The short definition
A data-driven app is a piece of software whose primary job is to pull data from other places, make sense of it, and help someone make a decision. It is different from a regular web app, which tends to help someone do a task. One produces a report, the other produces a booking.
Three real examples
Sales pipeline dashboard. Pulls deals from HubSpot, invoices from Xero and payments from Stripe. Shows a single view of lead-to-cash. The sales lead opens it every Monday morning and knows within 30 seconds where to focus the week.
Fleet utilisation report. Reads vehicle movements from a telematics provider, job data from a scheduling system, and fuel card spend from the accounts package. Tells the operations manager which vehicles are earning their keep and which are not.
Stock and supplier tool. Watches warehouse stock levels against sales velocity and supplier lead times. Flags what to reorder, when, and from whom, before the stockout actually happens.
How it differs from a regular web app
- It usually reads data from three or more places, not one
- The user mostly looks at it, rather than typing into it
- The hard work sits in the joins and the rules, not the screens
- It is judged on decisions made, not clicks logged
How it differs from a BI tool
Business intelligence tools (Tableau, Power BI, Metabase) are excellent at generic reporting and bad at bespoke logic. A data-driven app is worth building when your rules are specific to your business: "if this customer is more than 45 days overdue and has a live support ticket, do not send the second reminder email yet". A BI tool cannot reach that. An app can.
When you do not need one
If one spreadsheet covers it and two people check it once a week, leave it alone. Data-driven apps earn their keep when multiple systems are involved, multiple people need the same view, or the rules are too fiddly for a spreadsheet to stay honest.