31 October 2026 · 12 min read

Month four: why culture fixes fade

A culture fix is easiest to believe in the week it is announced. It gets tested later, on an ordinary Tuesday, when nobody is looking.

"Month four" is a figure of speech, and no survey names a month. It stands for the moment the review has been filed, the all-hands has faded, and the first test arrives without an audience.

The Tuesday

Callum runs operations at a distribution firm of about a hundred and twenty people. Four months ago an outside review told the board what some of them already suspected: sarcasm at two depots that new starters picked up within a week, and managers who talked over people in meetings. The board agreed a list of actions. Callum stood up at an all-hands, said the new standards out loud, and put them on a slide. People clapped, a little. It felt like a turning point.

This Tuesday, in the planning meeting, Gareth, a team leader who was at the all-hands, answers a quiet colleague's suggestion with "Right, and which planet is that on?" Two people laugh. Callum is looking at his laptop. He heard it. He says nothing, because it is one remark, the meeting is running late, and calling it out would make it bigger than it was.

On the way back to his desk he passes the meeting room. The slide with the standards is still the screensaver, and nobody has looked at it for weeks.

Callum has not failed. He has reached month four.

Callum is invented. The survey below is not.

How often do culture fixes last?

Just 6% of organisations said behaviour improved permanently following intervention (WorkNest, via HRreview, 21 August 2026). It was a survey of 1,017 HR professionals and business leaders, and 45% experienced an initial improvement before the problematic behaviour returned.

Read that for what it is. The page does not say where the respondents work or what the percentages are shares of, so this is not a measured relapse rate. It is what a thousand people responsible for the problem said happened. The shape is still the point: a fix that works at first and does not stay.

WorkNest's own HR consultant, Sita Anand, says it plainly in the same piece. One difficult conversation is not enough to assume the problem has been dealt with, she says, because without clear expectations and consistent follow-up, old patterns can quickly return.

What comes back is worth a look too, because it is not what a board pictures. In the WorkNest survey of 1,017 HR professionals and business leaders, persistent negativity or disruptive behaviour was the most frequently identified form of toxic behaviour (59% of respondents); 35% cited employees undermining managers or colleagues, 30% passive-aggressive behaviour and 13% bullying or intimidation (WorkNest, via HRreview, 21 August 2026). Those shares add up to more than 100, so people named more than one form. Read in that order, the commonest thing on the list is a tone. That is Gareth's remark, and that is why the fix fades so quietly: a tone never trips a process. It needs somebody in the room, on the day, to answer it.

Why do culture fixes fade?

The nearest hard evidence comes from a neighbouring field, workplace resilience training. Across 42 samples in 37 studies it had a small effect, and the effect faded: 0.26 soon after the training against 0.07 at a later follow-up, by that field's usual measure of effect size (Vanhove, Herian, Perez, Harms and Lester, Journal of Occupational and Organizational Psychology, 2016). There is an exception worth knowing. For people at higher risk of stress the effect grew instead, so the fade belongs to programmes given to everybody.

Learning research suggests why repetition matters. In a review of 317 experiments on how people remember, the best gap between two study sessions depended on how long the thing had to be remembered: the longer the wait, the longer the best gap (Cepeda, Pashler, Vul, Wixted and Rohrer, Psychological Bulletin, 2006).

Be careful with both. The first is about resilience, and the second is about recall tasks, not about teams. The step from either to Callum's Tuesday is ours, and it is a reading, not a finding: a standard heard once, on a good day, is one dose, and what keeps it alive is hearing it again, later, when it has been tested.

There is a second reading, and I think it is the more uncomfortable one. An all-hands tells a room what the standard is. The first slip that goes unanswered tells the room what the standard is worth. Everyone present hears both, and they are entitled to believe the second, because it is the one that cost something. It is what a new starter learns in week three: the wall says one thing and the Thursday catch-up says another.

What one unanswered remark teaches a room

The quiet colleague is called Ines. She read the review when it came round, and she took the slide at its word. She had decided to start speaking in planning meetings, and she thought about that suggestion on the drive in.

Next Tuesday she has nothing to add. Nobody will record that. It will not appear in an action list or a survey, and if Callum noticed at all he would call it a quiet meeting. The review found sarcasm that new starters picked up within a week. This is how a new starter picks it up.

The room pays as well. In Will Felps's experiment, groups with one planted "bad apple" performed 30 to 40% worse than groups without one (This American Life, episode 370, "Ruining It for the Rest of Us", 2008, transcript). That is the programme's account of one experiment, and it is no measure of Callum's firm. One remark does not make Gareth a bad apple either. The point for Callum is how far a group's work can move on one person's behaviour, and how little it takes to leave that behaviour alone.

How you can tell a fix is fading

None of these will reach the board. All of them can be seen from inside a meeting.

• The standard stops being said aloud. In month one it was in every meeting. Now it is a screensaver.
• The senior person looks at the laptop. Callum heard it, and the room saw him not hear it.
• "It was only one remark" becomes a reason. Each slip is too small to raise, and the sum of them is the old culture.
• The action list starts to read "ongoing". Nobody can say when anybody will next look.
• New starters learn the old way first. They copy what goes unchallenged, not what was on the slide.
• Quiet people get quieter. Ines has nothing to add, and the meeting finishes on time.

Why a second all-hands will not hold it

Two answers come first, and both are reasonable. The first is another event: an all-hands, an offsite, a refreshed values slide. The second is a document: a tighter policy, a longer list of behaviours, a form.

Both are a single dose. The event works on the people in the room that day, and attention is the thing that fades. The document works only if somebody reaches for it, and nobody reaches for a policy in the middle of a sentence in a planning meeting. Month four is a Tuesday on which the room is different: busier, tired, and without the board in it.

How to make culture change stick before month four

None of this needs a product.

Put the check in the diary now, with a name against it

Pick a date three months after the all-hands. Name the person who will ask three people who sit in meetings one question: "What have you seen in the last month that the review would have flagged?" Write the answers down. A check with no date is the first thing month four removes.

Make the standard small enough to fail in a meeting

"Respect" cannot be seen to fail on a Tuesday. "Nobody is talked over" can. Choose two or three behaviours specific enough that someone could say tonight whether they held today.

Decide now what the first slip looks like, and who answers it

Callum stayed quiet because the remark was small and nobody had agreed that small was in scope. Agree in advance that anyone can say "that one lands badly", lightly, in the room, and that the senior person present will say "agreed". The first person to do it should not have to defend it alone.

All three depend on a person remembering on the day, and that is the weak point of any fix.

How MeaningFuel carries the repetition

The first two are for you to set up once. The part that should not depend on anyone's memory is the repetition.

The daily insight

A company chooses the values and skills it holds, and each working day everyone gets one short insight on one of them. For Callum's firm the ones that bear on this are Respect, which is treating every person with dignity and courtesy; Accountability, which is owning your commitments, including when they go wrong; and Feedback, which is saying something honest and asking for it. In an ordinary week a reader might meet a line on how people are spoken to when nobody senior is watching, and another on what it takes to say "that one lands badly". It will arrive in month four as it did in month one, for as long as you keep it running, because nobody has to remember to send it.

It is not a fix for Gareth's remark, and incivility itself cannot be chosen: a problem cannot be chosen for the daily culture. What the daily insight keeps alive is the vocabulary a team would need to answer him.

The module: Civility and Respect

Civility and Respect is the module about stopping incivility and bullying. Whoever runs the account or the team books it, it invites the people on that team, and those who join get one short lesson by email each working day: a scene, a question with four answers, and a small action for the day. In its last lesson the team chooses two civility norms of its own, specific enough to say tonight whether you kept them today, practises them daily, and chooses again every quarter. The idea borrows from CREW (Civility, Respect and Engagement in the Workplace), a programme from the US veterans' health service that has been tested in controlled studies and rests on a team choosing its own norms.

Two things in it speak to Callum's Tuesday. One lesson hands the people who see a slight a few small moves for the moment, and one of them is a question that hands the floor back to Ines. Another says a slight from a lead is heard louder, so the lead's repair counts for more, and the lead repairs first. The team lead runs two ten-minute debriefs from a written guide, and check-ins follow two and four weeks later, which is the part of a fix that usually has nobody's name on it.

What Callum could read in month four

Insights is live whenever you look, and a summary arrives weekly or monthly, as you choose. Callum could read how much of each team takes part, week by week, pooled, with nobody named, in month four as easily as in month one. That is a line that does not rely on his memory, and it is all it is. Taking part is not behaviour. A team can answer every morning and still laugh at the wrong moment. A falling line tells him where to go and ask, and a steady one tells him nothing about the planning meeting.

What it does not do

It is not an investigation, and it promises no result. It puts the second dose, and the third, on a working-day schedule.

Questions on making culture change last

How long does culture change take?

Longer than the project plan says, and no honest study here gives a figure. What the WorkNest survey shows is the shape: an early improvement, and then, for many, a return. So judge a fix by what is happening after attention has moved on, not by the first few weeks.

Why does culture change fail?

Often because the diagnosis was treated as the cure. A review and an all-hands are single doses, and the first real test comes later, on an ordinary day, without an audience. If nobody has agreed what happens then, the old behaviour returns to a room that has learned the standard was for the launch.

What should you do when the old behaviour comes back?

Answer it the same day, lightly, in the room, and expect to do it more than once. Treat the return as the test you always knew was coming, not as proof the fix failed. Then check who else saw it, because they are deciding what the standard is worth.

How do you know whether a culture change has stuck?

Ask the people in the meetings, on a date set in advance, what they have seen in the last month. Look for specific behaviours and not feelings: whether anyone was talked over, and whether the senior person answered. Three honest conversations will tell you more than a slide.

Next Tuesday

The review was one dose. Month four is where you find out whether anything is still being taken.

Callum's Tuesday comes round again next week. Whether Gareth's remark gets a reply will depend on what happened in the weeks between, not on the slide.

• Culture Change
• Behaviour Change
• Habits
• Managers
• Workplace Culture