Rebuild or replace?

The decision to continue working with an existing leadership team is the part that gets skipped in most accounts of turnarounds, partly because it is unglamorous and partly because the standard narrative has the new executive arriving with a plan and a hiring budget. The rebuilding version has no equivalent story arc. It consists of a sequence of specific, mostly small interventions, and the interesting thing about it is how quickly the results change once they are made.

Reading who you actually have

The first task is to form an independent view of each person, and the obstacle is that the fastest available source of opinion is the previous management, whose assessments produced the current situation. A leadership team that has been performing badly usually comes with a ready-made explanation of who is weak, and that explanation deserves no weight at all until it has been tested, because in a company where the structure is the problem the people identified as weak are frequently the ones who kept saying so.

The assessment that works is behavioural and takes about three weeks by watching what each person does when a decision is genuinely theirs to make. Watching what they do with a problem that has no owner, watching whether their description of what is wrong is specific and matches what the numbers say, or general and matches what the previous leadership said. Or watching who other people go to when they need something resolved, which is often not the person on the org chart.

What emerges from three weeks of that is usually a different map from the one on the wall. The person described as difficult turns out to be the one who has been objecting to something that genuinely warranted objection. The person described as reliable turns out to be reliable at doing what they are told and unable to decide anything. Neither of those is visible from performance reviews written under the old structure.

The most common finding

In most of these situations the dominant problem is placement rather than capability. People are in seats that do not match what they are good at, usually because the seats were created around whoever was available at the time and never revisited as the company grew.

Let’s take a pattern: a strong individual contributor was promoted to run a team because that was the only way to pay them more, and they are now doing a job they did not want and are not suited to, while the thing they were excellent at is being done by nobody. A commercial person is running delivery because delivery was on fire two years ago and they stepped in. Someone with real judgment about the product is sitting in a support function because that is where they were hired and nobody reconsidered.

Fixing this does not require anyone to leave but the seats to be redrawn around the people, and it produces disproportionate results because it converts underperforming people into performing ones without any change in headcount, cost, or institutional knowledge. It is also fast: a redeployment can be decided in a week, while a replacement takes a year to complete.

Redeployment without demotion

The difficulty with moving people is that any move away from managing a team reads as a demotion, and a demotion that is not handled carefully produces the resignation the whole approach was designed to avoid.

What makes it work is separating scope from status. The individual contributor moved out of a management seat keeps the title, keeps the compensation, and is given a mandate that is visibly important: owning the product economics, owning the largest client relationship, owning the thing the company is worst at. The move is framed as putting them on the problem that matters most, which is true, rather than as removing them from a team, which is also true and is not the useful framing.

This works because most people who were promoted into management to secure a pay rise did not want the management. Offered a route back to the work they were good at, without losing standing or income, they take it with relief. The ones who genuinely want to manage are visible immediately in that conversation, and they should stay where they are.

Raising the standard without changing the person

The second lever is the standard itself, and it moves faster than most incoming executives expect.

In companies that have drifted, the standard has usually not been stated for years. Nobody has said what an acceptable response time is, what a deal has to look like before it gets approved, what a project plan has to contain, or what happens when a commitment is missed. In the absence of a stated standard people converge on whatever was last tolerated, and the level descends slowly because each individual instance of tolerance is reasonable.

Stating the standard explicitly, once, in specific terms, changes behaviour in weeks, as most people in a drifting company are aware the standard has slipped and have been waiting for someone to say what it should be. The first few times the new standard is enforced, particularly on something visible and small, the organisation recalibrates and the enforcement stops being necessary.

The people who do not recalibrate become visible quickly, which is a useful secondary effect. A clearly stated standard sorts the team more accurately in a month than a year of performance management, because it removes the ambiguity that lets underperformance be explained as a misunderstanding.

The hard cases

Three situations account for most of the difficulty, and none of them resolves through general goodwill.

First, the person who expected the job. Someone internal was passed over for the role the incoming executive now holds, and they are still in the company, usually senior, usually well regarded by the team. Ignoring this does not work and neither does pretending it did not happen. The conversation to have is direct and early: acknowledge that they wanted the role, say what would have to be true for them to get one like it, and give them something substantial to own in the meantime. Some of them actually can become the strongest allies in the rebuild, because they know the company better than anyone and now have a route forward. The ones who cannot get past it declare themselves within a month.

Second, the long-tenured underperformer everyone protects. A person who has been there twelve years, is genuinely liked, and has not been effective in some time. The organisation has built a protective structure around them and will resist any change. The move that works is to make the role rather than the person the subject: define what the seat requires now, in writing, and let the gap be evident to everyone including them. Frequently they arrive at the conclusion first and ask for something different, which is a far better outcome than a decision imposed from outside.

Finally, the person who is capable and has stopped trying. Often the most valuable person in the building, disengaged for a specific reason that is usually still true. Finding out what happened takes one honest conversation, and in most cases the reason is that they raised something important and were ignored. Acting on that thing, visibly, recovers them, and it also tells the rest of the organisation something about how the company now works.

What rebuilding does not mean

The approach is often heard as a commitment to keeping everybody, and it is not that. Rebuilding a team is compatible with removing individuals, and the two are not in tension. What it rules out is the assumption that poor results imply poor people and that a new team is therefore the remedy.

It also does not mean lowering the bar in the name of continuity. The standard goes up, and it goes up early, which is the opposite of tolerance. A rebuild that keeps everyone and asks less of them produces a stable team and the same results, which is the failure mode on the other side of the argument.

And it does not mean the composition never changes. It usually does change for specific reasons that were examined rather than assumed. The difference from the replacement approach is arithmetic: two-three considered changes take two months, and a new leadership team takes a year.

Why it can produce better outcomes

The case for rebuilding rests on three things, of which only the last is about the people.

Speed, because the year available is not long enough to hire and land a leadership team and also fix the business, and one of those has to be given up. Knowledge, because the accumulated understanding of clients, systems, and history inside the existing team is the main input into every decision the first year requires, and it cannot be bought in.

And the effect that is hardest to quantify and easiest to observe: a group of people told that the problem was the structure rather than them tends to start behaving as though that were true. The diagnostic work they had been doing privately for two years becomes available. In practice most of the material improvements in a first year originate from inside the organisation rather than from the person who arrived to run it, and a team that has just been replaced has nobody left who was there long enough to supply them.