The 7 most important questions relating to automation in credit management

3. To what extent can I automate my processes in credit management? 

Gerhard: A high degree of automation provides for more efficient processes. Yet despite this, the aim is generally not to achieve 100% automation. Why is this? There is often a disproportionate amount of time, money and effort involved in automating complex decisions that are subject to exceptions and special cases. Complex decisions are therefore typically left to qualified employees, allowing automation efforts to focus on those parts of the process that can be effectively standardised. Each company decides for itself how far automation should actually go, as well as what exceptions and decision-making procedures it wishes to use.  

4. What about your own practical experience? 

Gerhard: Every company has different requirements, processes and systems. As such, the processes to be automated in credit management also vary accordingly. Companies with limited IT resources typically use standard connectors for integration into their own systems (such as ERP, CRM). In other words, they use existing software from an external provider that covers their requirements as closely and effectively as possible. Although creating a custom solution using an interface (API) can certainly offer greater flexibility and individualisation in terms of automation, this requires more IT resources. 

In many cases, a step-by-step approach to automation is also adopted. Companies then generally prefer to start with a standardised, browser-based web solution, as this allows them to gain some initial experience with external data and potential follow-up processes. Once they’re satisfied and have enough knowledge, they then move on to integrating these into their own systems.  

5. What data should credit managers definitely include?   

Gerhard: A reliable and globally consistent database is critical to securing successful automation in credit management. In fact, this is the only way to apply the credit policy across-the-board internationally. It generally makes sense to combine both internal and external information to increase the overall usefulness and thereby improve the efficiency of automation.  

For existing customers, their payment history with the company represents a valuable source of information. This is then typically supplemented by external information such as annual financial statements or scores/ratings, which can forecast the probability of default. Using integrated data that highlights relationships with other companies and the associated risks has now also become established as a best practice.  

When considering whether to enter into new business relations, the first priority should be to securely identify the potential business partner. Having access to external reference databases is also helpful when performing a master data comparison. Since no internal payment history is yet available, external data from so-called payment history pools is used. This shines a light on the potential business partner’s ability to pay on time based on their track record of paying other companies in the past. More traditional data, such as annual financial statements and credit ratings, is also incorporated when checking new customers.

6. What’s the best approach for credit managers to adopt here? 

Gerhard: Intelligent automation should also be accompanied by a solid credit policy, which sets out all steps performed. For a credit check, it could look like this: 

• Assigning partners to classes (for example A-customers, B-customers, C-customers) 
  This classification is based on the relative importance of the customer for the company. Data such as historical revenue or margins is often used to perform this classification. For new customers, this must then be done on the basis of assumptions.
  
• Defining concrete measures for the risk assessment of each business partner class
  These measures differ based on whether they relate to an initial or ongoing risk assessment, as well as the scope of external information to be collected and how often this is updated.  
  
• Interpreting and deriving follow-up measures 
  The business partner assessment is then performed on the basis of the data points and rules defined in the credit policy. Suitable measures should be defined and subsequently implemented via IT here based on the results of this assessment (for example a red, yellow or green traffic light system). Subsequent steps could then potentially include automatically generated instructions, for example requesting the employees responsible to perform more in-depth manual checks, or fully automatic measures such as imposing a suspension of deliveries. Technical automation is not only capable of depicting the defined rules, but also applying them quickly and in a broad scope without errors. 

7. How long does it actually take to automate credit management processes?

Gerhard: It’s difficult to say exactly how long a project of this kind will take, as it depends on both the scope and complexity of the process. However, it’s important to draw up a realistic schedule and make sure that all individuals and departments involved are brought on board in good time. Aspects like the availability of IT resources, as well as maintenance windows, external parties that are required, training sessions and tests must also be taken into account.  

Subscribe to our blog updates!