Need-based customer segmentation helps businesses gain a more detailed understanding of how customers interact, buy, and consume products and services. This project allowed our client to offer the most relevant propositions in a personalized way, according to each market’s specific dynamics.
We analyzed our client’s historic sales for each customer on an invoice level. This enabled us to create data-driven profiles of each end customer based on the frequency, recency, monetary value, progression over time, and potential. The profiles were used to segment customers by their needs and assign them to respective service models, which helped in preparing tailored lists for each sales rep.
Need-based customer segmentation helps businesses gain a more detailed understanding of how customers interact, buy, and consume products and services. This project allowed our client to offer the most relevant propositions in a personalized way, according to each market’s specific dynamics.
We analyzed our client’s historic sales for each customer on an invoice level. This enabled us to create data-driven profiles of each end customer based on the frequency, recency, monetary value, progression over time, and potential. The profiles were used to segment customers by their needs and assign them to respective service models, which helped in preparing tailored lists for each sales rep.
We also partnered with the client’s marketing team to assist in bringing the customer segments to life through building customer personas. This enabled sales teams to better relate to the needs of each segment and service them accordingly.
As a by-product of the project, our clients also tailored their service models, creating a new “mass” model, where low-value, low-interaction customers were serviced by inside sales and marketing instead of dedicated sales reps, allowing for better focus on larger customers.
The approach
Need-based segmentation is a complex project that consists of various stages:
- Setting Business Questions
Together with the client, we start by listing and grouping all of the business questions and business initiatives that the client would like to address in their usage-based customer segmentation.
- Selecting Variables
We create an extensive list of datasets/variables to include in the segmentation. Variable selection is driven by the business questions, data availability, and data science techniques. At this stage, more is better.
- Iterating & calibrating
A strong, interpretable, and consensus-driven customer segmentation is only created through a combination of iteration and quick precise calibrations. The iterations are needed to cut down the list of variables to a manageable and easy to interpret list, as well as to get to a practical number of segments.
- Describing segments (visual and data-driven)
A segmentation that wider business users cannot understand and adopt is a bad segmentation. At this stage, the key is to communicate the segments in a relatable way and align them with other internal initiatives and data sources, so that the business owners can validate them.
- Segmentation/Maintenance
For the B2B environment, we offer periodic maintenance. Usually, in B2B space, new customers do not necessarily flock in on a daily basis. This is why periodic maintenance that is done offline is the optimal choice.
For the B2C environment, we implement and automate continuous segmentation, due to constantly changing user base (e.g. Telecoms). It requires algorithm deployment and ensuring proper model functioning.
Need-based segmentation lifecycle
- Build RFM (Recency, Frequency, Monetary value) profile per customer based on available data.
- Find the optimal number of segments and allocate customers.
- Calculate additional metrics for the Persona Building Process. Metrics include frequency of purchases per quarter, the average number of purchases per month, % of shopping bag value increase/decrease, product split ratio, etc.
- Prepare Questions for Persona Building Process. Prepare two types of questions – data-driven (source – data) that focus on purchasing behavior and patterns identification and soft questions (source – business expertise) that focus on involvement and other marketing initiatives.
- Draft Personas per Each Segment. Based on data metrics and business knowledge, prepare draft Personas for a business workshop.
- Verify Personas. Do the verification both on a high level and customer by customer level and make sure data points per Persona make sense from the business perspective.
- Map ‘Personas’ to the service model. Assign each Persona to matching service model components corresponding to its needs.
- Ongoing Adjustment. Revise the descriptions regularly, approximately once a year.
Benefits for business
Need-based data-driven customer segmentation results in customer segments identified through a machine learning technique and logically mapped to respective service model corresponding to customers’ purchasing behavior and overall needs.
The results below represent a practical example of output for one market. The treemap chart illustrates the revenue split among the different customer segments. As per the below example, Champions, which are 17% of all customers, contribute to 79% of the total revenue, having the highest spending power among all customers. Champions also place orders five times more often than Loyalists do, with an average purchase size that is almost two times higher than Loyalists. Insights like these help the client to interpret the results through business logic and act on them effectively together with the sales teams.
The action plan coming out of the segmentation comprises of specific targets covering not only lagging but also leading indicators of sales, drawn from the strategic goal of what each customer has the potential to become. This potential is being nurtured with need-based service models covering different levels per each customer segment to optimally convert as many clients as possible to Champions and keep the Champions as satisfied and happy customers.
Business impact
