How to use predictive analytics for better marketing electricity and magnetism connect to form

#

We do this by sending well-timed content, by personalizing some of that content, by enticing them to take tiny steps toward our goal. These are often known as “micro conversions” – a white paper downloaded. An online calculator used. A demo scheduled.

Well, predictive analytics can let you outsource some of that work. By analyzing tens of thousands (even millions) of prospect actions, it can estimate when each individual prospect might be most likely to complete one of those little micro-conversions.

Don’t let that capability make you worry about losing your job – there’s still plenty of work for you to do. But just like it’s not a good use of your time to manually reformat typos in your mailing list (gmial.com to gmail.com, for instance), it’s not a great use of your time to run an assessment on each individual prospect as they move through the sales funnel.

Doing that for just 100 prospects might well take up your entire day. So we let the algorithms of predictive analytics do it. While you go make sure your team members are working well (for instance), and go make sure IT understands the needs of your new app, and…. You get the idea. While you go do the rest of your job.

If there’s predicted demand for the fuzzy slippers, but not enough inventory to cover the orders, the marketer has some options. They could increase the price of those slippers, thus making a higher margin on the inventory they do have. Or they could give their best customers the opportunity to buy those fuzzy slippers first.

That’s where the predictive analytics comes in. By being able to analyze hundreds, even thousands of attributes about your best customers, the predictive analytics system can create a profile more detailed than anything you, the human, would have time to define.

Remember – once the predictive analytics algorithm knows how to pick the audience, make the ads, and personalize them, it can zoom through that work at computer speed. The same speed it processes any other data. That’s way faster than the click… type… click… double-click pace we humans work at.

Want proof of how well this works? One Harley Davidson dealership increased its leads by 2,930% in three months thanks to predictive analytics. Half of those leads came from lookalike audiences that the dealership had never before considered reaching out to. But the AI knew just where to find them.

If that strikes you as unfair, I get it. Some of us are a little cool on this approach too. But marketers have actually been doing this for at least a decade; they were just doing it at a more simplistic level. Catalog companies used to print different prices for people in different zip codes. More recently, airlines and travel websites have perfected the technique.

Of course, this cuts into the margins the company makes. But if they’re still doing well enough at even the lower price, it’s a win. They’re also getting the benefit of making a sale. Once you’re a customer, they can market to you more accurately and successfully.

This tactic is similar to segmenting, except it’s more like segmenting 10.0. You’ll be segmenting your customers and prospects based on every data point you’ve got – well, you won’t be doing that, the predictive analytics algorithm will do that.

When human marketers create personas, we tend to have to stick to 3-5 key personas. It’s just too much work and time to create a persona for every tiny little instance. We do our best, of course, but at some point, you have to go home to sleep and you have to address other demands of your job.

But compared to what an AI-driven predictive analytics program can do, this is child’s play. The AI can crunch every element of data – terabytes and petabytes of it – to find “clusters” of different persona types. It will see similarities among customers and prospects that humans wouldn’t see unless we had way more time and focus than we do.

Predictive analytics and AI are just better tools than spreadsheets and even good CRMs and content management systems. Think of those old systems like a shovel, or maybe even a spade. AI and predictive analytics are more like backhoes and mining equipment.