A lot of marketing, often little substance: How AI creates real added value in your company
12:53

 Der Markt für Künstliche Intelligenz (KI) boomt. Immer mehr Anbieter preisen ihre Lösungen an. Doch bei genauer Betrachtung entpuppt sich so manche vermeintlich intelligente Neuigkeit als Mogelpackung, schließlich lässt sich allein mit dem Label „Künstliche Intelligenz“ ordentlich Geld verdienen. Mittelständische Unternehmen stehen dadurch oft vor einem Dilemma: Einerseits möchten sie KI-Lösungen einführen, um beispielsweise von Effizienzsteigerungen zu profitieren, andererseits ist die Auswahl von Tools aufwändig und bleibt oft ohne Ergebnis. In ihrem Beitrag erklären die enomyc-Autoren Carla Dausend und Leonard Rampf, welche Fehler Unternehmen vermeiden sollten und was beim Einstieg ins Thema zu beachten ist.

Right now, AI is mainly one thing: a license to print money. There's hardly a new product, service, or solution that doesn't have the “contains artificial intelligence” label. Sales talents know how to exploit the hype surrounding these eager-to-learn algorithms: according to one study, four out of ten European start-ups currently describe themselves as “AI companies,” even though they do not actually use AI technologies. According to Forbes, simply mentioning the buzzword can lead to companies receiving 15 to 50 percent (!) more investment than other technology companies.

Due to the many operational and strategic challenges facing small and medium-sized businesses, many companies could really use a boost like AI right now, whether for process optimization, control, or efficiency improvements.

But not everything labeled AI is actually AI. This was the experience of an enomyc customer's purchaser who wanted to introduce a corresponding tool. Although the tool promised many AI-based solutions, our analysis revealed that it was ultimately nothing more than classic data analysis. A great thing – just without artificial intelligence.

So, a lot of marketing, but often alarmingly little substance. The result: frustration and high costs instead of real efficiency gains, a lack of integration instead of strategic relevance. Above all, practice repeatedly shows that many solutions fall short and are not tailored to the business model of the company in question.

Four typical pitfalls companies should avoid – and what really matters instead


The introduction of AI in small and medium-sized businesses offers enormous potential if approached correctly. However, the initial stages are often marked by misunderstandings, unrealistic expectations, and inefficient selection processes. If you want to reap long-term benefits, there is one thing you should do above all else: focus on your own goals and problems, not just on the latest tools and technologies.

These four pitfalls are particularly common:


  1. Focusing on technology instead of problems
    Many companies begin their AI journey by searching for the “best AI” instead of asking themselves: What problem are we actually trying to solve? The result: Solutions are implemented that do not meet actual needs. A reverse approach is more promising: start with the specific business problem and only clarify in a second step whether AI can make a meaningful contribution. Recommendation: Start with a structured use case workshop in which operational pain points and data potential are systematically identified.

  2. Tool fetishism instead of a target vision
    Whether chatbots, recommendation engines, or predictive maintenance – in many companies, the discussion is dominated by the choice of tools. But tools are only a means to an end. If you don't formulate a clear target state, you can easily get lost in endless comparison tables and test phases. Recommendation: First develop a target vision for processes and results. What exactly needs to be improved? How do we measure success? Only then should the question of technology be addressed.

  3. Lack of customization
    An off-the-shelf solution rarely fits perfectly. Many companies purchase AI products that are not compatible with their own data structure or business model. The result: implementation effort, low acceptance within the team, and disappointed expectations. Recommendation: Ensure that technology partners work closely with your specialist departments. Pilot projects help to check in advance whether the solution fits the specific requirements.

Integration hurdles are underestimated
Even the best AI tool is useless if it cannot be integrated properly. Issues such as data availability, system compatibility, data protection, and governance often come up too late and then lead to expensive surprises. Recommendation: During the selection phase, check how well a solution fits into your existing system landscape. Bring IT and data protection officers on board at an early stage.

The consequence of all four pitfalls mentioned above: Despite high investments in time and resources, the desired effects are slow to materialize, the team is frustrated, and the keyword AI is “burned” for the time being.

Our approach: expertise, clarity, and genuine solution development



Our starting point is always the specific challenge that the customer needs to solve, not the technology. We have tool expertise, but we always question the economic benefits. We know the market and can evaluate offers realistically. However, if necessary and appropriate, we also develop our own solutions. We know which providers really have substance and are not just riding the wave.

We also work in a modular and adaptive way. 80% of our solutions follow a standard logic. They are tried and tested and can therefore be implemented quickly. The remaining 20% are customized to ensure that the tool meets expectations. Whether AI, machine learning, classic data analysis, or rule-based automation: what matters is the impact on everyday business.

Our projects are therefore about solutions that deliver measurable added value: technologically sound, economically viable and usable in everyday life.

Best practices from our projects


Our approach is used where companies are stuck in their operational processes, whether in logistics, sales or purchasing. The spectrum is broad, but the problems are often similar.

Three typical examples:


  1. Logistics: A retail company with its own fleet of vehicles plans its routes manually – current factors such as sick leave, traffic, or delivery priorities are not taken into account. Our solution: An AI-supported route module with proven logic, supplemented by company-specific requirements. Added value: Up to 15 percent fewer kilometers driven per delivery and significantly greater schedule reliability in the event of short-term cancellations.

  2. Sales: Many CRM systems have untapped potential. We help identify relevant cross-selling opportunities based on existing sales data – with a modular analytics approach that can be tailored to specific product ranges and purchasing behavior. Added value: An average of 5 to 10 percent more revenue per active customer relationship and a significantly higher success rate for sales campaigns.

  3. Purchasing: Many risks in the supply chain go unnoticed – until it's too late. Our system automatically analyzes dependencies on strategic suppliers and links this information to external data sources such as news or creditworthiness information. Noteworthy constellations are identified at an early stage. The logic behind this is standardized, while the evaluation remains individually controllable. Added value: Risks are systematically screened and identified at an early stage, resulting in up to 25% fewer delivery failures and significantly fewer (expensive) ad hoc procurements.

Fewer buzzwords, more impact

There's no question about it: choosing the right business analytics or AI solution is critical to success and highly relevant to competitiveness today. At the same time, it is very complex and can quickly overwhelm the resources and expertise of medium-sized companies.

What goals should be achieved with the technology and which KPIs (key performance indicators) are relevant? These questions should be asked at the beginning of the journey into the realm of artificial intelligence, not at the end. After all, those who blindly rely on AI risk high investments that may not generate any added value. With our expertise, we help you gain orientation, identify the tools that are important to you, and find the right solution – perhaps one that is developed specifically for you.

 

Questions and answers about the use of AI in small and medium-sized enterprises

 

What is the best way for companies to get started with AI?

The most sensible way to get started with artificial intelligence is to start with a clear understanding of your own business processes: Where are bottlenecks, inefficient processes, or recurring manual tasks? The analysis should not be technology-driven, but based on business pain points—for example, in planning, customer service, or forecasting.

It is important to take a pragmatic approach with a manageable pilot project that generates real value – for example, by reducing the workload of employees or improving the basis for decision-making. In this way, AI does not become an end in itself, but is used in a targeted manner where traditional methods reach their limits. This focus also helps to build internal acceptance: once initial successes are visible, momentum is created for broader applications.

What should you look for when choosing a solution?

A good AI solution is not recognized by its complexity, but by its suitability for everyday use. It integrates seamlessly into existing processes, can be operated with available data, and shows quick results – e.g., through automated routine decisions or shorter response times in customer service.

Medium-sized companies should focus on concrete experience: Are there reference projects from similar industries? Can the solution be tested on a small scale? What is the effort involved in integration and training?

Those who start small remain in control – and can expand solutions modularly if they are successful. The same applies to digital technologies: evolution beats revolution.

How can you recognize a reputable provider?

A credible provider not only brings technology to the table, but also understands the operational reality of its customers. They don't talk in buzzwords, but provide well-founded answers to questions such as: Which processes can be improved? What data is needed? What can be measured in concrete terms in three months?

Professional providers also support not only the technical introduction, but also the organizational implementation – including change management, training, and ongoing support. They think in terms of the problem, not the product.

In small and medium-sized businesses in particular, it is crucial to rely on partners who develop realistic scenarios – and take responsibility for their implementation.

In which cases can consulting or support from an external consultant be useful?

SMEs often lack not ideas, but capacity, methodological expertise, or a clear picture of how an AI project can be successfully implemented. External consulting can provide valuable guidance here, especially in three typical scenarios:

Uncertainty about the right solution: Often, the first step is not selecting a tool, but asking: Can our problem be solved with AI at all?


    • An external sparring partner helps to distinguish between technological hype and real added value – and thus protects against misinvestments.
    • Limited internal resources: When operational teams are already working at full capacity, there is often no time for structured feasibility analyses or well-founded selection processes. Here, an external consultant with a clear methodology can provide support – from use case screening to pilot support.
    • Strategic relevance on a tight budget: Targeted prioritization is crucial, especially when investments are high and leeway is limited. Consulting ensures an objective assessment of potential, supports risk evaluation, and creates a basis for decision-making—for example, for management, investors, or supervisory bodies.

 

Get industry insights now!