What One Badly Handled Conversation Costs a Company

One conversation, four clocks

Take the most common case in any shared-service centre: an annual review in which a good employee gets „meets expectations” instead of the „exceeds” they were counting on. The conversation takes thirty minutes. That is what the calendar shows.

Now the clocks the calendar does not show.

The employee’s clock. If the conversation was handled well — with specifics, with criteria, with a date for the next one — the employee goes back to their desk and works. If it was handled badly — in generalities, without justification, with „keep trying” at the end — the employee goes back to their desk and for the next several days does three things instead of one: works, replays, and searches. Searches for an explanation among colleagues, searches for comparisons, and increasingly, searches job listings. In my experience, the drop in real output after a badly received review lasts two to three weeks, sometimes longer — depending on the situation and the person. And it shows up in no report, because the employee is still at work, still answering emails, still closing tasks. Just more slowly.

The manager’s clock. The same conversation, handled badly, comes back to the manager in three instalments. First: the employee asks for a second meeting „to clear things up”. Second: the manager explains to their own boss why morale in the team has dropped. Third: the manager prepares the documentation they didn’t prepare before the conversation, because back then it seemed unnecessary. From the manager’s chair, one badly handled annual review means anywhere from a dozen to several dozen extra hours over the following month.

That spread — a dozen to several dozen — is not accidental, and it is the most important thing I can say about this bill. The size of the cost does not depend on what happened. It depends on how wronged the employee feels. The stronger that feeling, the longer they dwell on it, the more often they come back to it, the more people they draw in. The same decision — the same rating, the same refused raise — communicated with criteria costs a dozen hours; communicated in generalities, several dozen. The difference is in the conversation, not the decision.

HR’s clock. When the employee didn’t get an answer from the manager, they go to HR for one. HR doesn’t know the context, so they gather it: a conversation with the employee, a conversation with the manager, sometimes mediation. Each such escalation is three to five hours of an HR business partner’s time. In a shared-service centre where one HRBP covers several hundred employees, those are hours that exist for nothing else — and review escalations never come one at a time. They arrive in the same window of the year as all the others.

The team’s clock. This one is the most expensive and the least visible. An employee living through a bad conversation doesn’t live through it alone. They talk. The team listens and draws conclusions about how the company treats people — not from the values on the wall, but from this story. If the team has ten people and each spends half an hour on corridor conversations about it, that is five hours of team time on one conversation that lasted thirty minutes.

Let’s add up what can be counted in hours — manager, HR, team. Conservatively, at the low end of the ranges, it comes to about twenty hours. At the high end — over fifty. That is three to seven working days of the organisation’s time for one thirty-minute conversation. On top of that come two to three weeks of the employee’s own reduced output, which I leave out because it cannot honestly be converted into hours.

This is an estimate, not a measurement. But it is an estimate from twenty years of watching this up close, from every side of the table — and I am fairly convinced it is an underestimate.

The improvement plan: when the process costs more than the decision

The second case where the bill is most brutal is the performance improvement plan — the PIP.

In my experience, companies run performance improvement plans in one of two modes. The first: as a real plan with criteria, support and a deadline, after which the employee either returns to form or both sides know it is over. The second: as documentation of a decision that has already been made. In the second mode, nobody says so out loud — and that silence has a price.

A plan usually runs six to eight weeks, longer in justified cases. Throughout, the manager holds weekly meetings whose outcome is known in advance, writes notes, consults HR. The employee knows they are on a plan, so their output — the thing the plan was meant to improve — drops further. The team sees that a colleague is „on a plan”, and some people quietly start checking how far they are from the same.

Sometimes the plan ends after a week or two, because the employee resigns. On the calendar it looks like a saving. It isn’t: the company loses the same person, only faster, without any documentation of value, and without an answer to whether they could have been kept.

And then, one way or another, what was known from the start happens. Recruitment of a successor begins only now — and there is no single number here. Some searches are quick; some drag on for three months, and you never know in advance which one you’ll get. Then onboarding the new person in an accounting or settlement process to full independence: three to six months — and in some roles and teams, ten to twelve.

Count the whole cycle, from the first conversation about the plan to the moment the successor works independently: six months to a year. For the most complex processes — over eighteen months. Throughout that time, the position is not running at full capacity, and the manager is handling three things at once: the plan, the recruitment and the onboarding. They received extra time for none of them.

It is worth putting those two numbers side by side. The conversation that starts the whole cycle takes an hour. The cycle takes a year.

I am not claiming improvement plans are bad. I am claiming that a plan run as theatre costs the company more than an honest conversation about parting ways — and more than an honest conversation about what specifically has to change for there to be no parting. Both of those conversations are harder than theatre. And both are shorter.

Why managers can’t hold these conversations

Because nobody taught them. That is not an accusation; it is a description.

In a shared-service centre, promotion to a first managerial role is earned through results as a specialist. Then comes training on the review system — how to fill in the form, what the deadlines are, where to click. There is no training on what to say to a person who has just heard they won’t get a raise for the second year running. No training on how to open an improvement plan so that it is a plan and not a sentence. No training on how to receive a resignation from the best person on the team without burning the relationship in the first minute.

The manager learns this alone, on their own people. That is how I learned. The cost of that education was borne by my employees, my team and my superiors — for years, before I learned to do it properly. Today I know that most of those mistakes could have been avoided in a single day’s workshop. Back then, there was no such workshop.

What changes when the conversation is handled well

I want to be precise, because „a well-handled conversation” sounds like a line from a brochure. This is not about being soft. It is about four things you can check.

Criteria instead of judgements. The employee leaves the conversation knowing what specifically has to happen for the next one to end differently. Not „be more proactive”. Rather: what result, by when, measured how.

Documentation before, not after. A note after the conversation, sent the same day — to the employee, not to a drawer. When it exists, there is no second meeting „to clear things up”, no escalation to HR, and no reconstructing the conversation three months later.

The date of the next conversation set at this one. This closes the loop of uncertainty that starts the employee’s clock and the team’s clock.

One version of the truth. The employee, the manager and HR hear the same thing. When that’s the case, HR doesn’t have to gather context — because the context is in the note.

All four have one effect in common: they remove the employee’s sense of being wronged, and that sense is the multiplier of the entire bill. None of them takes more time than a conversation handled badly.

The bill for the company

I won’t give you a figure, because any figure would be invented. Instead, here is a way of calculating it that you can apply to your own data.

Take the number of annual reviews in your organisation. Assume, conservatively, that one in five is handled in a way that starts all four clocks. Assign each twenty to fifty hours of manager, HR and team time. Multiply. With two hundred reviews a year and one in five handled badly, that is eight hundred to two thousand hours — half to one full-time position, in January and February alone.

Then take the number of improvement plans per year and check how many ended with the person leaving. Assign each of those six months to a year of a position working below full capacity. Multiply. With ten plans a year, half of which end in a parting, that is two to four and a half years of one position’s work — spread across the organisation so that it is never visible in one place.

That sum is a cost that appears in no budget. And it raises one question: how much of that time would stay in the company if managers knew how to hold these conversations. In my experience — most of it. And that is exactly the one-day workshop I didn’t have twenty years ago.