Business intelligence and reporting

Almost every company we work with already has the data. What it does not have is one place where the data agrees with itself - the store reports one revenue figure, the accounts report another, and the difference is a definition nobody wrote down.

This is not a dashboard project. It is deciding what the numbers mean, then building something that produces them the same way every time.

The definitions come first

The hard part is never the chart. It is that an order means something different in the store, the warehouse and the accounts, and until that is settled every report is arguable.

• When does an order count - placed, paid, shipped, or not returned within thirty days?
• Is revenue with or without VAT, shipping, and the discount applied after the fact?
• Which customer record is the customer, when the same person exists three times?
• What does a cancelled and re-placed order do to a number that was already reported yesterday?

Written down once, these stop being arguments. That document tends to be worth more than the reports built on it.

What gets built

• A scheduled pull from the systems that hold the data - store, ERP, CRM, accounting, ad platforms
• A reporting store separate from the live database, so a heavy query cannot slow down the checkout
• The definitions above, implemented once, in one place
• Dashboards for the handful of numbers looked at weekly, and scheduled reports for the ones read monthly
• Alerts for the numbers that only matter when they move

What we will talk you out of

A dashboard with forty tiles is a dashboard nobody reads, and so is a daily email that never changes. We would rather build six numbers somebody acts on than a wall everyone admires once.

Veelgestelde vragen

Which number is being argued about?

The fastest way into this is not a list of reports you want. It is the one figure two systems disagree on today - that argument usually contains the whole project.

Start uw project