Data doesn't create value. Decisions do.

The phrase “data is the new oil”, attributed to Clive Humby, is often cited without its most useful qualification: like oil, data takes work before it produces value. Collecting it isn't enough. You need to define its purpose, check its quality and make it available to the people deciding.

Many companies kept only the first half. They accumulated data, bought BI licenses and built teams that publish dashboards. The script repeats: six months later there are forty dashboards, and leadership still asks for the report in Excel.

The dashboard is the last mile

The BI tool may be unsuitable or badly configured, and that deserves checking. But even a good platform cannot decide for the business what each metric means, reconcile incompatible identifiers between ERP and CRM on its own, fix errors at the source or guarantee that a chart will change a decision.

A project can also fail before the first visualization: the ERP and CRM don't share a customer code, margin is calculated differently by each team or someone opens a spreadsheet to “fix” a number they dislike. When that isn't resolved, each new dashboard can become another argument, and people return to what they know.

If the road doesn't exist, the last mile leads nowhere.

The question almost nobody asks

Before building a report, a dashboard or an agent, ask which decision, operational task or obligation it will support. If there is no clear use, it probably isn't a priority. Using data well means knowing which decisions are made each week, which tasks depend on information and which evidence must be retained.

Designing backwards

The method is simple and counter-intuitive. Start with the decision, for example which customers to call today to collect. Define what you need to know to make it: due dates, payment history, amount at risk. Identify which sources each piece comes from and whether they're connected today. Only then decide the form: a prioritized list every morning, an alert or an answer when someone asks.

When design starts from a concrete use, the result changes nature. Instead of a board with twenty charts, a person might get a prioritized list, an alert or the evidence needed to meet an obligation. Data stops being something you only look at and starts supporting the work. Success is measured by better-grounded decisions, completed tasks or fulfilled obligations, not by the number of dashboards published.

The scenes in these episodes are composites of common situations, not customer cases. Any figures in them are illustrative.