> This text was published in the ‘BCR Receivables Technology Report 2025’

There is no doubt about the pioneering role of AI – also in the factoring sector. It generally leads to efficiency gains through automation, accelerates internal innovation development, and supports decision-making.¹ The process-related changes are sometimes so far-reaching that one must inevitably ask about their impact on a company’s structure, strategy and even culture: What should be considered when implementing AI in order to avoid serious mistakes in further development? We want to discuss some must-do’s, no-goes, and nice-to-haves here.

The use of AI has a direct impact on the structure of a company: Processes and roles change, are broken up or redefined. Such changes must not be left to their own devices, but should at best be accompanied by a parallel adaptation of the company’s own strategy. In other words, the aim here is to develop and (seamlessly) integrate an AI strategy into the general corporate strategy. However, as such changes usually involve a lot of work – including the adjustment of responsibilities, roles and functions – the structure is often left untouched. At worst, this encourages the formation of silos in which AI projects are driven forward in isolation from one another.² At the very least, it creates unrest among employees because it is not clear who is responsible for which tasks.³

So, what needs to be done to successfully implement AI in a company? At this point, we should be aware that something like this does not happen overnight, but can take months or years. It is therefore all the more important to proceed step-by-step – starting with a fundamental clarification: Which AI use cases harmonise with the company’s objectives? What added value can be expected? As a rule, one of the following goals can be achieved with AI: increased sales, cost savings, expansion of the product portfolio. The more precise the answers to questions about opportunities ( and risks) are, the easier it is to estimate and prioritise the benefits and costs.⁴ Ideally, a specification sheet is developed during this phase in which all requirements and expectations of the AI project are specified.⁵

In this initial step, it is important to get everyone involved (managers, developers, IT employees, project management, etc) on board in order to increase their acceptance, understanding and, last but not least, trust in the AI implementation. It helps to involve opinion leaders who are in favour of the whole thing, as experience has shown that such change processes meet with greater resistance. It is also important to ensure continuous evaluation in order to receive regular feedback from those affected.⁴

Next, you should look at the necessary data quality and availability. Is the data sufficiently available, clean, structured and accessible? How can data protection and security be guaranteed? Once this has been clarified, it’s time to address the fundamental question: buy or develop in-house? In addition to the cost factor, the dependency on other providers must also be taken into account. In principle, it can be said that in-house developments are recommended if the application relates to the company’s core business. No matter whether make or buy, the next step is integration into existing systems, including a test run and final release. This pilot phase is followed by the scaling phase, in which the application is gradually rolled out on a larger scale.⁵

We have already mentioned that AI-related changes in the organisation must also be mapped structurally and strategically, otherwise there is a risk of inefficiencies. At this point, A Chandler and H Mintzberg, who postulate an interaction between strategy and structure, are well worth quoting.6 However, let’s look at another factor that is directly influenced by the structure – the culture of a company. More precisely, it is about promoting an open culture of communication and innovation throughout the entire organisation.⁷ The topic of innovation culture is often underestimated, but it is extremely important to take the human factor sufficiently into account. A practised culture of innovation makes it possible to engage with new technologies, develop new ideas and think outside the box. At the very least, this can lead to a creative approach to AI, as well as to new products, services and business models. It is therefore extremely important to create a flexible environment that encourages experimentation and a willingness to take risks. External and internal training courses that familiarise employees with the effective use of AI should not be forgotten at this point. It is also worth considering the extent to which it would make sense to set up cross-functional teams in order to promote a culture of mutual exchange.⁸ Ultimately, all of this is about creating trust in something new.

On your AI journey, i.e. your company’s path to AI, you will go through several phases, all of which are associated with specific challenges (information phase, preparation phase, development phase and integration phase). It is important that you realistically assess your AI readiness at the beginning and start your individual AI journey on this basis. This should be understood as an ongoing process that never really ends.⁹ Incidentally, AI readiness can now also be seen as an attraction factor in the search for talent: digital natives are more interested in companies that are already equipped for the future than those that are still struggling.¹⁰ As you can see, the effects of AI implementation are sometimes far-reaching.

There are also those who describe AI as exaggerated hype. This applies to individual characteristics and is related to the fact that AI often cannot fulfil some of the expectations placed on it, or cannot do so without further developments. The hype cycle model for technical innovations can be used to gain a better understanding of this. According to this model, expectations skyrocket when a technological innovation becomes known (example: ChatGPT at the end of 2022). From this ‘peak of expectations’, the innovation then plunges into the ‘valley of disappointment’. Those who stay on the ball and gain experience instead of jumping off will find their way to the ‘path of enlightenment’ and from there to the ‘plateau of productivity’. AI models outside of GenAI have already reached the ‘plateau of productivity’ or are well on their way there.¹¹ For you, this means – remain critical of some expectations (internal or external). But also continue to be curious and try things out – after all, AI is about discovering new possibilities and limits.

AI has become a buzzword. But what lies behind it is much more than just a trend. It is an industrial revolution that will significantly change our entire economic and social lives – and we are only at the beginning. AI is capable of giving companies an enormous boost in efficiency and quality if you embrace it and don’t get stuck in a rut.¹⁰ I encourage you to start your own AI journey in factoring.

¹ Adopting AI-Driven Change Management: Key Strategies for Organizational Growth
² neuland.ai: KI-Strategie – häufige Stolpersteine vermeiden
³ L´Ortye, Hendriks: Struktur folgt Strategie, oder?
⁴ Intel: KI-Implementierung: In kleinen Schritten zum Ziel
⁵ Mittelstand-Digital: Schritte zur Integration von KI in KMU
⁶ Alfred D. Chandler: Strategy and structure, 1962
⁷ Handelsblatt live: Eva-Christiane Diemar – KI? Nie ohne Strategie!
⁸ Plamen R. PhD: The Chicken or The Egg: Does Structure Follows Strategy or Strategy Follows Structure?
⁹ Mittelstand-Digital: Wie KI-ready ist Ihr Unternehmen & Fahrplan für Ihre KI-Journey
¹⁰ inpact media, Mirko Heinemann: Wann ist KI sinnvoll – und wann nicht?
¹¹ Gartner Hype Cycle: Wie man Technologie-Hype interpretiert