Monday, January 23, 2017

Inside Predictive Analytics: Or How I Knew You Were Going to Read This.


Web Retailers would love to somehow tell buyers from shoppers. With predictive analytics, they can.

Depending upon the site and the products and services offered, some website visitors will purchase a product, fill out a form for more info, or schedule a demo.  Most however, will take no action. 

Conventional website analytics can tell you where those visitors are coming from, what browser they’re using and other well, useless information for a digital marketer. Other software will tell you which Calls To Actions (CTAs) performed best, while some others (like heat maps) will tell you where their eyes focus when they peruse your site. All valuable information but still no way of telling buyers from the window shoppers.

Leave your tarot cards and crystal ball at home, predictive analytics uses data mining, statistics, modeling, machine learning, and/or artificial intelligence to analyze historical data to make calculated predictions about the future. It’s been around for years of course. Mortgage companies use it to decide how much to lend, doctors use it to determine the likelihood of developing certain diseases. Where it is new is in digital marketing. Those advertisers who do use it right now not only will have the winning hand, they will have the upper hand on their competitors.

First the basics: when you browse the Internet you leave a trail creating a map of where you’ve been, what you’ve seen, and what you clicked on from “cookies” or tags that sites use to identify you.  This creates a profile and your preferences.  The practice of serving you ads based upon your profile is called “remarketing.” Still not there at predictive analytics for digital marketing. but we’re getting close.

Remember all those people I mentioned who visited your website? Well some got there by clicking the wrong link, or liked your photo, or was doing research, or a school project, or a full deck of missteps and arbitrary visits. Others likely got there and changed their mind: not a fit. Some though, do want your product – just not right now due to budget schedules or inadequate revenues when it’s B2B. Or maybe it’s the wrong size, color, or flavor if it’s B2C?  Regardless, the timing isn’t right. 

Predictive analytics in digital marketing does a few things. First it analyzes your history against those of others, more importantly, others who purchased. Buyers have patterns so when they came, what the clicked on, how many visits they made before they purchased. It works from the inside out, painting a picture of all buyers and seeing how closely you align to the profile.  Amazon is a great example of B2C best practices. It shows me ads for vitamins, 29 days after I received my order of 30 capsules. And last week I purchased new pillows. Today Amazon showed me bed linens figuring I would need those as well. Well played, Amazon. Well played.

But what about B2B where leads come in through website visits and forms? You have all these prospects, leads, and a lead is a lead, right? But which ones are the buyers and who should sales call first – or at all? You can sometimes check out their company based on their emails even isolate IP addresses, but you still don’t know who to call first. Enter predictive analytics that acts as lead scoring taking into such things as the number of visits made, the total time spent on your site, which pages were visited, whether they downloaded an eBook, whether or done it all before, job title, their company size compared with current customer company size.  You are only limited by your data and/or the database of your digital marketing agency.

Predictive analytic practices are not without their critics. The Big Brother factor turns a lot of people off in this increasingly connected world wherein privacy is a commodity.  But the intent is rarely malicious and that’s why trusted vendors mean more than ever. Face it.  Advertising cannot easily be escaped either on or offline.  So rather than condemn retailers who strive to learn my preferences and buying habits, I commend them.  Show me ads for paraphernalia of my football team, serve me coupons to my favorite restaurants, and save the ads for medicines for ailments I don’t have, for chips I don’t eat, and video games I won’t play.  Learn me. Know me. Deal me in.

Frank Bocchino is marketing communications manager for Revana Digital, a Digital Marketing agency leveraging paid search conversion rate optimization, SEO, and industry-innovative predictive analytics and geo-targeting that result in sales.


No comments:

Post a Comment

What say you?