The goal of lead scores – the rankings of customer prospects based on the likelihood to purchase – is to make marketing and sales processes more efficient by signaling a potential buyer’s purchase intention. When the prospect is ranked from 1 to 100, it makes it easier for a marketing team – and a sales team – to prioritize their campaigns.
A lead score can also lower the cost to convert a new client or purchase by decreasing the average amount of time sales personnel take to complete a sale. This efficiency can also translate into better ROI on marketing spend.
A client in the consumer retail space was actively trying to understand how to acquire new customers. They purchased a list of leads from a recognized list provider and wanted to understand how best to use this list in their marketing program. Because there is a lead activation cost associated with each lead, the marketing and sales team wanted to make sure that they only activate those leads most likely to convert. They approached Wrench to provide a lead score tool to determine how to prioritize leads to achieve a cost-effective advantage.
Sales and marketing teams often develop short lists from lead sources based on intuitive guesses on what attributes are associated with individuals who are more likely to convert. However, there is a more scientific approach that takes into account more complex sets of attributes and systematically determines how different attributes predict a likely purchase – this also removes intuition and relies on data. This is where Wrench comes in.
Here’s where we’ll provide a look “under the hood” to shed light on the data-driven power of the Wrench lead score tool.
The initial step involved building a reusable algorithm to score how likely a lead contact will convert to a customer. Using machine learning and 3rd party integrations, Wrench built a client-customized training set from the lead list and known customers. The resulting model provided an ongoing method of scoring new leads and determining each lead’s likelihood of becoming a customer.
The predictors for a customized model can be as simple as basic demographics or related information provided by most list providers. However, in this case, the client chose to add additional elements using Wrench’s match score capabilities. This feature uses AI deep learning modules to assess how well a lead contact aligns with the specific brand or product attributes being positioned by the client’s marketing team. These brand and product assessments are individual scores for each lead contact and can be used separately or in conjunction with a lead score modeling.
An additional scoring feature offered by Wrench is capturing the personality traits of lead contacts. While this feature is an active ingredient in Wrench’s Persona product, the standalone scoring of personality traits offers an additional predictor for determining customer potential, which was used in this case, to make the list of “likely to convert” leads more robust. Moreover, understanding the likely personality traits of customers, the client could develop content with specific persuasion angles that were more likely to prompt the recipient to act.
In this case study, Wrench’s lead score model produced two important features. First, it generated the relative contribution of all the predictors used in the model. Second, the actual predictive power of the model could be assessed prior to implementation, giving decision-makers the ability to evaluate the expected cost-benefit. In other words, the client was able to vet the list of customers, their lead scores, and their personality dimensions to determine if they were on the right track before implementing a marketing program.
The model uses the brand and product match scores as important predictors of likelihood of becoming a customer. In this case, the predictive power of conventional characteristics like demographics, location, profession or social connection contributed only 18% to the predictive power of the lead score model. Conversely, brand and product related scores measuring lead contact affinity with brand characteristics contributed 60% the predictive power. Personality traits rounded out the contribution with an increment of 22%.
The CPG case study the lead score model, using additional Wrench components in model building, consistently showed marked improvements in sales and marketing metrics when high scoring leads were compared to non-scored leads. In fact, the client achieved a 100% increase in average order size, from $25 per order to $50 per order.
Lead scores drive improved sales and marketing metrics and increase sales conversions by ranking customers based on their likelihood to purchase. Wrench’s proprietary technology enables sales and marketing teams to more efficiently target customer prospects, reach prospects more likely to purchase and improve sales and marketing results. By leveraging a variety of data, Wrench’s AI algorithms save time and marketing costs and improve both sales and marketing results.