SME Digitalization in Germany in Numbers: Where Businesses Really Stand

Why are average figures on digitalization of limited use to you?

In the morning you read that most companies use artificial intelligence. At noon you hear that only one in four does. In the evening a development bank reports that digitalization in small and mid-sized businesses has slumped. All three statements come from reputable sources, and all three are correct. They simply measure different things.

It starts with the word Mittelstand, the German term for the small and mid-sized, often owner-managed companies that shape the country's economy. The Institut für Mittelstandsforschung (IfM) Bonn, a research institute dedicated to this sector, counted around 3.52 million small and medium-sized enterprises in Germany for 2024. That is 99.2 percent of all companies. KfW, Germany's state-owned development bank, includes in its SME panel all private companies with annual revenue of up to 500 million euros. According to KfW, 81 percent of these firms have fewer than five employees. The official statistics on the use of information technology, by contrast, only cover companies with ten or more employees.

To see where you stand, you therefore need the figure that fits your size, and you need to know how it was produced. Every figure here comes with its source and year. The guiding question of this analysis is: where do small and mid-sized businesses really stand on digitalization, and which figure is the right yardstick for your business? To answer it, we first look at how many businesses digitalize and how large the gap by size is, then at what they digitalize and where cloud and AI stand, and finally at why the figures contradict each other and what research says about impact and obstacles.

How many German SMEs are digitalizing at all?

The broadest basis is the KfW SME Digitalisation Report 2025, published in April 2026. It draws on the KfW SME Panel, in whose latest wave 13,079 companies took part. The survey asks about digitalization projects completed in the preceding three years.

The result is sobering. Between 2022 and 2024, 30 percent of small and mid-sized companies completed at least one such project. That is five percentage points fewer than in the previous survey. Spending on digitalization fell to 23.8 billion euros in 2024, down 8.1 billion euros on the year before. KfW attributes the decline mainly to the weak economy and the fading of one-off effects from the pandemic years.

According to KfW, SMEs spend almost eleven times as much on traditional investment as on digitalization. Over the long run the report still points upward: the share of active companies and the level of spending are higher than when the survey began in 2017.

How wide is the gap between small and larger businesses?

The average hides the most important finding. Digitalization in small and mid-sized businesses is above all a question of size. According to KfW, larger SMEs with 50 or more employees carry out digitalization projects 2.3 times as often as firms with fewer than five employees. In the period from 2014 to 2016 that factor was 1.9. The distance has grown.

Larger SMEs therefore spend almost 19 times as much as the smallest firms. They make up only 2 percent of SMEs but account for 41 percent of digitalization spending.

That does not mean small firms are stingy. Measured against revenue, they often carry the heavier load. 13 percent of active firms with fewer than five employees spend 7 percent or more of their annual revenue on digitalization. Among larger SMEs this applies to only 1 percent. Small firms can therefore afford fewer failed attempts. It matters all the more that a project starts at the right bottleneck. How to find it is described in the article on why the process comes before the tool.

What exactly do businesses digitalize?

KfW also asks about the type of project. Most often, SMEs digitalize their contact with customers and suppliers, which according to the report frequently means new websites, online ordering and payment systems, social media, or data exchange with partners. Next comes the renewal of IT structures, meaning new hardware, systems, or applications.

The table shows a pattern. Small firms start at the interface with the customer. Larger ones work their way inward: they connect departments, reorder workflows, and build knowledge. KfW therefore reads the customer interface as a typical first step for late adopters.

For sales, that is good news. The first step lies where inquiries arise. What such an interface has to deliver is covered in the checklist for a website that brings in inquiries. Specialized service providers can handle the implementation, for example quik-services, a provider of websites, SEO, and lead generation and a business partner of Owners Council. The second step leads inward, for instance to a shared customer file. Which features a small team needs for that is clarified in the comparison of CRM for small businesses.

Cloud and AI: where do companies stand according to official statistics?

Germany's Federal Statistical Office (Destatis) surveys every year which technologies companies with ten or more employees use. For 2025 it reported that 54 percent buy paid IT services from the cloud, meaning computing power, storage, or software over the internet. Among small companies with 10 to 49 employees the figure is 51 percent, among medium-sized ones 65 percent, and among large ones with 250 or more employees 86 percent. The cloud is used mainly for email (76 percent of users), data storage (71 percent), and office applications (68 percent). Only 23 percent of cloud users obtain customer relationship software this way.

For artificial intelligence the slope is steeper. In 2025, 26 percent of companies used AI technologies.

In 2024 the office had still reported 20 percent. The comparison is limited, however, because the statistics have used a different definition of an enterprise since the 2025 reporting year. In a European comparison Germany is above average: Eurostat gives a share of 20.0 percent for the EU in 2025, after 13.5 percent in 2024.

Eurostat also measures digital intensity, an indicator built from several technologies. In 2024, 73 percent of small and medium-sized enterprises in the EU reached at least the basic level, compared with 98 percent of large enterprises. The EU's target for 2030 is more than 90 percent. Where AI saves time in selling today is shown in the comparison on AI in sales.

Why do the figures on AI use contradict each other?

In September 2026 Bitkom, Germany's digital industry association, reported that for the first time a majority of companies use AI: 57 percent, after 36 percent in 2025 and 20 percent in 2024. The Federal Statistical Office, by contrast, gives 26 percent for 2025. This is not a contradiction.

Bitkom surveys larger firms, one year later, and with its own wording. The very smallest firms, which dominate the SME sector in terms of numbers, do not appear in any of the three surveys. For every figure I therefore use a simple check, the three-question test.

1. Who was asked?

   Size, industry, and number of companies surveyed. A statement about firms with 20 or more employees says little about a business with six people.

2. What was counted?

   Use, planning, or a completed project. A company in which one person is trying out an AI tool counts as a user, depending on the question.

3. When, and over what period?

   A reference date, one year, or three years. With fast-moving developments such as AI, a gap of one year changes the result considerably.

More revealing than the headline are the barriers Bitkom reports. Of the companies without AI, 85 percent cite a lack of technical know-how, 66 percent legal uncertainty, and 63 percent costs that are hard to calculate. What the European AI regulation actually requires is assessed in the article on the EU AI Act in sales and marketing.

What does research say about impact and hurdles?

Figures on the current state say nothing yet about the benefit. For that it is worth looking at studies, each with its reach.

Rietmann (2024) deserves the closest look. Between September and December 2020 he spoke with one member of senior management at each of 28 hidden champions, meaning little-known mid-sized market leaders, in four rural regions of Germany. 89 percent of the firms are manufacturers, 54 percent are family-owned, and they range from around 20 to around 5,700 employees. He sorts them by two questions: how does management weigh opportunities against risks, and how available are skilled staff, IT, and network connections?

Four types emerge. Providers of purely digital products have digitalized almost all of their processes. The largest firms regard digitalization as indispensable and start with production and administration. Traditional family firms digitalize individual support processes, which often remain unconnected, and buy in the knowledge they lack. Skeptics, mostly small family firms, see more risk than benefit, do without a dedicated strategy, and stop at a few administrative projects. Market leaders are therefore not automatically digital front-runners. Rietmann stresses that the types generalize and that individual firms deviate. Three shorter findings round out the picture.

• Internal capabilities decide. Bielawska (2025), in a literature review, examined existing studies and Eurostat data on small and medium-sized enterprises in Poland, Germany, and Denmark. According to this work, internal capabilities are the strongest predictors of AI adoption. Hurdles are data quality, skills gaps, costs, integration with legacy systems, privacy, and cultural resistance.
• The return does not come by itself. Noja and colleagues (2026) examined country-level data for the 27 EU member states from 2016 to 2024. In some of their models, digitalization was negatively related to indicators of firm performance. The authors read this as adaptation costs and conclude that complementary investment in skills is needed.
• Digital firms find the next transition easier. Kesidou, Ri, and Roper (2026) surveyed 964 small and medium-sized enterprises in the United Kingdom. Digitally advanced firms were more likely to also adopt innovations that reduce emissions.

This fits with an analysis by KfW and the research institute ZEW from April 2026, based on panel data for 2017 to 2022. A 10 percent higher digital capital stock was associated on average with 0.159 percent higher productivity. Among the most digitalized companies it was 0.808 percent. The benefit thus grows with maturity.

The digital divide among SMEs separates businesses that see a project through from businesses that stop at buying software.

First steps for this week

Use the figures as a yardstick for an honest assessment of where you stand. Which few metrics you then look at regularly is described in the essay on data-driven decisions.

• Write down which digitalization projects your business has really completed in the past three years, meaning with a changed routine in day-to-day work.
• Add up last year's spending on software, hardware, service providers, and training, and set it against revenue.
• Check at which points customers can contact you digitally today and where phone calls, paper, or retyping are still needed.
• Record whether customer data sits in one shared place or in the inboxes and spreadsheets of individuals.
• Ask the team who already uses AI tools, for what, and with which data.
• Choose one single project for the next 90 days, with one person responsible and a measurable result.
• Apply the three-question test to every study you come across before you derive pressure from it.

Back to the opening question, which was where small and mid-sized businesses stand on digitalization and which figure is the right yardstick for your business. Above all, the figures show a gap by size: small businesses spend less and use cloud and AI far less often than large ones. Which figure applies to you depends on who was surveyed, what was counted, and which period was covered. An average across all businesses is therefore rarely a good yardstick; the figure for your size class is.