Management Reporting& Information Systems
Figures drawn from the systems that hold them, defined once, validated properly and presented so that a decision can be made in the meeting rather than deferred to the next one.
The argument about the numbers is the problem. Not the chart.
Reporting projects fail when they begin with the visualisation. A dashboard built on measures nobody has agreed simply produces a faster disagreement.
We start with definitions. What exactly counts as a completed job? Does a cancelled order appear in the month it was placed or the month it was cancelled? Is a part-time member of staff one head or half? These questions are tedious and they are the whole project.
Once the definitions are settled and written down, we build the mechanism that produces those measures from source, validates them, and presents them the same way every period.
What this service is built to remove.
The monthly pack takes several days to assemble.
Manual extraction and re-keying is a recurring cost and a recurring risk. Automating the assembly also removes the transcription errors nobody catches.
Two departments report different figures for the same measure.
This is nearly always a definition problem. Settling the definition in writing usually resolves it permanently — and often reveals that both were right under different assumptions.
By the time we see the numbers, the month is over.
Reporting cadence should match operational cadence. Work managed weekly needs a weekly view, not a retrospective produced three weeks later.
The report is produced by one person who is now on leave.
Where reporting depends on an individual's spreadsheet knowledge, it is a single point of failure. Making it systematic removes that dependency.
Nobody trusts the figures enough to act on them.
Trust comes from traceability. If a total can be opened to show the underlying records, challenge becomes verification rather than argument.
We have a dashboard, but nobody looks at it.
Usually because it shows what was easy to measure rather than what people are accountable for. Starting from the decision, not the data, fixes this.
Components of a reporting build.
The quotation states which of these are in scope for your engagement.
Multiple sources to a single decision.
The validation stage is the one most often skipped, and the one that determines whether anyone believes the output.
Where this service is normally applied.
Five stages, definitions first.
Stage one is the one clients are most tempted to shorten, and the one that determines whether the finished report is believed.
- Stage 01Decisions and definitionsWhich decisions the reporting must support, then a written definition for every measure. Output: an agreed measure dictionary.
- Stage 02Source assessmentWhere each measure can actually come from, how reliable that source is, and what has to change if it is not good enough. Honest findings, including “this cannot be measured yet”.
- Stage 03Consolidation and validation buildAutomated collection, transformation and validation rules, tested against periods you already have figures for so the output can be reconciled.
- Stage 04Presentation design and reviewDashboards and reports designed around the actual meeting or working routine, then reviewed with the people who will use them.
- Stage 05Parallel run and handoverOne or two periods produced alongside your existing method to prove agreement, then handover with documentation and a support window.
What you hold at the end.
- A written measure dictionary owned by your organisation
- A working reporting system with scheduled refresh
- Management dashboard and operational views
- Validation rules and an exception report
- Reconciliation record from the parallel run
- Documentation, export routes and a support window
And who it is not for.
A good fit
- Organisations spending days each month assembling a management pack
- Boards or management teams receiving figures too late to act on
- Businesses where the same measure is calculated differently by different teams
- Operations with data in several systems that never meet
- Organisations with recurring statutory or funder reporting obligations
Probably not a fit
- Organisations with no consistent underlying operational record yet
- Requirements for advanced statistical modelling or forecasting science
- Situations where no one is willing to settle the measure definitions
- Cases where an existing reporting tool is adequate and simply unconfigured
If the underlying operational data is inconsistent, reporting cannot fix it. In that situation we would normally recommend an operations system first, or an IT systems review to establish why the data is unreliable.
Often taken alongside this one.
Frequently asked.
In most cases yes — through a documented interface, a database connection, a scheduled export or, where nothing else exists, a structured file drop. Stage two of the engagement establishes exactly what is possible for each source before anything is promised.
Where a system genuinely cannot release its data in any reliable form, that is an important finding. It affects what that product costs you in the long run, and it belongs in the record.
Sometimes you should not. If a spreadsheet is produced in twenty minutes by two people who both understand it, that is a reasonable arrangement and we will say so.
Replacement is worth it when the assembly takes days, when it depends on one person, when the figures are disputed, or when errors have already caused a poor decision. Those are the conditions under which the build pays for itself.
Where historical records exist and are consistent, we load them so trends start from day one rather than from launch. Where the historical basis differs from the new definitions, we will say so and mark the point of change explicitly rather than presenting an artificially smooth series.
You do. We facilitate the conversation, ask the awkward clarifying questions and write down what is agreed — but the measures belong to your organisation. Our contribution is making sure that each one is unambiguous, obtainable and actually connected to a decision somebody makes.
Yes, and they usually do. The measure dictionary is a living document. When a definition changes we record the change and the date it took effect, so historical comparisons stay honest instead of quietly shifting under the reader.
Rarely, at this scale. Most organisations of the size we work with are well served by a straightforward consolidated store with scheduled refresh. We do not recommend large data platforms for reporting requirements that do not need them — the running cost and the complexity outlast the enthusiasm.