Most workplace communication training is a two-hour workshop about listening, empathy and body language. People nod, the slides get filed, and nothing measurable changes. Six months later the same handoff is still getting dropped between shifts.
The training is not useless because the topics are wrong. It fails because “listen better” is not a skill anyone can practice, observe or correct. What does work is narrower and less inspiring: teach people a fixed sequence of words for a specific situation, then check whether they use it.
The short answer
Generic communication workshops have weak evidence behind them. Structured communication protocols, meaning a defined script for a defined moment such as a handoff, an escalation or a shift briefing, have strong evidence, including a measured drop in surgical mortality across 108 hospitals. Convert each soft skill into one protocol, one situation and one observable behavior.
What the evidence actually supports
Start with the size of the problem, because it is larger than most training budgets assume. In research conducted by Grammarly and The Harris Poll between October 1 and 28, 2021, surveying 1,001 US knowledge workers and 251 business leaders, leaders estimated their teams lose 7.47 hours a week to poor communication out of the 19.93 hours a week spent on written communication alone. Grammarly put the cost at roughly $12,506 per employee per year and $1.2 trillion across US businesses.
That is a survey of estimates, not an audited number, and it should be read that way. What it establishes is scale, not precision.
Now the other side. The strongest evidence for communication training comes from medicine, where the training was not a workshop about listening. Neily, Mills, Young-Xu and colleagues studied the Veterans Health Administration’s medical team training program across 108 facilities and published the results in JAMA in 2010. Facilities that ran the program saw annual surgical mortality fall by about 18 percent, against roughly 7 percent at facilities that had not yet started. There was also a dose response: the more structured briefing and debriefing a facility did, the further mortality fell.
The program taught named procedures. Preoperative briefings with a checklist. Postoperative debriefings. Explicit permission for any team member to stop the line. Readback of critical information. None of that is “communicate more openly.” All of it can be watched, scored and corrected.
That is the whole distinction this article is built on. If a training item cannot be observed, it cannot be trained.
The five skills, converted into protocols
1. Active listening becomes closed-loop readback
The workshop version asks people to paraphrase and maintain eye contact. The trainable version is a loop with three moves: the sender states the instruction and names the receiver, the receiver repeats the instruction back in their own words, the sender confirms or corrects. The loop is not closed until the third step happens.
Practice it on your own material, not on role-play scenarios about lost luggage. Take five real instructions from last month, three of which were misunderstood, and run them through the loop in a live session.
What you observe: whether the receiver read back, and whether the sender waited for it.
2. Clarity becomes one fixed escalation format
Ambiguity is worst when someone is worried and senior to no one. Give them a format so the words are already chosen. A workable four-part version: here is what I am seeing, here is what I think it means, here is what I am asking for, here is my level of concern.
The format matters less than the fact that there is exactly one. Two competing formats produce neither.
What you observe: escalations that arrive with a stated ask, instead of a description and a hope.
3. Empathy becomes a defined response sequence
You cannot train someone to feel differently in a 90-minute session, and it is slightly dishonest to claim otherwise. You can train the sequence a person follows when a colleague or customer is upset: acknowledge the specific thing, restate it in your own words, propose the next step, confirm agreement.
This is the sequence customer-facing teams already use under other names. Managers rarely get it, which is why difficult conversations in most organizations depend entirely on the individual manager’s instincts. If you are building a manager program, this is where the hours go, not into personality inventories. Our note on balancing soft and hard skills covers how to schedule the two together without the soft half getting cut first.
What you observe: whether the restatement happens before the solution.
4. Teamwork becomes a briefing and a debriefing
The VHA result came mostly from this. A briefing is a short, structured start-of-shift or start-of-procedure conversation covering who is present, what is planned, what is different today and what could go wrong. A debriefing is the same shape afterward: what happened, what surprised us, what we change.
Both need to be short and both need to be mandatory. A briefing that runs when the shift lead feels like it will be run on calm days and skipped on the days it would have mattered. Teams running this pattern across sites usually need the schedule and the attendance record in one place, which is what frontline workforce training programs are built around.
What you observe: briefing completion rate, and whether debriefings produce a change anyone can name.
5. Written communication becomes a house standard
The Grammarly figures are about writing, and writing is the easiest of the five to standardize because the artifact sits there and can be reviewed. Decide the small things centrally: subject-line convention, where the decision goes in the message (first), what belongs in chat versus email versus a document, and the maximum length before something becomes a document.
Then enforce it in the templates rather than in training. A template people already use beats a rule people were once told.
What you observe: how far into a message the reader has to go to find the ask.
What each protocol looks like on paper
| Soft skill | The protocol | When it fires | What you measure |
|---|---|---|---|
| Active listening | Closed-loop readback | Any instruction with a consequence | Readback rate on audited handoffs |
| Clarity | Four-part escalation format | Raising a concern upward | Share of escalations containing an explicit ask |
| Empathy | Acknowledge, restate, propose, confirm | Complaint or conflict | Restatement before solution, in call or session review |
| Teamwork | Structured briefing and debriefing | Start and end of every shift or procedure | Completion rate, plus changes generated per debrief |
| Written communication | House message standard and templates | Every message longer than three lines | Position of the ask in the message |
What to drop: non-verbal communication training
Almost every communication course contains a segment claiming that 93 percent of communication is non-verbal, usually as 55 percent body language and 38 percent tone. That figure comes from two small Albert Mehrabian experiments published in 1967, in which participants judged the feeling behind single spoken words where the tone deliberately contradicted the word. Mehrabian himself has said the ratio was never meant to apply to ordinary communication. It does not survive contact with a sentence, let alone a project update.
Cut the segment. It spends real hours teaching a number that is wrong, and its presence tells anyone in the room who knows the research that the rest of the material was not checked either.
Where this approach does not work
Protocols are a good answer to a narrow question, and it is worth being clear about the edges.
- They do not fix a culture where raising a concern is punished. Readback trains the words, not the willingness. If people are afraid, they will complete the loop and stay quiet about the thing that mattered. That is a management problem and training will not touch it.
- They do not transfer well to genuinely novel or creative work. A briefing format helps a surgical team and a shift handover. It does not help two engineers arguing about an architecture.
- They decay. The VHA dose response cuts both ways: facilities that stopped reinforcing lost ground. Assume a refresh cycle, not a launch.
- The healthcare evidence is healthcare evidence. High-consequence, high-repetition, tightly supervised work. Transfer to an office setting is reasonable but it is an inference, not a finding, and nobody should present it as proven.
Running it without adding administrative work
The reason protocol training gets abandoned is rarely the content. It is that briefings, refreshers and observation records are three different tracking problems, and most teams run them on spreadsheets until the spreadsheet is a month stale.
This is the part a training management system for enterprise L&D teams is actually for: scheduling recurring sessions against shift patterns, recording who attended which cohort, and holding the observation checklist next to the record instead of in someone’s inbox. Where the training is practiced rather than watched, and most of this is, you need instructor-led session management that can handle multiple sites and instructors on one calendar.
In regulated settings the record is the deliverable, not a nicety. A healthcare training management system has to show which clinician completed which protocol on which date, on demand. One multi-state health system standardized its nurse onboarding pathway across every hospital for this reason, and cut Joint Commission audit preparation from about a week to under a day. The detail is in our multi-site nurse onboarding case study.
If you are only tracking whether the training happened, and not whether the behavior appears afterward, you are measuring attendance. Training management reporting is worth setting up around the observable measure in the table above from the start, because retrofitting it later means asking managers to remember last quarter.
A reasonable first 90 days
Pick one protocol, not five. Choose the situation where a communication failure already cost you something you can name, because that gives you a before number.
Weeks 1 and 2: write the protocol on one page. Weeks 3 and 4: run it with one team, in a live session with real material. Weeks 5 to 8: managers observe and record, weekly, using the same checklist. Weeks 9 to 12: compare the observable measure against your starting point and decide whether to extend it or fix it.
If nothing moved in 90 days on one team, the protocol was wrong or nobody was watching. Both are findable. Neither is findable after a workshop.
Frequently asked questions
Is communication training worth the budget at all?
How long should a communication protocol be?
What do we measure if we cannot audit conversations?
Does the 93 percent non-verbal statistic have any valid use?
Should managers or external trainers deliver this?
How often does protocol training need refreshing?
Where to start
Take your five most recent communication failures, write down which situation each one happened in, and see how many share a situation. That is the protocol to build first.
If the constraint is tracking rather than content, book a walkthrough and we will show you how the scheduling, attendance and observation records sit together for a recurring protocol program.
