fees tied to agreed outcomes where that is realistic.

Outcome-based KPIs

Most supplier reports measure activity: tickets closed, hours spent, sprint velocity. These have their place, but they do not tell you whether the business is better off.

Use two layers, and keep each short:

Service KPIs (three to five). Response and resolution times by priority, availability of critical systems, on-time delivery of agreed milestones, defect rates after release.

Outcome KPIs (two or three). Linked to the purpose of the work. For an ERP rollout, it might be days to close the month or order processing time. For a customer service platform, first-contact resolution. For a cloud migration, the cost and stability of the migrated workloads. If you are planning one of these, modernization and cloud covers the governance side in more detail.

Agree the baseline before work starts. Without a baseline, outcome KPIs turn into opinions.

Steering cadence

Governance is a rhythm, not a document. A typical structure:

  • Weekly working meeting. Delivery leads on both sides. Progress, blockers, decisions needed at this level.
  • Monthly steering committee. Chaired by the client's accountable owner. KPIs, risks, change requests, budget versus plan, escalations.
  • Quarterly business review. With senior leadership on both sides. Is the relationship delivering what we need? What should change in the next quarter?

Two rules make the difference: the client owns the agenda and the minutes, and every meeting ends with decisions and owners, not just updates.

Change requests

Change requests are where budgets and timelines quietly drift. Each one may be justified, but together they can double the cost of a programme without anybody deciding that on purpose.

A simple discipline helps:

  • Every change request states the business reason, the cost, the impact on timeline and on other work, and who asked for it.
  • Changes above a threshold go to the steering committee. Below it, a named owner decides.
  • Keep a cumulative log. Review the total, not just each item.
  • Ask whether the change is needed now, or whether it can wait for a later phase.

Lock-in and exit

You do not need to plan to leave a supplier, but you need to be able to. That option is what keeps the relationship balanced.

  • Data. You own it, you can export it in a usable format, and the contract says so.
  • Documentation. Configurations, integrations and customisations are documented as part of the deliverables, not as an optional extra.
  • Intellectual property. Clear terms on who owns custom work.
  • Knowledge. Your own people or an independent party understand the systems well enough to brief a new supplier.
  • Exit assistance. A contractual obligation to support transition, with defined terms.
  • Notice periods. Reasonable and symmetrical.

When to replace a supplier

Replacing a supplier is expensive, so it should be a decision, not a reaction. It is justified when:

  • Performance problems persist after clear, documented escalation.
  • Their commercial interests consistently conflict with yours, for example pushing licences or scope you do not need.
  • Key people keep rotating and knowledge keeps leaving with them.
  • They cannot support where your business is going, in technology or in scale.
  • Trust has broken down at senior level on both sides.

When you decide to change, plan the transition carefully: overlap periods, knowledge transfer, data migration and a clear communication plan. The worst outcome is a rushed switch that keeps the same governance problems.

The supplier governance checklist

  • Does every major supplier have a SOW with business purpose, scope in and out and acceptance criteria?
  • Do we have three to five service KPIs and two or three outcome KPIs, with baselines?
  • Does a named person on our side own each supplier relationship?
  • Do we chair the monthly steering meeting and own the minutes?
  • Is there a change request process with thresholds and a cumulative log?
  • Can we export our data and do we have up-to-date documentation?
  • Is there an exit assistance clause and a reasonable notice period?
  • Have we reviewed each major supplier's performance in the last quarter?

Much of this checking can now be done quickly. I use AI agents to compare SOWs against delivery, reconcile change logs and flag gaps in contracts, which turns what used to be weeks of review into days. The judgement and the conversations with suppliers stay with people, and that is where they belong.

Closing thoughts

Strong supplier governance is one of the highest-return disciplines a company can adopt. It protects budgets, improves delivery and, in my experience, makes the relationship better for good suppliers too. If you want to see how I approach this as an ongoing role, vendor governance explains the scope. For programmes that have already drifted, why digital transformation programmes stall is a useful companion read.

If you are not sure whether your suppliers are delivering what you are paying for, send me a message through the contact form with a short outline of the situation, and I will tell you where I would look first.