What a scorecard is

The questions a conversation is judged against. It is your own definition of good service, written down.

The questions a conversation is judged against. It is your own definition of good service, written down.

What it contains

• Rating categories, the individual questions.
• Group sections, optional, to keep related categories together.
• Weights, where some matter more than others.
• A rating scale, chosen once for the scorecard.
• Critical categories, where a failure fails the whole review.
• Root causes, the reasons a reviewer picks to explain a rating, kept as separate lists for positive and negative.

Root causes are the part teams skip and then miss. A score tells you a review went badly; a root cause tells you why, and it is the only field that turns a quarter of reviews into a list of things to fix.

The four scales

The 4-point scale is the one to consider when reviews keep landing in the middle, because it removes the middle.

It is a definition, not a form

Writing a scorecard forces a team to agree what good service is here. Most teams discover during that exercise that they did not agree, and that discovery is worth more than the scores.

Short beats complete

Four to six categories. A long scorecard takes fifteen minutes per review, so fewer reviews happen, and reviewers start scoring the middle option to get through it.

Each category needs a clear answer

"Did the agent confirm the customer's issue before answering?" can be judged from the transcript. "Was the agent helpful?" cannot, and two reviewers will score it differently every time.

If you cannot point at the evidence, the category is not usable.

Critical categories, sparingly

One or two, for things that genuinely invalidate the interaction: giving wrong information, breaching a policy. Making everything critical makes nothing critical.

Know what it does before you tick it. A critical category rated below 50% fails every category in the review and the score becomes 0%. Where the category sits in a group you can limit the damage to that group instead, which is usually what people meant.

Weights are a message

Whatever you weight most heavily is what the team will optimise. Make sure it is what you care about, rather than what was easiest to define.

Expect to revise it once

After a quarter of real reviews, one or two categories will be ambiguous. Revising then is normal; revising every month means nobody can compare anything.

See also

• What makes a scorecard that people accept
• Creating a scorecard
• What rating categories are