Reporting features: why exports matter more than charts
The most used reporting feature in most business software is the export button. Why that is fine, and how to build it well.
The short answer
Look at the usage logs of almost any business application and the most used reporting feature is the export button. That is not a failure of the reporting features; it is people choosing the best tool for exploration, which is the spreadsheet or business intelligence tool they already know and which has years of features for slicing, pivoting and charting. The application’s job is to hold the data correctly, show the handful of numbers that drive daily decisions, and hand everything else over cleanly. A good export, and the same data through the API, serves more people better than a report builder ever will.
Charts, dashboards and exports
| Need | Best served by | Why |
|---|---|---|
| A decision made every day at a glance | A dashboard in the application | Right place, right moment, few numbers with comparisons |
| A question asked this week that nobody asked before | An export into a spreadsheet | Flexible, familiar, immediate |
| Regular analysis across systems | A business intelligence tool fed by the API or scheduled exports | Built for joining, modelling and charting |
| A document for someone else | An export in the format they need, or a scheduled report | Portable, printable, shareable |
| A custom chart of anything | Not the application | Every chart built in is one someone will want slightly differently |
Building exports well
- Every field, not the ones on the screen; users will ask for the rest within a week.
- Consistent formats: dates in one standard, numbers without formatting, identifiers as text, one row per record.
- The same filters as the screen, so what you see is what you export.
- A data dictionary: what each column means, in a document linked from the export.
- Scheduled exports to a mailbox or a storage location, for the reports people need every Monday.
- The same data through the API, so a business intelligence tool can connect without a person clicking a button.
- Permissions respected: users export what they may see and nothing more; exports are logged.
Where charts belong
In the application, for the questions asked every day by the people who live in it: today’s bookings, this week’s revenue against target, open items past due. Few, with comparisons, on one screen. Everything exploratory belongs in tools built for exploration, fed by exports and the API. An application that gets that split right has a small, excellent reporting surface and users who never feel trapped by it.
What this means for you
Ask for a few dashboards that answer daily decisions, exports that are complete, consistent, documented, filterable and scheduled, and the same data available through the API. Resist the built-in report builder. Your team will analyse in the tools they know, and the application will have spent its budget on holding data correctly rather than on charts nobody looks at.
Frequently asked questions
Is it not a failure if people export data to analyse it elsewhere?
No. It is people using the best tool for exploration, which is rarely the application itself. A spreadsheet or a business intelligence tool has years of features for slicing, pivoting and charting; rebuilding a fraction of that inside an application is expensive and always incomplete. The application's job is to hold the data correctly and hand it over cleanly.
What makes an export good?
Every field, named consistently, in stable formats for dates, numbers and identifiers; the same filters as the screen; documentation of what each column means; a scheduled version that lands in a mailbox or a storage bucket; and the same data available through the API for tools that connect directly. Most exports fail on the first two: missing fields and inconsistent formats.
Should we build a report builder into the application?
Rarely. Report builders are large features that end up as poor imitations of tools people already have. Build the handful of dashboards that answer daily decisions, make exports and the API excellent, and connect a business intelligence tool for the people who need to explore. It costs less and serves everyone better.