The Unbilled Hours: Auditing the Reporting Work That Sits Outside Delivery

Most organisations can account for the hours spent delivering a service. Far fewer can account for the hours spent describing it. A monthly return, a weekly dashboard, a tracker maintained for a meeting that no longer exists: each looks modest in isolation. Together they form a second operating model, one that consumes capacity without appearing in delivery plans, project budgets or resilience assessments.

That matters because operational resilience depends on spare capacity and clear priorities. When front-line teams spend their day assembling figures for internal audiences, the organisation is paying twice: once for the work and again for the management attention used to review it. The reporting burden deserves the same scrutiny as any other process cost.

The reporting estate is larger than the report list

Begin by asking teams to list requests, not reports. A report is a document; a request is any occasion where someone must find, clean, check, format or explain data for another person. That distinction surfaces the informal work that sits between formal outputs: the chase for a missing figure, the reconciliation before a board pack, the spreadsheet rebuilt because two systems disagree.

A simple register can capture the estate without turning the audit into another burden. For each request, record the requester, the frequency, the source system, the time needed to prepare, the time needed to validate, the time spent discussing it, the decision it supports and any statutory or contractual obligation. Add a final column: when did this last change a decision? The answers are often revealing. Some returns are legally required. Others are repeated because they have always been repeated.

Cost the whole chain, not the visible task

A report that takes two hours to produce may take six hours to deliver once chasing, checking, formatting and pre-meeting preparation are included. To capture that chain, ask a representative group of teams to log data work in 15-minute blocks for two weeks. Separate the log into collection, cleaning, validation, formatting and explanation. Then convert the total into full-time equivalent hours and apply a loaded cost.

Use ranges rather than false precision. The purpose is not to produce an accounting artefact but to make an invisible load visible. If 40 people each spend three hours a week on internal reporting, the organisation is consuming 120 hours a week before any analysis begins. That is a resilience fact, not merely an efficiency one.

Separate obligation from habit

Not all reporting is waste. Statutory returns, regulatory notifications and contractual reports must continue. The useful test is to sort requests into three groups: required by law or regulation, required by a customer or funder, and required internally. The third group needs the strongest justification.

For internal reporting, apply four tests. First, the decision test: what decision does this inform, who makes it, and what would change if the number moved? Second, the duplication test: does the same information already exist in another format? Third, the latency test: does the report arrive in time to affect the decision, or does it merely document what has already happened? Fourth, the precision test: is the requested accuracy proportionate to the use? A monthly management meeting rarely needs daily granularity.

Where no convincing answer emerges, the request is a candidate for retirement. Where an obligation exists, the question shifts to cost: can the same assurance be provided with less manual effort, fewer formats or a longer interval?

Cut in sequence, not by decree

Simply cancelling reports creates shadow reporting. Teams continue to collect the data informally because they fear being asked for it without notice. A better approach is sequenced and explicit.

Start with a reporting register and a 90-day moratorium on new internal data requests. Any new request should name the decision it supports and the existing request it replaces. Consolidate dashboards so that one source serves several audiences. Review meeting agendas and remove pre-reads that are not discussed. Publish a short data request charter covering lead times, purpose and the single source of truth. Retire one low-value report per function each quarter, and track the hours reclaimed. Redirect a visible share of those hours to resilience work, such as cross-training, incident preparation or process improvement, so that the benefit is tangible rather than theoretical.

Make the audit repeatable

The first audit will always find the most obvious duplication. The harder task is preventing the estate from growing back. Add reporting hours to project plans and business cases. Ask programme teams to estimate the ongoing measurement cost of any new service. Include a standing item in operational resilience reviews: what did we stop measuring, and what did we start measuring, and why?

Leadership signalling matters. If the only question asked of a new dashboard is what it adds, the estate will expand. If leaders also ask what has been retired, teams will treat measurement as a managed cost rather than an automatic good.

The aim is not fewer numbers. It is fewer numbers that cost more than they return. Measurement should earn its place in the operating model, just as delivery does. Until reporting work is counted, it cannot be managed, and until it is managed, it will continue to draw capacity away from the work that customers and regulators actually see.