Converting Sales Leads with Product Predictions

A telco company was looking to increase sales of a product range in one of their geographical regions. They used a standard sales funnel process to convert priority leads into new customer contracts, and relied on a lead targeting tool to identify and prioritise potential customers.

Having already made a significant investment in this tool, they were curious about how technologies such as A.I. and process automation could be used to enhance it, and achieve their overall objective. Their ideal solution would be capable of providing lead recommendations with the highest propensity for conversion.

The goal was to achieve at least a 30% increase in qualified leads for the top 3 product recommendations provided by the algorithm, and for at least 50% of these recommendations to be for non-core product offerings.

We built a solution capable of identifying patterns in historical customer data, and making accurate predictions on which products would be most appropriate to which customers. The system considers how variables such as business size, industry, and geographic location, influence conversion rates for particular products. By revealing pairings with a higher propensity to sell, the client’s sales team is able to nurture more appropriate leads, and achieve greater qualification results.

After mapping the existing end-to-end sales process, we were able to discover how A.I. enabled process automation could enhance it and accurately estimate the ROI.

We then created a solution that would collate and process the leads and sales data in real-time, and measured its performance against a specified success criteria to validate its business value.

Next, we fine-tuned the solution to improve prediction accuracy and overall performance of the solution before deploying it into production.

Now deployed in production, our A.I. enabled process automation solution meets the parameters set out at the start of the project, achieving a current accuracy level of 29% and growing. This is just under a 5% uplift on previous activity. In addition we delivered 54% of the product recommendations from long-tail, high margin, sticky products.