(See the video version at the bottom of this post.)
Today marks the 50th episode of my “Something I Learned Yesterday” podcast, so … I guess I’ve learned 50 things. And if you’ve been coming along for the ride, so have you. So congrats.
Bo Sacks distributed an article by Matt Turner titled “A top CMO reveals how AI is going to change sales and marketing across every company within 12 months.”
It paints a nice picture of marketers using AI to create individualized ad campaigns, “where each potential customer receives a different message at a different time that’s tailored to their needs and behaviors.”
It’s time to start planning for that now, because AI isn’t magic. You can’t just say, “hey, AI, make some individualized messages for these people.”
“Based on what?” the robot will reply.
Start thinking now about what kind of data you’ll want to have in place to power these lovely AI customizations.
And this doesn’t only apply to marketing. It applies to content creation as well.
You’ll want a few different kinds of data.
The simplest is behavioral data on your website. How people found your site, what they did when they got there, what topics they read about, whether they commented on articles, and so forth.
You’ll also want to track how they interact with your brand in email and on social media.
You might consider collecting what some people call “zero party data” through polls, questions, surveys, and so on.
If it’s possible, you might use 3rd-party sources to get additional information about your market through look-up services.
All this data will have to be stored in a central repository where it’s all tied to the individual, so you can link it all together. That’s what customer data platforms do.
But you have to start with some use cases. What do you want to collect, and what do you want to do with the data you collect?
So let’s walk through an example. I’m a homebrewer, so let’s say I owned a website that sells homebrewing supplies and equipment. There are all sorts of things I would want to know about my customers.
Is he a hobbyist or a professional? (I’m saying “he” just because it’s easier, not because homebrewers are all men.)
Does he do extract brewing or all-grain brewing?
Does he have the equipment to do lagering, or does he stick with ales?
How often does he brew? Does he brew with friends? Is he a member of a club? Etc.
But more important than that, I need to think about what I’m going to do with information.
Someone who’s in a homebrew club is probably going to be a better customer than someone who isn’t, so I could give discounts to club members. Then I could start to form relationships with the clubs, find out where they operate geographically, and even promote club membership to customers based on their location.
There are a hundred things to do, but my point today is that if you want to make whizbang customized AI wonders for your market, you need to start thinking about that now, and start collecting that data now, so when you get the technology, you’ll have the data you need to create the customizations you want.
By the way, if you need help with this sort of planning – coming up with your own use cases for your market – that’s what I’m here for. Give me a call.
Links
