Measurement and performance: two distinct processes?

Most organisations run one performance management process and ask it to do two unrelated jobs: produce data for promotion and pay decisions, and actually make people better at their work. MIT Sloan Management Review has argued these should be separated, and the argument holds up.

Why one process cannot do both

Measuring and improving pull in opposite directions. Measurement wants consistency, comparability and distance. Improvement wants candour, specificity and a manager who is on your side. Put them in the same meeting and the employee has to decide whether to admit a weakness to someone who is about to score them on it. Most people decide not to, which is how the annual review became a negotiation.

The result is familiar: a complex framework, a once-a-year ritual, data that is not accurate enough to make decisions with, and managers who disengage from the whole thing.

Measuring performance

If the purpose is informing talent decisions, the design goal is simplicity and scale. Sloan's version uses five dynamics: quality of work, timeliness, proficiency, how well the person partners with the team, and how well they partner with their manager.

Collected frequently, quarterly rather than annually, through a short survey the manager completes. Short is the operative word. It takes less time, it carries less bias than a long free-text assessment, and because it happens four times a year you get something close to real-time insight rather than an annual snapshot taken when everyone was tired.

Accelerating performance

If the purpose is improvement, the design goal is frequent, personal manager attention. Weekly check-ins built on open questions: what are your priorities this week, what do you need from me, how is the work going. Not scored, not recorded, not feeding into anything.

That separation is what makes the honesty possible. People will tell you what is not working when the conversation is not also the thing that determines their bonus.

What it looks like when done

A regional credit union cut its performance measurement to a six-item survey completed quarterly by managers, and separately introduced weekly manager check-ins. Sloan reports both employee satisfaction and performance improved.

Where to start

Ask honestly whether your current process is achieving either purpose, let alone both. Then look at running them separately. The easiest first move is introducing the check-ins, because it requires no system change and no committee, just managers having a short conversation every week.

The two-pronged version gives you measurement you can trust for decisions and development that actually develops people, which is what the single process was always claiming to deliver.

Where this is taught

The performance management cycle and where the formal review fits is a module of the Certificate in Performance Management. See also ten questions worth asking in a review and how to build a performance management process.

The original article is What We're Still Getting Wrong About Performance Management in MIT Sloan Management Review.