B2B Prospect Scores – Improving Marketing ROI in the Information Technology Vertical

We build a lot of B2B prospect scores for one of our clients.  One customer of this client recently tested a prospect score we built on their behalf.

Their goal was to determine how much “lift” the prospect score provided to their outreach campaign for their information technology networking solution.  That is, they wanted to know if the score led to (and by how much) the identification of more, higher likelihood prospects than using no score at all.

Their very thoughtful test yielded a result that surpassed even our, admittedly biased, expectations.

What are B2B prospect scores?

A prospect score is a number assigned to a list of non-customers to be contacted in an outreach campaign.  When this list is sorted by this score, the non-customers who are the “best” prospects rise to the top.

That is, if the prospect score works as expected.

These “best” or “high potential” prospects typically resemble current customers.   Converting these into actual customers is the job of the sales team.  The prospect score just tells them who they should focus on first.

What is lift?

“Lift” is marketing-speak for how much a B2B prospect score sorts a list of non-customers, from highest to lowest potential, compared to some other score, approach or no method at all.  Lift is presented as a percentage, such as 133%.

Consider a list of companies, some are customers, most are not.  Now sort this list by the prospect score, from highest to lowest potential.  And break the list into deciles.  A chart like this can then show the lift attributable to a prospect score built using this list.

This chart shows that over the first 30% of the sorted list (ordered by the prospect score), 70% of the actual customers can be identified (blue line).  This yields a lift attributable to the score, compared to using no score (a random selection in this case indicated by the red line), of 133% (i.e. [(70-30)/30]*100).

Of course, the lift that will actually occur in an outreach campaign targeting only non-customers will be different (and most likely lower).  But a prospect score that performs well in testing should also perform well in practice.

The size of the lift depends not only on how well crafted the score is but also on which alternative method the score is compared.  The lift attributable to one score vs. another might be very small.  However, compared to using no method for ranking non-customers, the lift could be quite large.

Why is lift important?

B2B prospect scores which yield appreciable lift reduce the cost of outreach campaigns.

Continuing with the above example, in order to identify 70% of actual customers without using a score, 70% of the marketing list would need to be used.  But by using the score to sort the list, these same 70% can be identified by using only 30% of the list.

This yields a cost reduction of 57% as the remaining 40% of the list does not need to be contacted (i.e. [(70-30)/70]*100).

So how do we build prospect scores?

Here at KDD Analytics, we build B2B prospect scores using statistical models.

We build these models using a sample of customers coupled with a randomly-selected sample of non-customers.  The goal is to arrive at a model that maximizes the differences (“discriminates”)  between customers and non-customers.

These differences are based on a set of “explanatory” characteristics.  In B2B, information technology marketing, these are foremost firmagraphics (such as number of employees, industry sector, enterprise revenue, etc.).  But they also include measures of information technology used at the business (such as spending on computers and software, likelihood of a certain vendor presence, number of PCs, etc).

Back to the test…

The customer of our client divided a prospecting list into two samples.  One sample used our prospect score, the other did not.

Then they initiated an actual calling campaign for their networking solution using both lists.  When the two campaigns were finished, our client’s customer compared the results.

In general, the degree of lift depends on the definition of a “successful” call.  For this test, a successful call was one in which the prospect expressed a genuine interest (follow-on call scheduled) and/or requested a quote.

The results of this test were a staggering 300%+ lift from using the KDD Analytics prospect score.

Granted this is a very high lift and we have to assume that the customer calculated lift correctly.  But it is consistent with our experience in B2B, information technology marketing.

B2B prospect scores can measurably improve outreach campaign performance.

Needless to say our client’s customer was extremely impressed with these results.  So much so that they immediately renewed their contract with our client for additional data sales and analytical services.

Of course, your mileage may vary.  But using B2B prospect scores to target potential new customers can measurably improve your marketing ROI.