AI Competitive Pressure: How Mid-Sized Companies Tell Whether They Are Falling Behind

A mid-sized company is falling behind on AI when competitors demonstrably deliver faster, price lower or serve customers better, and the difference traces back to their use of AI. Without that evidence, what you feel is pressure from reports and industry conversations. Five signals and a simple calculation show which case applies and what action follows.

The pressure is real. Since ChatGPT appeared at the end of 2022, the media report daily on companies changing their operations with AI, and managing directors hear at association meetings that a competitor is already “doing something”. Those who invest in haste pay for the lesson, and those who wait risk a genuine gap. This article helps to avoid both mistakes. It shows what surveys on the spread of AI actually establish, which signals reveal a gap and how the cost of waiting can be put into figures.

How widespread is AI among German mid-sized companies?

In January 2024 AI is still a minority activity among German companies, mid-sized ones included, and most do not use it. The ifo Institute reported in August 2023, from its business survey of June 2023, that 13.3 per cent of the companies surveyed in Germany use AI. A further 9.2 per cent planned to. In industry, about a third either used AI or planned to. The statistical office of the European Union had found for 2021 that 8 per cent of EU enterprises with ten or more employees used AI technologies.

The figures put into perspective the impression that everyone else is further ahead. Most of your competitors are in the same position as you: they are discussing and experimenting, and only a few applications run in regular operation. That does not rule out a single competitor having an edge in one process. The signals below test whether that is so. Why reports of success rarely show the whole picture is explained in the article on AI FOMO in companies.

How do you recognise a genuine gap?

A genuine gap shows in features that customers or figures confirm. The grid below separates verifiable signals from presumed ones.

Two strong signals justify a targeted review of the processes affected. Weak signals alone trigger no investment and belong in monitoring. An assumed example of handling a weak signal: a machine builder reads that a competitor is introducing an AI assistant for spare-part enquiries. That is a weak signal. It becomes strong when customers report that the competitor answers enquiries in minutes while its own answer takes a day.

What does waiting cost?

The cost of waiting is the difference between your process and the competitor's better process, multiplied by volume and value. The calculation turns a feeling into a figure. The example below uses assumed values.

The result shows two things. First, waiting is not free, and the calculation shows how much time remains for a review. Second, a process with a difference of a few hours is not a question of competitiveness. Which applications to tackle in which order is described in the article on AI priorities for 2024.

What do studies say AI delivers in daily work?

Studies show measurable productivity gains for clearly defined tasks, and they show limits. A field study by researchers at the Massachusetts Institute of Technology and Stanford University with more than 5,000 customer service staff found that an AI assistant raised the number of issues resolved per hour by 14 per cent on average, and by 34 per cent for novices. In an experiment with professional writing tasks, completion time with ChatGPT fell by about 40 per cent, and independent raters judged the quality of the texts to have risen. A study of 758 consultants, by contrast, found that the same tools reduced the number of correct solutions on tasks outside their capabilities.

Both findings lead to the same conclusion. The benefit arises in processes the user knows well, and it fails to appear where nobody checks the results. A competitor that introduces AI broadly without understanding its processes therefore has no secure lead. How executive boards turn pressure into a plan is shown in the article on AI activism.

What to do if the signals point to a gap

If at least two strong signals point to a gap, action follows in four steps, beginning within a few weeks.

1. Narrow down the process affected and name an owner in the business unit.
2. Measure the baseline: time and cost as well as quality of the process today.
3. Build a prototype with real data in the real process and test it against the baseline.
4. If it succeeds, plan the introduction, with a person responsible for operations and trained users as well as a metric after 90 days.

For the executive team there is a further point: those who must answer shareholders' questions on AI will find in the article on pressure from shareholders the questions for which reliable answers should exist.

AI competitive pressure and Niehues Consulting

Niehues Consulting is a consultancy for AI management and organisation design, seen from the delivery perspective, based in Düsseldorf. Founder Marin Niehues brings AI into operation, in projects ranging from VC-funded scale-ups to DAX 40 corporations and in more than 20 countries on six continents. In the service area of AI management the work runs from business analysis and requirements through the business case to the selection and sequencing of applications and the steering of the portfolio.

The approach follows the Niehues Delivery Model. For the question of a gap, three of its test questions matter most: which business result should change measurably? What is the baseline, and how do we measure success? Is the business case confirmed with real values? The starting point is a free initial consultation of 30 minutes, which you can arrange via the contact page. Scope and terms of an engagement are discussed there.