When the Dashboard Went Green and the Delivery Went Missing
It does not give you a partial view of delivery; it gives you a confident view of something else, dressed in the language of assurance.
Executive Summary
Somewhere in the past three years a great many organisations stopped managing delivery and began managing service levels instead — and largely failed to notice the difference. The outsourcing wave that promised to convert unpredictable internal delivery into contractually guaranteed outcomes has, in practice, produced a subtler and more troubling shift: the service-level agreement has quietly become a substitute for delivery management rather than an instrument of it. This essay argues that the substitution is not an accident of careless contracting but a structural consequence of how sourcing decisions were justified, how retained organisations were designed, and how programme governance now rewards the appearance of control over its substance. The all-green dashboard is not a lie told by suppliers. It is an accurate report about the wrong thing. Reclaiming delivery management does not mean abandoning service levels; it means refusing to let them stand in for the harder, less glamorous work of owning an outcome that no contract can fully describe.
The Comfort of the Green Dashboard
Picture the monthly service review, a scene now repeated in thousands of organisations. The supplier’s account manager presents a deck. Availability was 99.6 per cent against a target of 99.5. Incident resolution times were within tolerance. The change calendar was honoured. Every indicator on the scorecard is green, and where one amber crept in, a remediation action is already logged against it. No service credits are due. The meeting closes early. And somewhere on the same floor, the business sponsor of the programme this service is meant to support is quietly furious, because the thing they are trying to achieve is going nowhere.
Both parties are telling the truth. That is what makes the situation so hard to name. The supplier is meeting the contract, and the dashboard proves it. The sponsor is not getting delivery, and their experience proves that. The gap between these two truths is the subject of this essay, and it is not a gap that a better set of metrics will close — because the problem is not that the metrics are wrong. The problem is that we have allowed the measurement of the contract to replace the management of the delivery, and the two are not the same activity, performed by the same people, aimed at the same thing.
We should be honest about how comfortable this arrangement is. A green dashboard is a form of institutional reassurance. It lets a retained manager report upward that the service is under control. It lets a supplier demonstrate compliance. It lets a governance board minute that all is well and move to the next item. Everyone in the room has an interest in the green, and almost no one in the room is accountable for the outcome the green was supposed to represent.
How the Substitution Happened
The substitution did not arrive by decision. Nobody stood up in 2002 and proposed that the organisation stop managing delivery. It happened by increment, and each increment looked sensible at the time.
When a function was outsourced, the work of doing it left the building. What remained was the work of managing the people now doing it — the retained organisation. That retained function needed something concrete to manage against, and the contract provided it: the schedule of service levels, the credit regime, the governance calendar. These were tangible, defensible, auditable. Delivery management — the messy, judgement-laden business of understanding what actually needs to happen and steering a diverse set of contributors toward it — was neither tangible nor easily auditable, and so it slowly ceded ground to the thing that was.
- First, the measure became the target. The service levels were written as a floor beneath acceptable performance, but they rapidly became the definition of it. Anything green was, by definition, fine.
- Then the target became the conversation. Governance forums organised themselves around the scorecard, because the scorecard was what both sides had agreed to discuss. Delivery questions that fell outside the schedule had no natural home on the agenda.
- Finally the conversation became the capability. A generation of retained managers learned their craft entirely within this frame. They grew skilled at reading service reports, negotiating credits, and running the governance machine — and correspondingly deskilled at the older discipline of owning an outcome from end to end.
By the third step the substitution is complete and invisible. The organisation still believes it is managing delivery. What it is actually managing is conformance to a contract that describes only the part of delivery that could be written down in advance.
The Structural Forces That Sustain It
If this were merely a training problem it would be easy to fix. It is not, because several structural forces actively hold the substitution in place. Each one, examined alone, looks like good practice.
- The business case created the incentive. Sourcing decisions were justified on cost, and the case rested on a clean transfer of responsibility. To then staff a strong delivery-management capability on the retained side looks, to a finance director, like paying twice — buying the service and then paying people to do the job the service was meant to remove. So the retained function is kept deliberately thin, and thin functions default to monitoring rather than managing.
- The contract is the only shared language. The client and the supplier are separate organisations with divergent interests. The one thing they have genuinely agreed is the schedule of service levels. It is therefore the only ground on which they can meet without friction, and so every conversation gravitates toward it.
- Auditability rewards the proxy. A service level can be measured, evidenced, and defended to an auditor or a regulator. Delivery judgement cannot. In an era increasingly attentive to control and evidence, the measurable thing crowds out the important thing simply because it can be shown.
- The credit regime misdirects attention. Service credits are meant to align the supplier with the client. In practice they focus both parties on the avoidance of penalty rather than the achievement of outcome. A supplier managing to avoid credits and a client managing to claim them are, between them, managing everything except whether the work is any good.
- Green is safer than honest. For the retained manager, a green report is career-safe and an amber one invites scrutiny. The structural asymmetry of consequence quietly discourages the very honesty that delivery management depends upon.
Every force sustaining the substitution is individually defensible. That is precisely why it is so durable: no single decision looks wrong, and the failure becomes visible only in the aggregate, at the level of the outcome nobody owns.
What a Service Level Cannot See
At the heart of the matter is a category error. A service level describes the performance of a defined service. Delivery describes the achievement of an intended outcome. These overlap, but they are not the same, and the space between them is exactly where transformation succeeds or fails.
| What the Service Level Sees | What Delivery Management Must See |
|---|---|
| Was the service available and within tolerance? | Did the available service actually advance the outcome? |
| Were incidents resolved in time? | Were the right things being worked on at all? |
| Was the change delivered as specified? | Was the specification still the right thing to build? |
| Is the supplier compliant with the contract? | Is the supplier’s effort aligned with a goal the contract never fully described? |
| Are we avoiding service credits? | Are we building the capability the organisation will need next year? |
Read down the right-hand column and a pattern emerges: every question requires judgement about the outcome, and none of them can be answered from the scorecard. A service can be perfectly available and entirely beside the point. A change can be delivered exactly to specification when the specification itself was the error. The service level, by construction, cannot see any of this, because it was written before the outcome was fully understood — and delivery is precisely the process by which understanding of the outcome changes.
This is why treating the service level as a proxy for delivery is not merely incomplete but actively misleading. It does not give you a partial view of delivery; it gives you a confident view of something else, dressed in the language of assurance. The greener the dashboard, the more complete the illusion.
The Retained Organisation That Forgot How to Deliver
The most damaging consequence is not any single failed programme. It is what happens to the organisation’s own capability over time.
Delivery management is a craft, and crafts atrophy when they are not practised. When the retained organisation is designed to monitor rather than to manage, the people in it stop making the kind of decisions that build delivery judgement. They no longer wrestle with sequencing, with the trade-offs between scope and time, with the reading of a team that is quietly in trouble long before any indicator turns amber. Those muscles waste. And because the work has left the building, there is no longer a junior population learning the craft from the ground up. A decade of this and an organisation discovers it has forgotten how to deliver anything it has not bought — a profound strategic vulnerability disguised as an operating model.
I have come to believe that the true cost of the substitution is paid here, and paid late. It does not appear in the year of the outsourcing decision, when the savings are real and the dashboards are new and green. It appears years later, when the organisation needs to do something genuinely new — something no existing contract covers — and finds that the people who might once have led it now know only how to hold a supplier to account.
Reclaiming Delivery Management
None of this is an argument against outsourcing, and none of it is an argument against service levels. Service levels are a reasonable way to describe the performance of a stable, well-understood service. The error is not in having them; it is in mistaking them for the whole of the management task. Reclaiming delivery management means restoring the discipline the proxy displaced, alongside the contract rather than instead of it.
- Name the outcome owner, and make it a person, not a scorecard. Every significant piece of outsourced delivery needs a named individual on the client side whose accountability is the outcome, not the service level — someone whose job is explicitly to ask the right-hand-column questions and who is measured on the answers.
- Fund the retained capability honestly. Accept in the business case that managing an outcome across a contractual boundary requires real delivery skill on the client side. A retained organisation staffed only to read reports is not a saving; it is a deferred cost.
- Separate the two conversations. Keep the service review for what it does well — the health of the defined service — and hold a distinct, differently framed conversation about whether the outcome is actually advancing. Do not let the green scorecard close the meeting before the harder question is asked.
- Govern the outcome, not only the contract. The governance board should demand a view of the outcome that does not derive from the service report — an independent read that can be amber when every service level is green, and that is treated as legitimate when it is.
- Protect the craft. Deliberately keep enough real delivery work, and enough people practising it, that the organisation does not lose the ability to lead delivery it has not purchased. Treat delivery management as a capability to be preserved, not an overhead to be minimised.
The test of whether you are managing delivery or merely managing the contract is simple: can your governance report be amber when every service level is green? If it cannot, you are managing the proxy.
The Deeper Lesson
The story of the service level standing in for delivery is a particular instance of a more general failure, and it is worth naming plainly. Organisations under pressure will always prefer the measurable proxy to the unmeasurable substance, because the proxy can be reported, defended, and rested upon. The service level is only the current form of an old temptation: to mistake the instrument that describes control for control itself.
The outsourcing wave did not create this temptation, but it institutionalised it — by writing the proxy into a contract and then organising the entire retained function around the contract. The organisations that will get the most from their sourcing arrangements in the coming years are not those with the tightest service levels. They are those who remember that a service level was only ever meant to describe a service, and that delivery — the achievement of an outcome that keeps changing shape as you approach it — remains a thing that people manage, with judgement, and cannot be handed to a dashboard, however green.