How to Create Personalization Business Rules

Lately I’ve been exploring what defines personalization, and how low-tech personalization can succeed (no machine learning required!). I’ve also learned a lot from my colleague Kevin Nichols. And I’ve realized something. All personalization requires a low-tech approach. All personalization has personalization business rules.

Personalization business rules are the key to answering the question: what content goes to which people when?

Let’s take a look at how personalization business rules work, and how to create them.

What are Personalization Business Rules

Typically, personalized content is supposed to show someone content related to actions they’ve taken, or things about them. We often focus on “how does the site/app/service know that you’ve purchased an item?” and “how do they know the weather where you are?” Less often do we consider “how does the site/app/service know to show you weather-related information?” And perhaps more importantly: how does it know whether to show you weather-related information, product-related information, or both?

For example, let’s say we have a retail website capable of tracking purchases, pages you’ve looked at, and your location. In theory, this means the website could offer you umbrellas when it rains in your city, shoes to match your recent pants purchase, and a reminder to look again at the watch you were admiring. But the website only has one area dedicated to personalized content. What should you see?

This is where business rules come into play.

How Do You Create Personalization Business Rules?

Here are a few steps to build out your own personalization business rules.

1. Begin by making a list of all the possible messages. What is every action you might suggest to a visitor? For example, “check out our umbrellas”. Get detailed.
2. Compare the list of messages to the content you have available. It’s great to suggest “read more about our policies,” but if you have no content on your policies then you either need to create it, or remove that action.
3. For each message, identify the prompt. The prompt is how you will know to say the message. This is the technical piece: will you see their location? Did they purchase something? Etc.
4. Next come the business rules. For each message, note if it should only appear under certain circumstances. For example, if there’s a message like “Did you like our watches?” you might want to note “only show if user viewed watches” and “remove after 3 days.”
5. The other part of the personalization business rules is the priority. After making the notes in step 4, prioritize every single message, and every single prompt. This way your system will know which message to offer, even if the visitor does something that results in multiple prompts.
6. Consider the
7. Lastly, consider: how are we going to measure success?

Next Steps

Personalization business rules aren’t easy. And a lot of the early work is guesswork. But with time and effort you can set up a system that any CMS can implement.

Machine learning is great, but even the best of technology requires these steps to know what to do. Let’s not forget that!