Usually, I’m reluctant to cover reports that point to solutions offered by the report’s sponsor.
That’s the case with a new report, out today, from Forrester Consulting. Called “How Predictive Marketing Analytics Boost B2B Performance,” it was commissioned by predictive analytics firm EverString.
But this report offers some findings that could answer a key question about predictive analytics: How do you know if it works?
When one asks predictive analytics vendors, they talk about A/B testing, or comparing the business before and after using predictive tech. But that’s not always a very feasible solution.
For instance, if your company’s web site uses predictive analytics to predict which product I should be shown next — because my profile indicates I should be interested on this January day in a pair of warm socks, for instance — how does the site know the prediction is accurate?
I might buy the offered socks, and the site might see that its overall sales have increased since it started using predictive technology. But if I had been offered, say, a baseball bat, I might have bought that instead. The site can’t easily do an A/B test on me, for the same point in a given session. It can do that for other customers, but those are other customers.
And a rise in overall sales could have resulted from other factors, such as better navigation, lower prices, or whatever. In fact, other vendors focused on machine learning-based business intelligence make their living trying to exactly pinpoint the various — sometimes hidden — factors behind sales rises and declines.
Similarly with predictive lead scoring. In it, predictive technology generates a score for each prospect, individual customer or account that is supposed to indicate if they are more likely to become a new or more loyal customers.
“Better Business Results”
If you focus on a few high-scoring customers, and they turn out to be good customers, is it because you gave them extra attention? And when predictive lead scoring attempts to predict lifetime value, how do you filter out other contributing factors? A lifetime is a long time to track results and account for varying conditions.
That’s the problem with predictions: The future is built on so many things.
So a new report that attempts to look at the big picture could help. Forrester says the report is based on a survey of 150 B2B marketing executives whose companies have at least 5,000 leads. A fairly small sample size, but potentially large enough to point in a direction. Forty-nine percent of the respondents said their company uses predictive analytics.
The big takeway: “Predictive marketing analytics use correlates with better business results and metrics.”
That is, compared with those in the survey who do not use predictive analytics (which it calls Retrospective Marketers).
“Predictive Marketers,” the report notes, “are 2.9x more likely to report revenue growth at rates higher than the industry average.” They are also 2.1 times more likely to “occupy a commanding leadership position in the product/service markets they serve” and 1.8 times more likely to “consistently exceed goals when measuring the value their marketing organizations contribute to the business,” compared to the Retrospective Marketers in the survey.
Forrester analyst Laura Ramos, who was involved in the report, told me the main point is clear: “Predictive analytics pays off.”
“If you use it right, you can outdistance the competition,” she added. This doesn’t answer all the questions about validating predictive analytics, but it points to the bottom line.
As for the genesis of this report, Forrester explains:
“In October 2015, EverString commissioned Forrester Consulting to evaluate the adoption of predictive marketing analytics at B2B companies. To further explore this trend, Forrester developed a hypothesis to test the assertion that predictive marketing analytics enables B2B marketers to be significantly more effective in identifying and engaging buyers in specific accounts and across the buyer’s lifetime.”
Ramos said that if the hypothesis had been proven false, there would have been no report. She also said respondents were located independently of EverString, and the survey was conducted without mention of the sponsor.
Comments