Measuring is not enough: why your health score programme needs an implementation mandate

The familiar picture: dashboard green, customers gone anyway

Perhaps you know this situation. Some time ago, your company introduced a health score. Every customer has a traffic light, the report arrives punctually every month, with a list of customers at risk. Management knows who is wavering and why.

And yet exactly these customers cancel.

In the next quarter, the report lists the same reasons as in the last one: invoices the customer doesn't understand. Delivery promises that sales makes and order processing cannot keep. Service requests that get stuck between two departments. Nobody disputes these findings. None of them gets fixed.

We see this pattern again and again. It has little to do with negligence. The team that maintains the score measures, reports and escalates, and it does that well. It just isn't allowed to change the invoicing process, set rules for sales or release money for an adjustment in service. So every finding goes to another department as a recommendation. There it competes with day-to-day business and loses.

Across the market, too, little is moving. In Forrester's Customer Experience Index 2025, 73% of the brands studied worldwide stayed unchanged, 21% got worse and 6% got better. In Europe, 90% were statistically unchanged. The figures don't say why. But they fit the picture from the start: a lot measured, little improved.

Your programme doesn't lack a better score or another dashboard. What it lacks is a person or a team that is allowed to fix problems and has a budget for it. As long as someone is responsible for measuring and nobody for changing, the health score remains a reporting exercise. It reliably shows which customers are leaving. It keeps none of them.

Measuring and changing are two different assignments

Many health score programmes contain a silent assumption: whoever makes a problem visible also makes sure it goes away. This assumption does not hold. Measuring and changing need different authority, different resources and often different people.

The measurement assignment covers the score itself. Someone maintains the data, keeps the weighting up to date, reports which customers are slipping into risk and escalates when a value deteriorates. The result is information. Which signals should be escalated and what happens next is described in early warning signals with a clear action.

The implementation assignment begins where the report ends. Here a process in order processing changes, or a rule in the CRM, a template in customer communication or the responsibility for complaints. The result is a different workflow that the customer notices.

Established approaches to closed feedback loops separate these two roles explicitly. There, a coordinating function does not solve problems itself. It collects findings, merges duplicates, prioritises and passes them on to the team that has the resources and expertise for the solution. Each prioritised measure is assigned to a person or group, and they receive the necessary authority and resources. The same approaches note that such measures almost always cut across several functions and often require considerable investment.

It is exactly this second part that is missing from most project definitions. Only the measurement assignment is handed out, for three understandable reasons.

It is cheap. A score needs data access and some analysis time. A change needs a budget whose size nobody knows when the project starts.

It is free of conflict. A report states things; it demands nothing. Changing, on the other hand, means another department has to adjust its workflows, priorities or metrics.

It stays within its own area. The programme can run the score itself. The causes of the problems, however, usually lie elsewhere: in sales, in service, in accounting.

Why friction points lie between departments

Sales promises a delivery date. Order processing plans with different capacities. Service only hears about the delay when the customer calls. Accounting issues the invoice on the original date and later sends the reminder. Every department has done its part correctly. What reaches the customer is still an experience that makes them hesitate before the next order.

Most of the friction a health score makes visible arises at handovers: between quote and order, between delivery and invoice, between complaint and credit note. Information gets lost there because each team works with its own tool, its own goals and its own rhythm.

Whoever sees the friction point almost never owns the process in which it arises. The team that maintains the score recognises the late invoice correction as an early signal of churn. It cannot change it. The rule belongs to accounting, and so does the budget.

This is not a phenomenon peculiar to mid-sized companies. A Gartner survey of CX leaders from 2019 found that marketing is increasingly taking on responsibility for the customer experience, but that budget authority for it remains spread widely across different departments. Seven years later, a CX Dive report of 31 August 2026, based on a Forrester report from July, paints the same picture. In most organisations, budgets are still planned department by department. Product management, digital, service, operations and IT often influence the customer experience more than the CX team itself. The recommendation there: if the customer experience extends across functions, budget planning has to as well.

Without an explicit mandate, every finding is therefore read in the target department as a request, sits untouched and reappears in the next report.

The three symptoms of a programme without an implementation mandate

You can check whether your programme only measures or also changes things by looking at three symptoms, without a new metric, each with questions for the next leadership meeting.

Symptom 1: the same findings appear in every report

Lay the last few reports side by side. If the same topics keep coming up, such as late order confirmations, incorrect invoices or service requests left unanswered, then the measurement works and the implementation doesn't. A recurring finding shows that nobody had the assignment to fix it.

Questions to check:

• Which finding has been in the report for more than a quarter?
• Who is responsible for fixing it, and not just for observing it?
• Is there a date by which it is supposed to be fixed?

Symptom 2: measures are called "recommendations" instead of "decisions"

Pay attention to the wording. "It is recommended to review the handover process" is a request without an addressee. "The head of order processing will change the handover to service by the end of the month" is a decision. Whoever is only allowed to recommend writes recommendations.

Questions to check:

• How many measures in the last report have an owner and a deadline?
• Who is allowed to decide that a recommendation gets implemented?
• What happens if the affected area simply doesn't take up a recommendation?

Symptom 3: implementation fails on budget or priority in the target area

The finding is accepted, the measure makes sense, and still the answer is "no budget this year" or "other priorities". From the target area's point of view, that is understandable. The costs of the change arise there, while the benefit shows up elsewhere: in whether customers stay. This is the spread-out budget authority from the previous section, on a small scale.

Questions to check:

• Does the programme have its own budget for changes, or only for measurement and tools?
• Who decides when a measure in the target area competes with day-to-day business?
• Is customer retention among the goals of the areas that would have to change something?

How to read the result

If you cannot give a clear answer for one symptom, that is a warning sign. With two or three symptoms, the situation is clear: your programme has an assignment to measure but none to change. That can be fixed without setting the programme up from scratch.

What an implementation mandate actually sets out

A mandate is a short written agreement on who may change what for which finding, with what money and by when. The following five points translate the principle "responsibility plus authority plus resources" into decisions you have to make.

1. Who decides on which kind of change

Not every finding calls for the same intervention. Divide possible changes into kinds and name a decision-maker for each kind:

• Process changes, such as a new handover between sales and order processing
• Changes to tools and systems, such as a mandatory field in the CRM or a changed notification
• Rule changes, for instance on goodwill gestures, payment terms or escalation levels in service
• Customer communication, such as invoice texts, order confirmations or notices after a disruption

Your health score team can take on the coordinating role. The mandate then has to state that a referral is an instruction and no mere request.

Question to check: for each of these kinds, can a specific person say "I decide this" without having to ask anyone first?

2. Up to what amount changes are made without new approval

Many findings fail at the first euro. The mandate names an amount per individual measure up to which the person responsible decides without further approval, and the cost centre the money comes from. Where this money comes from in the first place is governed by the implementation budget in the next section.

The amount is specific to each company. It depends on revenue, number of customers and what a lost customer costs you.

Question to check: does the person responsible know up to what amount they may decide on their own today?

3. Which areas have to cooperate

Because friction points sit at handovers, their solutions cross departmental boundaries. The mandate sets out which areas are obliged to cooperate (such as sales, order processing, service, accounting and IT), what cooperation means (supplying data, providing contacts, allowing changes in their own area) and how much working time these areas plan for it.

Question to check: can a head of department reject a change in their area without anyone noticing? If so, the mandate is too weak at this point.

4. Which deadlines apply from finding to implementation

Without a deadline, every finding becomes a permanent topic. Set two points in time: by when it is decided whether and how the finding is tackled, and by when the change takes effect. A revised email text needs less time than a new handover between two departments.

Question to check: is compliance with these deadlines measured anywhere?

5. When to escalate, and to whom

Escalation is not a sign of failure. It is what makes the mandate enforceable. Triggers are a missed deadline, a budget limit that is too small, an area that refuses to cooperate, or two areas that cannot agree. The recipient is a named person with decision rights over all the areas involved, which in mid-sized companies is usually management.

Question to check: does the escalation point also have a deadline for its answer? Without one, the backlog simply moves one level up.

Once answered, the five points fit on one page. That is intentional.

Budget belongs in the project definition

Many health score programmes start with a budget for software, data integration and reporting, but without money for implementation. The assumption: if the score shows a problem, the funds will follow. They rarely do.

Why budget requested after the fact fails

The report shows that customers wait too long for feedback after the order confirmation. The team proposes adapting a template in the CRM and introducing a handover rule between sales and order processing. That costs IT time and a little external support. So someone writes a budget request.

From that moment on, the change competes with the trade fair stand, the warehouse extension and the next ERP update, usually without an advocate on the budget committee. Each individual change narrowly loses this comparison. In sum, the programme loses.

Whoever has the finding doesn't have the money. Whoever has the money has other priorities.

What a reserved implementation budget changes

When you define the project, set aside part of the funds for the changes the score will make visible in the coming months, without a project fixed in advance. The amount per measure from the mandate governs who controls the money. The budget makes sure it is there at all.

That changes the pace: a change within the limit needs no new budget request, and the time from finding to change shrinks from budget cycles to working weeks. It changes the commitment: if money is available, the report has to state what was implemented. And it changes the nature of the programme: the reporting exercise becomes an operating routine of signal, decision, change and review.

How large the budget needs to be depends on the number of your customers, the state of your systems and the number of handovers in your customer journey.

The price, stated openly

A reserved budget ties up funds before you know what for. In the first quarter there may be money available that nobody draws on, because the score is still settling in. A controller will rightly ask why a budget should be approved without a concrete list of measures. A separate article covers how the return on a measure can be made provable before it starts.

Three ways limit this risk:

1. Start small and tie it to rules: only for findings from the health score, only where at-risk customers are involved.
2. Agree on a return flow: unused funds go back into the general budget on a fixed date.
3. Report on usage: the report shows which finding was fixed with which amount. After a few cycles you have your own basis for adjusting the limit.

The alternative is not cheaper. You then pay for measurement, reports and coordination meetings without anything changing in the customer journey.

Who should hold the mandate, and who should not

Most mid-sized companies do not have a dedicated CX team of ten people. The health score sits somewhere between sales, marketing and finance. That seems practical. But this is exactly where it is decided whether findings get fixed.

Why analytics and marketing rarely enforce the mandate

The programme often ends up with analytics or marketing. The task fits there, but the clout is missing. An analytics team can show that customers drop off after a late order confirmation. It is not allowed to change the workflow in order processing. Marketing can send an apology email. The cause in invoicing remains untouched. Whoever carries only the responsibility can report, but cannot order anything.

Why management has to be the sponsor

Friction points arise between departments, so you need someone who stands above them. In mid-sized companies, that is almost always management. It doesn't have to run the programme operationally. As sponsor, it grants the mandate explicitly, releases the implementation budget and decides when two areas block each other. Without this backing, every allocation of authority and resources stays on paper.

What a small journey team with cross-functional rights looks like

You don't need a new department. A small, fixed group is enough:

• a coordinating role that collects findings, removes duplicates, sorts by impact and hands them over to the responsible area
• one responsible person from each affected area, such as sales, order processing, service and accounting, with the right to decide on changes from the programme in their area
• a fixed meeting with management in which blockages are decided rather than discussed

Record these rights in writing, ideally directly in the mandate.

The most common trap: authority without time

Often a committed person takes on the mandate alongside their job, such as the head of sales or whoever introduced the health score. On paper they have all the authority; in their calendar, not a single hour for it. Nothing is then rejected. It simply sits there.

So check capacity as well as authority. Without dedicated time, the mandate has only been assigned on paper. In that case, free up time or assign it differently.

How to retrofit a running programme

Score, data sources and reports stay as they are. All that is missing is the link between finding and change, and you can add it in four steps.

1. Collect open findings. Lay out the reports from the past few months side by side and write down every finding for which a measure was recommended but never implemented. Merge duplicates and mark topics that recur several times. The list will usually be sobering. That is exactly why it is useful.

2. Sort by impact and effort. For each finding, ask: how much does it contribute to churn? What does fixing it cost in time, money and coordination? High, medium and low are enough; if data is missing, mark the rating as an estimate. Among the high-impact findings, prioritise those with manageable effort, because they quickly show that the programme now changes things.

3. Name owners for the top findings. Not all at once. Each of the top findings gets a person or team with decision rights and budget. Record in writing what may be decided without new approval, which budget pays for it and by when the change is supposed to take effect.

4. Turn the agenda around. Keep your reporting meeting, but put the status of the assigned changes at the top: what has been implemented, what is stuck, and why? Anything stuck twice in a row goes to management. The score moves to second place and shows whether the changes are reaching customers.

Some findings will fall through, because their effort is out of all proportion to their impact. Decide that deliberately and take them out of the report. A finding that has been explicitly rejected is more honest than one that is back on the list every quarter.

How you can tell it is working

You won't see it in the health score itself first. That can stay unchanged for a while, even as a lot is already moving underneath. Three criteria from your own reports are more telling. The third is symptom 1 from above, this time read as a measure of success.

1. Share of findings implemented. For each reporting period, count how many findings were actually implemented: a changed process, an adjusted rule, a newly arranged handover. As long as this figure is missing from the report, the programme mainly measures itself. Pairs of activity and outcome figures are explained in a separate article.

2. Time from finding to change. For each finding, record when it first appeared in the report and when the change took effect. The yardstick is your own track record: the span should get shorter.

3. Fewer recurring topics. If the perennial issues disappear from the reports and make room for new topics, the programme is working.

Keep track of the three criteria properly. Then you will see in your own customer base what the mandate achieves.

A health score is worth as much as the authority to act on it. Who in your company is allowed to act on it today, and with what money?

If you want to check where your programme stands, the Reactivation Audit is a place to start. UNIT ONE works with mid-sized companies to make findings from health score and journey programmes actionable.

Frequently asked questions

What is an implementation mandate in a health score programme? A short written agreement on who may fix the problems the score makes visible, with what money and by when.

Who in a mid-sized company should have the authority to fix problems in the customer journey? Management as sponsor, because only management can set priorities across departmental boundaries. Operationally, a small journey team with dedicated time for this task is enough.

How much budget does a health score programme need for implementation? That depends on the number of customers, systems and handovers. What matters is the timing: plan the implementation budget into the project definition, instead of requesting each change individually after the fact.

Sources

• Forrester: Forrester's 2025 Global Customer Experience Index Rankings (press release, June 2025). forrester.com
• MarTech: report on the Gartner CX Management Survey 2019 (January 2020). martech.org
• CX Dive: report on CX budget planning with reference to a Forrester report from July 2026 (31 August 2026). customerexperiencedive.com