Write Coaching Feedback a Rep Can Argue With — And Then Fix

"You sound like you don't care" is not coaching

Here's feedback a rep got last month, verbatim from a manager's notes:

"Needs to build more rapport. Comes across as transactional. Not enough enthusiasm."

The rep read it, nodded, and changed nothing. Not because he's stubborn. Because there's nothing in there he can do on Monday morning.

"Transactional" is a label. It describes who the manager thinks the rep is, not what happened on the call. And when you hand someone a label, you've handed them an identity problem. Their options are to accept that they're a transactional person or to argue that they're not. Neither one improves the next call.

Evidence-based feedback does something different. It plays back the tape. It says: at 3:42 the customer said her water heater had been leaking for two weeks and you responded with "okay, and what's the model number." That's not a personality judgment. That's a thing that happened, and the rep can hear it, and there's an obvious alternative sentence he could have used instead.

The test: could a stranger verify it?

Before you write a line of coaching feedback, run it through one filter.

If I handed this note and the recording to another manager, could they point to the exact moment I'm describing?

If yes, it's evidence. If no, it's an impression.

Notice that the praise column has the same problem. "Great rapport" is as useless as "transactional." The rep doesn't know what to repeat.

Four things that count as evidence

You don't need a lab coat. You need four categories.

1. Quotes. What the rep actually said, word for word. Not a paraphrase. "You said 'we're pretty much the best in the area'" hits differently than "you oversold."

2. Timestamps. 4:15. 11:30. This is the single cheapest upgrade to your notes. It turns an argument into a replay.

3. Counts. Number of open questions before the pitch. Number of times the customer was interrupted. Number of times the price was defended before the objection was understood. Counts are hard to dispute.

4. Sequence. What came before what. "You quoted the price at 5:00 and asked about their timeline at 12:00" is a diagnosis in itself. The order was backwards.

That's it. Quote, timestamp, count, sequence. Every piece of real feedback you'll ever give is built out of those four.

Rewriting the labels you actually use

Let's do the ones managers say most.

"He doesn't build rapport"

What you probably observed: the rep opened with "Hi, this is Dave, what can I do for you," took the request, and started qualifying. Zero acknowledgment of anything the customer said emotionally.

Write it as:

At 0:20 the customer said "I've been calling around all morning and nobody's picking up." You went straight to "Sure, what's the address." A sentence like "Sounds like you've been at this a while — I can help you right now" costs you four seconds and changes the whole tone.

Now there's a target behavior and a sample line.

"She talks too much"

What you observed: the rep talked for 74% of the call, and the longest customer turn was 11 seconds.

Write it as:

You held the floor for about three quarters of the call. The longest the customer spoke uninterrupted was 11 seconds, at 3:05. Goal for the next five calls: at least one customer answer over 30 seconds. Ask "walk me through what's been happening" and then don't say anything until they stop.

"He's not confident"

Almost always this means hedging language. Go find it.

Counted seven hedges: "kind of," "I think maybe," "if that works for you," "sorry to bother you." The one that cost the most was at 14:20 — "I could maybe get someone out Thursday if that's okay?" Try: "I've got Thursday at 2 or Friday at 10. Which one?"

"She gets defensive on price"

Customer said "that seems high" at 8:40. You started explaining the warranty at 8:42 — two seconds. You never found out what she was comparing it to. Next time, one question first: "High compared to what you were expecting, or high compared to another quote?"

Each rewrite does the same three things: names the moment, names the cost, gives the sentence.

A call review template that forces evidence

The reason managers write labels isn't laziness. It's that a blank text box invites impressions. Structure fixes that. Here's a call review template that makes it hard to be vague.

Call: [rep, date, call type, length]

Outcome: [what actually happened — booked, quoted, dead, callback scheduled with date]

Two things to repeat

• Quote + timestamp:
• Why it worked (what did the customer do next?):

One thing to change

• Quote + timestamp:
• What the customer did after:
• The sentence to use instead:

Rep's own read: [filled in by the rep before the session]

That last line matters more than it looks. Have the rep listen first and write what they'd change. Most of the time they name the same moment you did, and the conversation stops being a verdict and becomes a comparison of notes.

The "why it worked" field is the one people skip. Force it. "It worked because the customer immediately volunteered that they'd had a bad experience with the last company" is a lesson. "Nice job" is not.

Where managers still slip

Mind reading. "You could tell he wasn't interested." You can't tell. What you can say is: "He gave three one-word answers in a row starting at 6:00."

Attribution. "He rushed because he had a bad month." Maybe. Doesn't matter. Coach the behavior on the recording.

Stacking. Six pieces of evidence about six different problems is not six times as helpful. It's paralysis. Pick one behavior. Everything else waits until that one is fixed.

Vague praise as a cushion. "Great call overall, just one thing." If the call was great, say which 20 seconds were great and why. Otherwise the rep hears the sandwich and discards both slices.

Run one week of this

Pick three calls per rep this week. For each, write nothing that doesn't have a quote, a timestamp, a count, or a sequence in it. If a line doesn't have one of those, delete it.

You'll notice two things fast. First, your notes get shorter. Second, your coaching sessions stop being negotiations about what kind of person the rep is and start being reviews of what happened at 8:40.

Reps argue with labels. Nobody argues with a timestamp. Whether you're pulling call monitoring feedback examples from recordings by hand or your scoring runs automatically, the standard is the same — if you can't point to the moment, don't write the sentence.