Balanced scorecard

What it means

The balanced scorecard is a frame that turns strategy into measures across four perspectives: finance, customer, internal processes, and learning and growth. It balances the money of today against what will earn it next year.

It is not a list of KPIs and not another dashboard. The substance is a chain of cause and effect: trained people improve a process, a better process keeps the customer, and revenue follows.

A scorecard is usually drawn as a strategy map: one page with the objectives from all four perspectives linked by arrows that show what drives what. A measure that cannot be tied to any objective on the map does not belong in the scorecard, however easy it is to measure. The map therefore also works as a filter against pointless numbers.

How it works in practice

In a small company two or three measures per perspective are enough, each with an owner, a target and a data source. Finance comes from the books, the customer from the CRM and invoicing, processes from lead times and complaints.

A scorecard does not replace OKR: the scorecard is a standing frame for measurement, OKR a quarterly push for change. The starting point is a strategic analysis, a PEST analysis or a look at the competition.

For the financial perspective we prepare the measures from the monthly close: margin, days sales outstanding, fixed costs as a share of revenue, each defined the same way every month. The company fills in the other three perspectives itself. We make sure the financial figures agree with the books and nobody adjusts them by hand in a spreadsheet before the meeting.

What to watch out for

A scorecard wired to a bonus through a single number is dangerous. If the bonus hangs on revenue, quarter-end brings invoices for orders never delivered: the tax shown on the invoice is payable anyway and the cash comes late or never.

The second mistake is too many measures. Nobody reads a scorecard with thirty numbers, and the meeting turns into reading a table aloud. If a measure has not triggered a single decision or question in six months, drop it or move it to routine reporting, where it does not take up management attention.