How do we grow when skilled people retire and roles stay empty?

In two years, the colleague who has run production planning for thirty years will retire. The inside sales position has been advertised for months, and hardly any applications come in. There would be enough inquiries. But even the current orders only get done with overtime.

This is what it looks like in many SMBs, or something close to it. The question behind it comes up again and again in our conversations with management teams too: how are we supposed to grow when we lack skilled people?

It will not resolve itself. By 2039, around 13.4 million people in the German workforce will reach the statutory retirement age, almost a third of everyone available to the labor market today. The younger generations will not be able to replace them in number. This is what Germany’s Federal Statistical Office has calculated.

13.4 millionpeople in the workforce reach the statutory retirement age by 2039

22 %of small and medium-sized companies see their business held back by a lack of skilled workers

18 %of companies using AI expect more employees as a result (without AI: 9%)

In the next upturn, skilled people will be missing

At the moment, the shortage of skilled workers is less pressing than a few years ago. In April 2026, 22 percent of small and medium-sized companies reported that missing skilled workers hold back their business. Over the last three years, the share of companies saying so has more than halved, according to the KfW-ifo Fachkräftebarometer. The main reason is the weak economy. If the economy recovers as expected, KfW and ifo expect the shortage to grow again. And the baby boomers are retiring either way.

For growth this means: in the next upturn, it is often not demand that holds a company back, but staff. Anyone who wants to take on orders then has to rebuild workflows now so that they manage with fewer people. Where order books are currently moderate, businesses have more room for this than in the boom years. Whoever waits rebuilds under pressure.

Growing without hiring more people

Until now, growth in SMBs almost always meant hiring more people. More orders required more workers, in production as in administration. This calculation no longer works, because the labor market does not supply the skilled people.

AI changes the calculation. It does not replace a skilled worker who sets up a machine or advises a customer. But it can take over a large part of the work around these tasks: sorting inquiries, preparing quotes, checking invoices, writing documentation, gathering knowledge. It is exactly this work that ties up many skilled people today who should really be doing something else. Where it falls away, capacity for growth appears:

• More orders with the same team. Quotes are prepared faster, and orders move through faster. What is left lying today can be accepted.
• Shorter delivery times. The waiting time between sales, engineering and costing shrinks. Whoever delivers faster wins orders, even without being cheaper.
• New services. The time freed up flows into service, advice or new offers. Which ones these can be is shown in the article on the business model.

Capacity does not appear because everyone gets an AI assistant. It appears when a workflow is rebuilt so that routine work falls away. And growth only follows when it is clear beforehand what the time saved is for. How that works is described in the lead article of this series.

Growing instead of cutting

Not every position that becomes free has to be refilled. When someone retires or moves on, a rebuilt workflow can take over part of the work. The work still gets done, over time even more of it, and nobody is laid off. That is realistic, even if it is rarely said openly.

Many companies, by contrast, expect AI to mean fewer staff. In the current survey by the digital industry association Bitkom, 45 percent of companies expect the number of their employees to fall slightly because of AI over the next two years. 12 percent expect a clear decline.

It is striking who calculates differently. Of the companies already using AI, only 9 percent expect a clear decline in jobs, compared with 15 percent of companies without AI. And they are twice as likely to expect AI to increase the number of employees, 18 compared with 9 percent. Those who use AI apparently see growth in it rather than cuts.

Anyone who still uses AI mainly to cut staff should know the consequences. People who fear for their own job do not show their own time savings, and the rebuild stalls. And when the next upturn comes, exactly the skilled people who are then needed will be missing.

What the workforce needs to know

Growing with the same team only works if the team comes along. Three things help.

A clear commitment. Whoever simplifies their own work with AI does not lose their job because of it. This commitment is the precondition for employees to show what AI really brings. If the goal is growth, it can be given honestly.

New roles instead of old job descriptions. Work shifts from processing to checking, deciding and looking after. Whoever records invoices today checks tomorrow the cases where the AI is unsure. Many departments also need someone who looks after the AI applications and helps colleagues.

Learning on your own workflow. 66 percent of companies rate their employees’ AI skills as low. 70 percent now train them in working with AI. Experience shows, however, that employees learn most when they help rebuild their own workflow, not in a seminar.

Where there is a works council, it belongs at the table early. For many AI tools it has a right of co-determination, for example when they can record performance or behavior. Involving it from the start often wins an ally instead of a brake.

Capturing experience before it leaves

With every retirement, knowledge leaves that is written down nowhere: which machine causes problems with which material, which customer needs which tolerances, how to get the old control system running after all. This knowledge can be secured while the experienced people are still there. They tell, the AI helps to capture and organize, and afterwards the result is available to everyone, including the AI applications themselves.

For growth this counts twice. New colleagues are onboarded faster. And new services can emerge from the secured knowledge. For the experienced people themselves it is often a form of appreciation: their knowledge stays, even when they leave.

So how do we grow when skilled people retire and roles stay empty? No longer mainly through new positions, but through workflows that achieve more with fewer people. AI is not the whole answer, but a large part of it. Provided that management decides early what the time saved is for, and tells the workforce openly.