Transformation in mid-sized companies: when a rebuild is necessary and when adjusting is enough

How do you know a rebuild is necessary?

By a single test: will your business still make money in five years if the market keeps developing the way it looks today. If the honest answer is no, a rebuild is due. If the answer is yes, but with worse numbers, then adjusting is the job.

This test is uncomfortable because it demands a statement about the market instead of about your own performance. Three signals point to a rebuild:

• The customer buys something different. Not less, but different. A company selling equipment whose buyers will purchase uptime instead of machines does not have a sales problem. It has a different business ahead of it.
• The service is delivered differently. When part of the value creation can be replaced by software, regulation or a new production method, the share you end up earning shifts.
• Competition comes from a different direction. A new player who runs your business as a byproduct does the math differently than you do. Price pressure from that direction cannot be answered with efficiency.

If none of these signals are there, a rebuild is the most expensive way to work on an earnings problem.

You can back the test with numbers instead of debating it in a meeting. Take the three largest product groups and model a five-year trajectory for each: volume, price and contribution margin, assuming the visible development continues. If two out of three groups fall below the threshold where they cover their fixed costs, the question is answered. This calculation takes an afternoon and ends the discussion about whether it is really that serious.

What separates adjusting from a rebuild?

The scope of the decisions, and with it the amount of attention they tie up.

The last line explains why the assessment matters more than the approach. A rebuild that never had to be one burns trust. Adjusting in a situation that calls for a rebuild pushes the problem into a time when there are fewer resources to handle it.

What phases does a rebuild go through when it works?

Three, and they overlap only at the edges. The most common shortcut is jumping from the first to the third.

• Clarify. What will hold in five years, what will not, and which capabilities are missing for it. This phase ends with a decided direction, not with an action plan. Three to six months.
• Build while the old business runs. The new business gets its own people, its own budget and its own metrics. It is not measured against the numbers of the core business, otherwise it loses every comparison. One to three years.
• Switch over. The new business becomes the standard business, the old one is wound down or sold. This phase requires the hardest decisions and is the one people most like to postpone.

Between phase two and phase three sits the point where most rebuilds get stuck. The new business exists, it does not earn enough yet, and the old business keeps funding it. As long as nobody sets a date for the switch, this state stays stable and costs money every year.

Who drives the rebuild day to day?

One person from management who gets time freed up for it. Not a project manager with a steering committee, and not the management team as a whole.

The difference is practical. A rebuild generates weekly decisions that cut across existing responsibilities: a sales territory gets redrawn, a price gets calculated differently, a resource goes into the new business instead of the existing one. Decisions like these get made by someone with a mandate, or by no one.

That includes a written release from other duties. Anyone who owns the rebuild while continuing to run their own unit unchanged will run the unit. That's not a question of character. It's a question of what comes first in the calendar. In practice: two days a week are locked in, a deputy takes over day-to-day operations, and both are stated openly in the leadership team.

How do you keep the organization with you?

By saying what stays. During periods of change, there's a lot of talk about the new and almost none about the existing, and that's exactly where the unease comes from.

Three statements help more than any presentation: which parts of the business keep running, which sites and roles are not up for discussion, and when the next update comes. The third point is the most effective. A fixed date takes the ground out from under rumors, even if there's nothing new to report that time.

Resistance is information, not noise. Someone in sales who has served the same customers for twenty years and says the new offering doesn't fit there usually has an observation you can verify. It deserves to be checked, not dismissed as clinging to the past. After the check, the decision still gets made, but it's better grounded and the person stays on board.

How do you measure whether the rebuild is making progress?

By figures that show up before revenue does. Revenue from the new business comes late, and until then leadership needs something that tells them whether it's working.

Three numbers are useful. How many customers have seriously evaluated the new offering, meaning a meeting of their own and a question about price or timing. How many people in the company are working on the new business and how much of their time actually sits there. And what share of investment goes into the new business, because the budget shows the real priority far more clearly than any explanation.

If these three numbers stay flat over two quarters, the rebuild isn't running, no matter how many workshops have taken place. That's the moment for a decision about mandate, budget and time, not for another workshop.

If you have an advisory board, a bank or shareholders outside of management, report these three numbers to them as well. It sounds like extra work and it works in both directions: the group gets a question it can ask, and management gets a date on which standstill becomes visible. A rebuild without that outside view loses its urgency after about a year, because inside the company everyone has gotten used to the slow pace.