Five Communication Behaviors You Can Actually Score

Edmundo Barrientos · 2026-08-12 · 3 min read

TL;DR: Behavior only exists inside a situation, so measuring it means creating one. Five communication behaviors worth scoring, the acts to look for, and how to keep the result data rather than opinion.

Why a simulation is the instrument

You can ask someone what they know. You cannot ask a questionnaire what they do, because behavior only appears when a situation calls for it. There are two ways to get one.

Observe real work. The most authentic evidence available, and the least comparable: only some roles leave a reviewable artifact, and no two customer calls carry the same difficulty, so a low score may describe a hard conversation rather than a weak performer.

Create the situation. A simulation, a role-play with a real counterpart and a real decision, holds the difficulty constant. Everyone meets the same customer, on demand, with no live account absorbing the cost of a poor attempt. And because the same scenario can be run again next quarter, the score becomes a trend rather than a snapshot.

Both belong in the design: simulations for the comparable, repeatable reading, real work to confirm the behavior survived a live customer. Where the scenarios come from decides whether the score means anything, because a generic scenario measures only generic behavior. The escalations, objections, and standards worth scoring are the ones already sitting in your own recordings and quality guidelines.

Five behaviors worth scoring

These decide outcomes in leadership, compliance, onboarding, and customer support, not just the sales team.

Behavior Look for Common failure
Escalation (support): a customer whose order failed twice, asking for a manager Names the problem back before proposing; one clarifying question before the fix; closes with an owner and a time Reciting policy in the first 20 seconds
Corrective feedback (leadership): a team member who missed three deadlines Opens on observable behavior, not character; states the impact; agrees on one change and a review date Burying the message between compliments
Consultative discovery (sales enablement): a prospect who opens with "just send me pricing" Open questions before presenting; summarizes back and checks it; surfaces the cost of today's situation Feature-dumping at the first buying signal
An unpopular decision (change communication): retiring the tool the team prefers Stated plainly inside 30 seconds; separates settled from open; owns it Burying it after five minutes of context
Supervising an AI recommendation: why an output was accepted or overridden Names the check they ran; states the limitation in the output; gives their own reasoning Full deference ("the system said so")

The last one is the newest and the least measured anywhere. McKinsey's HR Monitor 2026 reports demand shifting away from routine task-based capabilities toward skills focused on analyzing and interpreting AI outputs. That is their finding; the corollary is that this is now a communication behavior in most professional roles, and no completion report can see it.

Score it like measurement, not opinion

  • Define the behaviors before the practice. Criteria written from what you saw are a description, not an assessment.
  • Score the transcript, not the impression, citing the line that evidences each act.
  • Hold the scenario constant across the cohort. Scores compare only when situations do.
  • Track change per behavior per person. "Communication: 3.4 out of 5" cannot be acted on; "acknowledges before solving: 4 of 5 attempts, up from 1" can.

Three limits worth stating out loud

  • It measures can, not always will. Real work validation stays in the design.
  • Fluent is not effective. Score the outcome the behavior should achieve, never the polish.
  • If scores drive pay, people optimize for the rubric. Keep development measurement separate from evaluation.

References

  • McKinsey & Company. HR Monitor 2026: A turning point for the people function, 2026.

Which of these would you score first, and what would you accept as evidence that someone actually did it?


About the publisher: this article is published by Grasply (grasply.ai), a capability builder. Grasply turns what a company already knows, its docs, calls and recordings, into skills people use at work, for employees, partners and customers. Practice turns knowing into doing, and Grasply shows proof that the work changed.