B2B growth with advanced lead scoring
Context and objectives
A European B2B food and drinks distributor wished to gain a deeper understanding of their client base and to increase the potential number of leads.
They were facing two main challenges
poor data actionability for their customer base: they mostly had consumption behavior data and few external data manually filled in by the sales team.
an overwhelming number of prospects to accurately qualify from a list provided by (expensive) B2B data services.
Approach
The first step of the project involved gathering internal data about the behavior of the customers, such as their consumption patterns, such as the type of products and services, or vendor machine data and the purchase frequency.
We then collected relevant B2B open data (financial & sectorial) and created extra variables linked to company size, EBIT, and other relevant factors. This process was crucial in providing a more comprehensive view of the clients database.
Next, we developed an algorithm designed to differentiate between “high” and “low” potential clients. This was a critical phase, as it allowed the company to understand their customers better and identify potential leads. We faced a challenge here as the criteria for defining a high/low client were initially restrictive. However, after some refinements, the algorithm was able to clearly differentiate between the clients.
Results
We managed to enrich over 50% of the CRM entire database and more than 80% of the clients who made a purchase in a last 3 years with relevant information. The enrichment revealed the ideal sectors and firmographic types linked to higher commercial performance. This insight was important as it helped them target their marketing and sales efforts more effectively.
The lead scoring algorithm was also a success. It not only provided the client with a deeper understanding of the type of companies they were working with, but it also identified 1,800 companies with the highest potential. This was a manageable number for the organisation, enabling them to focus their resources effectively.
In terms of deliverables, the client received
Two datasets. The first dataset was their customer base, now enriched with relevant data. The second was a list of target companies, each scored based on the likelihood of being a profitable client and their respective financial & sectorial data.
A trained model
Project code
Comprehensive documentation
A Power BI dashboard
