According to SiriusDecisions, 68% of companies use marketing automation systems to do lead scoring, yet only 40% of sales people are getting value from it. When marketers are sitting on mountains of data, how could that be?The challenge with traditional scoring found in marketing automation is that it wasn’t built with a predictive-first approach. Point values are manually assigned and there isn’t any emphasis on the combination, concentration, or recency of signals. Points are just additive and the point differences between activities (e.g., downloading a whitepaper or visiting a pricing page) are essentially arbitrary. While it’s a starting point, it is easy to see why it is prone to false positives and why it is not trusted by reps.
What we have seen from Infer’s customers is that when your sales and marketing teams know exactly which prospects to invest their time and energy into, they gain a distinct competitive advantage. They can operate more efficiently and capture more of the revenue opportunity.
It would seem that amongst all the data being collected the answer is there, but it is actually quite a hard problem to solve. It requires connecting different data sources and analyzing tens of thousands of disparate activities. Thanks to the tribal wisdom we’ve built up at Infer over the years, we’ve figured out tricks for getting data out of automation systems (e.g. Salesforce, Eloqua and Marketo), matching records, and distilling insight by using data science and machine learning. We’re confident that we can extract and analyze more behavioral data than any other solution out there, and do it better.
For one customer we’re analyzing approximately two million activities per week and updating behavior scores for hundreds of thousands of their leads every night. By using our advanced modeling techniques, we can identify which 10% of the company’s leads are 6.1x more likely to buy in the next two weeks. And in a similar vein, our model captures 92% of the sales team’s won amount in just the top 28% of its leads. An effective behavioral scoring solution can capture insights like this and should be sufficiently robust to work even without demographic data.
Time and time again, our models have out-predicted competitors in real-world settings. That’s because we’re taking a very scientific approach with both our behavior and fit scoring, which makes our predictive models significantly more accurate than anything else.
If you have questions about behavioral scoring, contact us and we’d be happy to help out.