Offering the nicest price
Some of us aren’t going to like this tactic very much. It has a whiff of being sneaky, even kinda sleazy. But alas – it works.
Because the algorithms know so much about us and how we respond to ads and offers and products we’re searching for, they know that we respond to different price points. And so they can offer different prices to different customers.
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.
Here’s how it works: If you live in an ultra-high-income zip code, the price for a particular Christmas wreath might be $175. If you live in a lower income zip code, the wreath would be $125.
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.
For many companies, even if they lose a little bit on the first order, they’ve got a sophisticated enough marketing system to make up the loss later when you buy again.
Create far more refined customer personas
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.
So you pick the personas that make up the largest chunk of revenue, you build content and a buyer’s journey that best addresses their needs as you can, and you call it good enough.
And that is pretty darn good. It’s way better than just treating everyone the same, that’s for sure. And this level of segmentation and personas works – you’ll get 50-300% more results just by treating these groups differently.
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.
The AI can then address those clusters’ needs with content they’ll like most, via channels they prefer, at times when they are most likely to respond.
The result? Dramatically higher numbers of leads, better leads, and leads that move through the sales funnel faster.
We’ve barely scratched the surface of what predictive analytics can do for marketers. This post could easily be extended into a book.
But we have covered enough to show you what’s possible. And hopefully enough to show that predictive analytics aren’t out to steal your job.
Just think of AI and predictive analytics as computers 2.0. They’re a power tool to manage the mountain of data your business accrues every hour.
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.
Marketers get to play with the big toys now.