close-icon
Subscribe to learn more about this topic
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Email Targeting

Email only users who are ready to buy

datarevenue-icon
by
Markus Schmitt

Better than trigger rules

Until a couple of years ago, the state of the art in selecting users for an email campaign was behaviour-based segmentation (or trigger-based targeting): You analyse the user journey and identify certain micro-behaviours (triggers) that tell you that this customer is probably now interested in the email you want to send.

This study by MarketingSherpa shows how behaviour-based targeting is the best option so far:

stats graph

Using AI to find out when a user is most interested in a product is a very large improvement on even this technique. At ImmobilienScout24, we found that revenue from emails increases on average 245% with proper AI-based behaviour targeting.

No more guessing - AI gives objective answers

Segmenting users by manually finding rules is often a bit of guesswork. You ask the product manager which factors he thinks identify the right users, then you do some ad hoc data analysis with the engineering team, and then you build your rules. Maybe you test a couple of different setups with an A/B test. But in the end, you’re never quite sure whether you have the right rules. AI turns this process upside down: With few or no assumptions, the algorithms searches all the possibilities and finds the exact rules that work best. It’s like running hundreds of thousands of A/B tests, but in one hour, for a few cents.

Simple continuous improvements

Your users’ behaviour changes - so do your products. Traditionally you’d have to adjust your manual rules continuously, a costly and complicated task. With AI-based targeting, you can always rerun the algorithm on newer data, updating its rules and learning. This leads to simple continuous updates. Same goes for new data (e.g., you just integrate your CRM data into your data warehouse): Just retrain the algorithm with the new data source.

How does it work?

You train an algorithm to learn how to connect behaviour to actions: What behaviour on your website means someone will purchase soon? Once the algorithm can predict the likelihood of a purchase, you apply it to every single user daily and select users for your email campaign based on their likelihood to purchase.

Get Notified of New Articles

Leave your email to get our weekly newsletter.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Keep reading

No items found.
No blog posts found.