Everybody Agrees With This. Almost Nobody Does It.
Everyone knows why change initiatives fail. Almost nobody builds the thing that would stop it.
I can tell a room the three things that keep a decision alive in about ninety seconds.
Somebody has to own it, by name. It has to live inside something that was already happening. And it needs cover from above for longer than anyone plans for.
The room nods. Almost always. Nobody has ever pushed back on the list itself.
Then I ask them to name the person for the decision they made in March, and the room goes quiet.
The gap between the nodding and the silence is what interests me. Everybody agrees with the list. Almost nobody walks out of the room and builds it.
First, where this sits
If you work in this field you have already noticed that two of those three sound familiar.
Ownership, embedding into business as usual, and executive sponsorship have been in every change methodology since Kotter. I am not claiming to have found something the field missed. If you have run a Prosci-based programme you have met all three under different names.
Three things are different in what follows, and they matter more than the overlap.
The standard models are about getting a decision adopted. They take you to the point where the new way is live and people are doing it. That is where they largely stop, and it is where my question starts. Adoption is not survival. Most of what I have watched die was adopted successfully first.
"Embed it into business as usual" is advice everybody gives and almost nobody follows literally. What organisations actually do is build a new governance structure to manage the embedding. A steering group. A monthly review. A dashboard. That new structure is usually the most fragile thing in the whole arrangement, and I will make that case properly below.
Sponsorship models assume sponsorship is something you can secure and maintain. In government and in donor-funded work, you cannot. The discontinuity is structural and everyone can see it coming. Advice built on holding sponsorship steady is not difficult to follow in those settings. It is inapplicable, and pretending otherwise wastes everyone's time.
So: familiar list, different question, and one place where I think standard practice actively causes the failure it is trying to prevent.
Now, why nobody does it.
Naming a person means naming someone to blame
Ask a leadership team to name one person for something, and watch what happens. You will get a directorate. You will get "the safeguarding team". You will get a working group with seven members and a terms of reference.
What you will rarely get is a name.
There are good reasons for this, and none of them are carelessness. In most organisations, naming an owner does two uncomfortable things at once.
It names who will be asked when it goes wrong. Nobody volunteers for that, and few managers will impose it on someone they like.
And it exposes the resourcing. The moment you write a name against a decision, somebody asks what that person will stop doing to make room. Usually the honest answer is nothing, because the decision was made without anyone costing the carrying of it. A group can absorb an unresourced responsibility indefinitely, because a group is nobody. A named person cannot.
So the group stays a group. Everyone is relieved. And the decision has an owner in the paperwork and nobody in the building.
I worked with an organisation where everybody already knew what the culture problem was. We ran the training. I facilitated the conversations. Leadership commissioned the work and turned up to most of it.
Nothing changed for months.
The leaders had committed to paying for a process. Their own behaviour stayed exactly the same, and their behaviour was the actual problem.
What moved it was one line in a document. The head of the organisation was named as the person accountable for the behaviour standard. Not the HR director, not a working group. The person at the top, whose own conduct was part of what was being examined.
Things started shifting within weeks. Nobody was disciplined. It shifted because there was finally one person who had to answer for it.
The version that works is smaller than people expect. One person, one decision, and something specific they are accountable for producing. And in the public sector, attached to the post rather than the individual, because individuals rotate and posts do not.
Using something that already exists looks like you did not try
This is the one I would defend hardest, because it runs against what most people are taught.
When an organisation commits to something, it builds something new to carry it. This looks like seriousness. A new steering group signals that we mean it. A new dashboard signals investment. A new monthly review signals that somebody is watching.
Attaching the same decision to a report that already goes out on a Friday signals almost nothing. It is invisible. Nobody gets to announce it.
And it is the version that survives.
A new routine created to carry a new decision has no protection of its own. No history, no constituency, nobody whose own work runs through it. It is the first meeting cancelled when the calendar tightens, and the second cancellation is easier than the first. Within two quarters it is a recurring invite that people decline without reading.
A routine that predates the decision has already proved it can survive a bad month. It survived last year's crisis. People defend it, because their own work depends on it running.
So the more impressive-looking option is the more fragile one, and we choose it precisely because it looks impressive. The thing that would actually hold the decision is the thing that makes it look like we did not take it seriously.
That is an uncomfortable trade and it is rarely made consciously.
Another organisation had a fraternisation problem. Relationships across reporting lines were common enough that nobody really treated them as unusual. A policy was written. For a long time it made no difference, because nobody acted on it.
Then an enforcement committee was set up and it did the job properly. Cases were heard. Outcomes were visible. People started coming forward. New fraternisation stopped.
An enforcement committee is a new structure, which is the thing I have spent this whole section warning you about. So it is worth saying why this one worked.
It produced consequences that people could see coming. A steering group set up to review progress produces a report. This produced outcomes that people actually felt. That is a more useful test than whether the structure is new or old.
The test I would put to any decision: if nobody chased this for a month, would anything still show up? If a number, a record or a return would appear whether or not anyone remembered, it is attached to something real. If nothing would appear and nobody would ask, it is not attached to a routine. It is attached to somebody's memory.
Asking about leadership continuity means saying they will leave
The third holder is the one people find hardest to raise, and the reason is social.
To ask "what happens to this when you are no longer in post" is to say out loud, in the room, that the person in front of you will not be there. In a lot of rooms that lands as rude. In some it lands as a good deal worse than rude.
So the question does not get asked. The decision proceeds on the unstated assumption that the current arrangement of people is permanent, which nobody in the room believes and nobody wants to be the first to say.
In government the absurdity is sharper, because everyone knows the rotation schedule. It is published. And still the conversation about what happens after it rarely takes place before it happens.
For donor-funded work the end date is in the contract from day one. Everybody has read the contract. The close-out conversation still tends to begin in the final quarter, when the people who could have transferred the work have already started looking for their next post.
Nigeria's crash helmet law is the clearest public example I know.
Enforcement started in 2009. Checkpoints, penalties, a lot of national attention, and compliance moved quickly. For a while it looked settled.
The law was never repealed. It is still there. Enforcement is what went, and the reason is on the record. In May 2022 the Corps Marshal of the Federal Road Safety Corps, Boboye Oyeyemi, told Vanguard: "We have to slow down a bit. When implementing, we met a brick wall in terms of religious and cultural issues. Many of our staff members were physically assaulted and some were killed."
That is an institution saying publicly that it could no longer carry the cost on its own.
There was nothing underneath the enforcement. No routine producing evidence without being asked. Nobody accountable outside the agency doing the enforcing. So when the enforcement stopped, the compliance stopped with it.
Ask anyone in Nigeria when helmet wearing stopped being normal. Nobody can name the day.
The way through is to stop treating it as a question about a person. Nobody has to say "when you leave". The question is: sooner or later somebody else will be making these calls, so what does this look like then? That is a question about the decision, and it can be asked without anyone in the room being asked to contemplate their own departure.
If the honest answer is that it continues if the next person happens to care about it, then the decision is riding on a coin toss that nobody in the room controls.
Why change initiatives fail at the same three points
None of the reasons are about competence.
Nobody fails to name a person because they do not understand ownership. They fail because naming one exposes something about resourcing that the organisation has agreed not to look at.
Nobody builds a new steering group because they have never heard of business as usual. They build it because the alternative looks like indifference.
Nobody skips the succession conversation because they think leadership is permanent. They skip it because raising it costs something socially, right now, and the cost of not raising it lands eighteen months later on somebody else.
Every one of the three requires saying something slightly uncomfortable in a room, in exchange for a benefit that arrives long after the discomfort. That trade is difficult for individuals and much more difficult for institutions, which is why the list everybody agrees with is so rarely the list anybody builds.
What to do about it
Three questions, and none of them require you to say anything awkward about anyone in the room.
Instead of asking who owns this, ask: if this stopped happening for a month, who would notice first, and what would make them notice? You will get an honest answer, because it is a question about a mechanism rather than about accountability. If nobody can answer with a name and a trigger, you have found something.
On the routine, the useful question is: what already happens every week that this can live inside? It moves the conversation away from building and towards attaching, which is where the durability sits.
And on leadership: sooner or later somebody else will be making these calls, so what does this look like then? Nobody has to name a departure. The decision is the subject of the sentence.
Three questions, asked before the launch rather than after the review. They cost one meeting.
The alternative is what usually happens. Everybody agrees the list is correct, nothing on the list gets built, and eighteen months later somebody is asked to write a review explaining what went wrong.
Nothing went wrong. Nothing was holding it.
Mofoyeke Omole is an organisational change consultant working at the intersection of policy, safeguarding and people. Her work has contributed to four national and regional policy instruments, and has been tested with more than thirty organisations and over a thousand practitioners, leaders and public officials.
