Marketing Analytics

In gearing up for 2016, I did some research on Marketing Analytics trends for 2016 and found a great article from MarketingProfs.com.  The article’s first point on the use of predictive and prescriptive analytics ties into another area I’ve been looking into, lead scoring.

In the article, the author, Joshua Reynolds, says “Predictive analytics is an incomplete approach because it only gives you a likely outcome if nothing changes. It doesn’t tell you why outcomes are likely, the correlations driving those outcomes, or—perhaps most importantly—how to change those outcomes.”

He continues “Prescriptive analytics is something of a step beyond predictive analytics in that it tells people not only where they’re headed but also the moves they can make to improve the outcome. However, that is still basically a black box approach; most prescriptive platforms don’t let people understand why the platform made certain recommendations.”

This, to me, defines lead scoring.  Lead scoring is predictive in that the score is supposed to predict when the lead will purchase.  However, it doesn’t take into account if the variables change (decision maker leaves, they buy an alternative, etc.).

Lead scoring is also Prescriptive because it attempts to direct next steps based on the score.  If, for example, your lead scoring system ranks on a 0-100 scale, 0 being the worst, a lead with a score of 40 might have the lead scoring system tell you the best path to take is to follow up with some generic marketing email and then drop the lead because their score isn’t high enough.

Reynolds also mentions that Prescriptive analytics is often viewed as a “black box approach” meaning the users of the information don’t understand how that information was created.  How does this apply to lead scoring?  Marketo has a 54 page “Guide” to lead scoring.  They also have a 2 page “cheat sheet” that includes a glossary of terms (BANT anyone?) https://www.marketo.com/cheat-sheets/lead-scoring/.

My guess is anything that requires 54 pages of explanation and then the 54 pages requires a dense 2 page cheat sheet isn’t being understood by the majority of users.

Reynolds then introduces a concept that is incredibly relevant to lead scoring and another project I’ve been involved with, INgageHub.  Reynolds says:

“Enter explanatory analytics. Marketers are smart people, and as such they (not computer code) should be asking the analytics questions, driven by their curiosity and intuition. They should then bring in the machine to investigate the correlations that matter.”

What is more effective than generating a lead score is to give your people the ability to ask your target audience questions and then engage with them to find out, directly from the source, their level of interest.  Lead scoring becomes very simple, they either tell you directly that they are not interested or they don’t engage, they score a 0.  If they answer the short question set, they score 100 – a lead that should be forwarded to sales.  INgageHub gives companies the ability to do this.

In addition, like Reynolds says, as the data from your audience starts to grow, then you have the ability to use machines to unlock the underlying information in your data.  The most powerful aspect of this data is that is comes directly from your audience.

Smart questions and powerful data can now replace antiquated lead scoring techniques.