Chatter Performance Scorecard: The Grid, Ready to Copy

Every agency eventually builds a leaderboard, and every leaderboard eventually ranks the person who was handed the best inbox. The scorecard below exists to stop that. It scores the work, not the fans someone inherited.

Copy the grid, keep the scale, and resist the urge to add up the column.

What is a chatter performance scorecard for, and who uses it?

It turns “how is she doing” into seven answerable questions, asked the same way every month. That is the whole purpose: comparability across people and across periods.

• The owner uses it to allocate shifts. The peak evening should go to the highest timing and closer scores, not the highest revenue line.
• The team lead uses it as a coaching agenda. One low criterion is one conversation to have, with examples attached.
• The chatter uses it to know what good looks like before the review, not during it.

It is not a ranking, and it is not a pay formula. The moment it becomes either, people start writing for the grid instead of for the fan.

What does the complete scorecard look like?

Seven criteria, one scale, no total. Score each from 1 to 5, where 1 is “costs us threads”, 3 is “safe unsupervised”, and 5 is “I would hand them a difficult account”.

#	Criterion	What you are reading	Score 1	Score 5	Source
1	Conversion at comparable volume	Conversion rate inside their volume band	Bottom of the band	Top of the band	Dashboard + band split
2	Pitch timing	Where in the thread the pitch lands	Pitches in the first few messages	Waits, and the wait is deliberate	Read conversations
3	Closer quality	The last line of a sales message	Ellipsis, or a closed question at the close	Open, specific, no trailing dots	Read conversations
4	Voice discipline	Drift from the creator’s reference messages	Own vocabulary surfacing	Indistinguishable from the reference	Read conversations
5	Floor-price discipline	Behaviour under a haggle	Discounts, or invents a bundle	Refuses and keeps the thread alive	Read conversations
6	First-response time on live threads	Median wait on threads inside their block	Threads outlive the block	Answered inside the block, every time	Dashboard
7	Handover hygiene	The note left at the end of the block	Missing, or unusable	Next shift never re-pitches a purchase	Handover notes

Recorded but not scored, in a separate row of the same sheet:

Period: ____________          Chatter: ____________
Threads handled: ______       Volume band (1 / 2 / 3): ______
Blocks worked: ______         Creators covered: ____________
Gross sales attributed: ______   ← recorded, never scored
Conversations sampled for this review: ______

And the review note, written before the meeting, not during it:

Scorecard | [chatter] | [period] | reviewer [____]
Scores: 1)__ 2)__ 3)__ 4)__ 5)__ 6)__ 7)__
Strongest criterion, with one example thread: ____________________________
Weakest criterion, with one example thread: ____________________________
One thing to change before the next period: ____________________________
One thing to keep doing: ____________________________
Compared with last period, what moved: ____________________________

How do you score each criterion?

Four of the seven need conversations read end to end. Pull a sample across different blocks, including one quiet afternoon and one peak evening, so you are not scoring the easy hours.

• Criterion 1. Split the roster into thirds by threads handled, then read conversion only inside a third. Never across bands.
• Criterion 2. Count the messages before the pitch. In our corpus, pitching before the sixth message drops conversion by roughly a third, and the optimum sits after about ten exchanges (the reasoning is in when to pitch a sale).
• Criterion 3. Look only at the final line. Two habits cost you conversion: the ellipsis is the worst-performing closer we have measured, and a closed question at the moment of closing costs several points of conversion. See how to end a sales message.
• Criterion 4. Read five consecutive replies against the creator’s reference set. Drift shows up in punctuation and jokes before it shows up in vocabulary.
• Criterion 5. Find one haggle. There is always one.
• Criterion 6. Median, not average. One catastrophic thread should not decide the score.
• Criterion 7. Read the note the next shift received, not the one that was written.

One more habit belongs in the coaching conversation even though it does not carry its own row: a personal callback placed at the exact moment of the pitch costs points, even though it feels like it should help. It is the most counter-intuitive result in our corpus and people do it instinctively: the personal callback mistake.

What do you record about revenue instead of scoring it?

Three lines, all of them context and none of them scored: gross sales attributed, threads handled, and the volume band those threads put the person in. Revenue belongs in the sheet (a review that cannot see the size of the period is not a review), but it belongs above the grid rather than inside it, because it measures the inbox and the grid is trying to measure the person. The longer argument for that is in creator agency metrics that matter; the scorecard’s job is narrower, which is to keep the number in the room without letting it rank anyone.

What you compare	What it actually measures	Ranks correctly?
Gross sales	Who inherited the best fans	No
Sales per hour worked	Inbox quality, again, per hour	No
Conversion, whole roster	Volume differences between people	No
Conversion inside a volume band	How well comparable work was done	Yes
Messages before the pitch	A habit our data ties to conversion	Yes, as a leading signal

The correction is one step: build the bands before you build the ranking. Comparing someone working cold threads with someone holding long-standing whales produces a number that looks precise and means nothing.

What are the three classic mistakes?

Averaging the seven into one score. A chatter scoring 5 on closer quality and 1 on floor discipline averages to something unremarkable, and the average hides the only fact worth acting on. Read the profile, never the mean.

Changing a criterion mid-period. Comparability is the entire value of the grid. If a criterion is wrong, note it, finish the period, and change it at the quarterly revision.

Scoring from the dashboard alone. Criteria 2, 3, 4 and 5 are invisible in aggregate numbers. An agency that scores only what its tooling counts ends up rewarding volume and speed, which are the two things that push people to pitch early.

When do you revise the scorecard?

Revise the grid quarterly, and run the review monthly. Weekly reviews measure noise; quarterly reviews arrive too late to change anything.

Three triggers justify an off-cycle revision:

1. Two reviewers cannot agree. Rewrite the criterion until two people reading the same five conversations land within one point of each other.
2. A criterion stops separating anyone. If everyone scores 4 or 5 on it, it has become a house standard. Move it to the onboarding checklist and free the row.
3. The work changes. A new platform, a new creator with a different audience, or a change in what the team is asked to sell all move what “good” means on rows 2 and 5.

Keep the old grids. The first thing anyone asks in a review is what moved since last time, and that answer only exists if the criteria stayed still.