Making the decision that someone is “the one” is important to get right. If you’re going to invest deeply in a relationship, you want to be sure that it pays off in the end. After all, a long-term relationship should benefit both parties involved: the marketer and the customer.
(Please forgive me if you thought we were talking about personal relationships–that’s a topic for a different blog than this one!)
In terms of generating revenue, it’s often a better investment of time (and perhaps money) for a marketer to figure out how to turn a casual customer into a high value customer than to try to find and acquire more new customers. There are a number of ways to tackle this challenge:
- Turn one-item purchasers into multi-item purchasers
- Turn trials into subscriptions and subscribers into renewers
- Turn one-time purchasers into loyal customers
The list of specific approaches is quite long and varies across business models, but the real challenge isn’t so necessarily coming up with an effective tactic. Most often, it’s more difficult to identifying which subset of casual customers is most likely to become a high-value customer.
With Lytics Predictive Audiences, you can do just that, and avoid outcomes like over-investing resources in an audience that’s unlikely to provide value in return.
How predictive modeling works
At its heart, when you use lookalike modeling, you’re taking a group of customers that you want more of and using them to identify new individuals who resemble them in certain key attributes.
The first step–identifying the customers you want to create more of–is really a task that has to align with your larger organizational goals. It’s a strategic choice you’re making. You need to understand what goal you’re trying to achieve. Increase customer lifetime value? Increase brand engagement? Increase average transaction value? You want to choose your customer segment to match this goal.
When making a lookalike model, you have to balance the goals of reaching the most possible people and ensuring that they’re similar enough to your target that your model has predictive value. In Lytics, that means balancing two variables–reach and accuracy–in a way that aligns with your organizational goals.
Reach and accuracy are somewhat, if not precisely, inversely related. If your model is very accurate, you’ll have a smaller new audience. If you want to maximize the size of your audience, some members will only bear a passing resemblance to your target.
If you have to invest a lot of money and effort to achieve your goal (say if you’re selling an expensive good or service), you might want to emphasize accuracy. And if you’re looking to boost volume sales of a lower cost product, then you’ll emphasize reach.
Then Lytics AI/ML engine can get to work identifying the key shared attributes in the group. An example of what those attributes are might very well include both behavioral features and content affinity.
Lytics will find the subset of your source audience that matches these attributes, at which point the creative element of marketing comes into play. You need to decide what sort of experience, offer, or message you want to delivery to induce the new audience to take the action you want. Maybe it’s a special subscription or renewal offer, a product recommendation, or something else entirely–but you can push it out through any and all of your marketing channels.
Using Lytics Predictive Audiences in your business
Part of the beauty of lookalike models is their simplicity. You identify your goal and which of your customers’ actions contribute to that goal. You decide what pool of new customers or prospects you want to search to meet that need. Once Lytics Predictive Audiences does its work, you apply your creativity to determine what actions you can take to make it happen.
If you want to find out more about how Predictive Audiences, request a demo.