How do we change the company with AI while it keeps running?

After a heavy storm, the forest quickly shows what holds. The spruce monoculture often lies flat, trunk against trunk. The mixed forest next to it is still standing. Trees have fallen there too, but the forest as a whole has held.

That is not because nothing changes in the mixed forest. On the contrary: it renews itself all the time. Old trees fall, young ones grow back. And because many species stand side by side, no storm hits them all at once.

SMBs work in a similar way. Many have existed for decades because they have reinvented themselves again and again: with new machines, new markets, with digitalization. It is no coincidence that the idea of sustainability comes from the forest. In 1713, the Saxon mining official Hans Carl von Carlowitz called for a “continuous, steady and sustaining use” of the forests. Thinking long term is the same attitude in the forest as in an SMB.

Artificial intelligence is the next of these renewals. It reaches deeper than the earlier ones, and it comes faster.

Why it concerns the whole company this time

Digitalization mainly transferred existing workflows into software. The paper form became an input screen, the filing cabinet a network drive. The work itself stayed as it was. Only the tool was new.

With AI it is different. It does not only support the work, it does parts of it itself. It reads inquiries, writes quotes, checks invoices. This changes who does what, who decides what, where knowledge sits and what customers pay for. That concerns not one department, but the whole company.

That is why a new tool alone is not enough. 45 percent of all companies are spending more on AI in 2026 than in the previous year. Yet no company that uses AI says it uses its full potential. 28 percent make little use of the possibilities, and 59 percent none at all, according to the current survey by the digital industry association Bitkom. Companies that have a strategy for digitalization as a whole, by contrast, use AI up to nine percentage points more often. This is shown by an analysis from KfW.

45 %of companies are spending more on AI in 2026 than in the previous year

87 %say they make little or no use of what AI can do

up to +9percentage points more AI use with a strategy for digitalization

The other articles in this series show why AI rarely reaches the result, how it shifts business models, how companies grow despite a lack of skilled people, and how they stay in control when they use AI. Everything leads to the same point: it is about rebuilding the whole company. The question is how to succeed without losing what makes the company strong.

What stays

Rebuilding does not mean throwing everything overboard. What makes an SMB strong decides how the rebuild succeeds: its values, its closeness to customers, its quality, its responsibility for the workforce and its independence. They are not an obstacle to change, but its measure.

This includes the strengths that define SMBs: engineering, optimization down to the detail, precision and reliability. AI can amplify exactly these strengths. Whoever has learned over decades to design a plant better than others has the knowledge an AI needs to really help.

This is also where Europe’s opportunity lies. Europe does not have to become a better Silicon Valley and copy the pace and size of the big platforms. It can build on what it already does well. An economy of many specialized SMBs is not a disadvantage. It is a mixed forest. And reliability becomes a competitive advantage when customers want to know exactly whom they are trusting with their data and workflows.

What changes

What changes can be summed up in five points.

• The workflows. They are rethought from the result for the customer, not just made faster.
• The roles. Employees do less processing and more checking, deciding and looking after.
• The knowledge. What sits in a few heads today is captured and becomes usable for everyone, including the AI.
• The prices. When work gets cheaper, what customers pay for shifts. This is the subject of the article on the business model.
• The leadership. Management steers the rebuild itself instead of handing it to IT.

Rebuilding while the business keeps running

Many SMBs have had bad experiences with large IT projects. They are in good company. In 2018, Lidl stopped the introduction of a new merchandise management system after seven years. According to reports in Lebensmittel Zeitung, around 500 million euros had been invested by then. At the confectionery maker Haribo, there were supply shortages before Christmas in the same year. The switch to a new merchandise management system was seen as one reason.

The lesson: the big bang, in which everything is switched over at once, is the biggest risk. The better way follows the mixed forest: not everything at once, but one workflow after another. Each is rethought from the ground up, not just improved a little. Each is implemented while the business keeps running. And no dependency on a single provider that decides prices and direction alone.

In practice, a path in stages has proven itself. This is how we walk it with our clients:

1. 01Weigh & Assess. Set goals and key figures, clarify the framework of data, systems and law.
2. 02Capture & Connect. Capture the knowledge from heads, files and folders and bring it together.
3. 03Design & Rethink. Redesign workflows from the customer’s benefit and check feasibility.
4. 04Plan & Prioritize. Rate every workflow by benefit, effort and dependencies, and set an order.
5. 05Implement & Automate. AI takes over routine work, step by step, while the business keeps running.
6. 06Embed & Establish. Train teams, set clear rules for AI, make the new the standard.
7. 07Expand & Grow. What works at the core carries over into further areas.

Every stage ends with a result that can be measured. That sounds like a long road, but it is not. The first workflow can run after three months. While it takes effect, the next one begins. Step by step does not mean slow. It means safe.

What management has to do

A rebuild of this kind cannot be delegated. Three things lie with management.

Tell the future. In Germany, people talk a lot about risks and little about pictures of the future. The futurist Florence Gaub writes that a future does not have to be good for us to fear it less, or not at all. It has to be something we can influence. For management this means: drawing a picture of where the company will stand in ten years. And showing which step leads there today. Thinking back from the goal like this makes the future something you can shape, for the workforce too.

Keep up the pace. Stages are no reason to wait. The signs in the article on the business model show that markets are already shifting. Better one workflow with a measurable result every quarter than a large project every two years.

Bring the workforce along. Whoever simplifies their own work with AI must not lose their job because of it. Only once that is clear does the company grow like a mixed forest: with many who grow along, instead of a few who carry everything. How that works is described in the article on growing with the same team.

So how do we change the company with AI while it keeps running? Workflow by workflow, with what makes the company strong as the measure. In the end the whole company is affected, but never everything at once. This is how the mixed forest renews itself too, tree by tree, and why it survives the storm. Whoever shapes this renewal themselves also decides what stays.