Learn, explore and unleash your inner chef.

 What is Machine Learning (ML) and where can you deploy it?

With Machine Learning, we program an intelligent and flexible set of rules which task is to learn what to do in order to achieve a desired result. You could say it is a smart program which can learn on its own how to do a specific task.

Usually, human feedback is required in order to “train” the program. When the program is not sure what to do or if what it has done is correct. The person will provide feedback which will be used to complete the current task and used as reference in future for similar situations (the program “learns” to achieve the desired outcome).

One of the O2C processes that has high potential for automation and machine learning is the payment application. The process takes as input the remittance advices and the payments received from customers, and produces as output the correct application of the payment to the corresponding invoice.

The Machine learning program could take as input the old remittance advices and payments, invoices the payments were applied to and “learns” how to use a new remittance advice and payment in order to find the correct invoice and apply the corresponding payment. The program could also raise a warning to the responsible person for payment or remittance advices that were not processed or have a higher chance of being wrong (as the program is learning it’s task).

Keep in mind that Machine Learning solution implementation is relatively recent and every business case scenario should be analysed on a case by case basis before starting a project. 

How can success be achieved in Order Processing Automation?

Mainly by focusing on the right activities to be automated and transforming by taking one step at the time. When talking about automation it is important to stress that one-size does not fit all, meaning it is important not to blindly try to automate all activities. Separate transactional activities (e.g. Input a new order in the system) from judgmental activities (e.g. Out of stock, decision on which customer is impacted, and what follow-up actions to take). Also, separate high volume (e.g. Sales order data correctness check) VS low volume activities (e.g. customer credit limit calculation). The high volume transactional activities are a good place to start with. 

What are in your opinion the best automated processes that improve Customer Experience?

One of the key differentiating factors can be the Customer Portal integrated with seamless direct contact with Customer Service. While providing an easy to use self-service web portal you can:

• Be available 24/7 for order taking
• Give additional information and visibility to the customers such as pending payments, open claims, or even credit situation
• Offer additional services (e.g. Simple stock overview, online return request etc.)
• Get insight and information from customers that otherwise would be impossible to have. 

How does the future of customer experience and customer facing supply chain look? 

Again, one size does not fit all! The optimal process will depend on the industry and the type of relationship you have with your customers. Generally speaking, the self-service solutions will grow and cover a larger variety customer service activities.

It will be ok to rely on FAQ and support material, as well as chatbots (as long as we know it is an intelligent robot we are talking to). At the same time, for many industries, the automated self service will have to be paired with exceptional chat/phone/video/physical support done by a person.

As the processes and support offered becomes smarter, “human touch” will evermore become a premium service we offer customers.

 Why should O2C optimisation be derived through utilisation of day2day operational experience? 

Smaller/shorter/quicker feedback implementations enable fast learning and quick iterations. You need to involve people working in the operations as they are the end users and they can connect their needs and desired business results to the to-be process.

When setting up an O2C Optimisation Project the selection of the right people with the right skills is key to success.

On a higher level you want two groups of people: (1) people with process and technology knowledge that can objectively analyse the as-is (usually sourced outside the organisation) and (2) people who know the as-is and the company in depth and will make sure the future solutions will fit and will be implemented (usually sourced inside the organisation).

As a rule of thumb, any program should have a life span no longer than 1 year, as running it longer, results in people leaving the program, new people joining, and usually there are few results produced in the first year and there is a higher overall risk for the program.

Solid Global design with cluster implementation approach can be more flexible, less risky and offers faster results (e.g. Design 6 months, first country implementation in 3months, next countries rollout 1 per month).

What is the role of the Organisation Design in an O2C project? 

While reviewing the processes and designing the new technology is most often the focus of O2C projects, it is the processes, technology, and the organisation dimension that offers most opportunity for growth.

Supply chain, Sales, and Finance are the main functions involved. It is common to have different teams sitting in different locations trying to reach independent and sometimes competing objectives (e.g. increasing the level of service VS decreasing total cost to serve).

Including Organisation Design as part of the scope of O2C projects, allows among other things to:

• Break the silos between functions and physical locations
• Create the “One Team” mentality
• Review and update the Governance and the decision-making processes in order to have a lean O2C communication flow
• Optimise the span of control creating teams that are both efficient and effective in running the business (not too big, nor too small, with the right roles and skills in each team)

What would you like to achieve by attending the Automation, Robotics and AI for O2C conference? 

I am looking forward to network with like-mindeded people working on O2C processes in different industries. Every time I meet a new person, it is a great opportunity to try and help by sharing my knowledge, challenge each other, and learn something new.