Sales Anomaly Detection

Background:

The (Swedish) sales force is in a situation were they have the sales responsibility for more customers than each one of them can manage according to our sales process. There are currently several initiatives ongoing to clean up this situation by a more structured way to categorize and prioritize which customer each sales person is to work with. The customers who wont be categorized as top priorities will be categorized as part of “Kundbanken” - a group of customers who are still important to Ahlsell and can generate a great deal of revenue but not eligible for a high touch treatment. In order to be able to monitor the sales progress of the customers in Kundbanken new tools are needed. The PO area for Sales (Analytics) has worked on a tool for sales anomaly detection and ran a pilot in Region 86 utilizing a machine learning model to detect when customers purchasing behavior is deviating.

Why is it important?

If Ahlsell Sweden are to continue its growth and maintaining the high quality service to its customers it is vital to understand our customers behavior even when we can’t have physical meetings on a regular basis. Being able to step in and being a speaking partner to our clients when we see sales falling or rising is vital in order to understand where in the life cycle our customers are.

The aim of this discovery?

Based on the back of the pilot done, create a roadmap for moving this from a pilot phase into a integrated part of A-Sales.

Stakeholders:

Christian Morton Fincham - Head of Sales Sweden

Marcus Doverhjelm - Head of Sales Support Sweden

Magnus Salmi - Head of Region Sydost