Data and Information

    Reporting that holds

    By Redaktion techport.ai, IT-Beratung · Last updated on

    In many mid-sized companies reporting is a monthly craft: exports from two or three systems, assembled in a spreadsheet that has grown over the years and that exactly one person understands. That works, but it costs days every month and produces figures nobody can trace back to the source.

    The way to better reporting does not run through a tool. It runs through the question of which decisions are to be made with which figures, and through a binding definition of those figures.

    How you notice it

    • The monthly report takes several working days and arrives when the month is long over.
    • Meetings debate whether the numbers are right instead of the matter at hand.
    • There are many dashboards but no decision that was made because of them.
    • When the responsible person is on holiday, there is no report.

    Why this happens

    Reporting grows out of individual requests. Someone needs a figure, someone builds an export, the export becomes routine. Nobody decides on definitions along the way, because the question seems unambiguous at the moment it is asked. Only when two areas calculate the same metric differently does the missing standard become visible, and by then the correction is uncomfortable because both sides have worked with their version for years.

    How we go about it

    1. Settle the questions before the tools. We collect the decisions that get made regularly and derive the metrics needed for them. Metrics without an associated decision do not get built, however interesting they might be.
    2. Fix the definitions. We define per metric the formula, data source, scope and refresh cycle, binding for the whole company. That collection is the most valuable part of the initiative and outlives any tool.
    3. Connect sources and automate. We build the path from source to report so that it works without manual steps, and document it. Where several systems come together we examine whether a dedicated analysis database is worthwhile.
    4. Enable self-service in measured doses. We provide the departments with reviewed data areas from which they can build their own analyses, and separate those clearly from the binding reports. That creates freedom without competing truths.

    What you gain

    • Reports that arrive on time and without manual work.
    • Meetings that discuss content rather than numbers.
    • Departments that can answer their own questions without blocking IT.

    From our projects

    The effort for the monthly report is almost always underestimated because it is spread across several people and nobody sees the total. When we add it up, it regularly amounts to several person days per month in mid-sized companies, permanently. That is the economic core of the topic, not the nicer presentation. The second reliable finding concerns the number of metrics: the first collection often contains more than a hundred, while typically twenty to thirty are genuinely decision relevant. Reducing to that set is the step that makes reports readable.

    Häufige Fragen

    Do we need a data warehouse or is an analysis tool enough?

    If your metrics come from one system, its own reporting or a tool that accesses it directly is enough. A dedicated analysis database pays off when you combine data from several systems, when you need historical states that the source overwrites, or when analyses place a noticeable load on the source systems.

    How do we stop everyone building their own numbers?

    Through a clear separation: binding metrics with documented definitions, maintained centrally, and an area for individual analysis that is explicitly a working draft. It also helps to establish that only the binding figures are decision relevant in meetings. Self-service cannot be banned anyway, and it is useful.

    Let us talk about Reporting that holds

    In a thirty minute first call we work out where your biggest lever sits and whether we are the right people for it.

    Further reading

    Back to the field Data and Information