In the year since we launched Intellimize, we’ve had the privilege of working with innovative marketers like Chime, Stella & Dot, LendingTree, Segment and Perkville. They’ve become active users of our predictive personalization system that tailors their web sites for each visitor automatically using machine learning. We’ve seen how they’ve put predictive personalization to use day-to-day, and we want to share some of what we’ve learned.
One thing that surprised us was how quickly marketers embraced the differences between predictive personalization and A/B testing approaches. There was pent-up demand within our clients to try many new ideas in order to drive more conversions. Our customers were able to meet that demand testing, on average, 44 different experiences containing 127 variations across 5 different pages on their mobile and desktop websites. Because predictive personalization allows marketers to run many individual tests simultaneously, the average campaign running on Intellimize optimizes across 94 million possible versions of the page. This means our clients are effectively personalizing down to the individual level.
What does this mean for marketing going forward?
The ability to test and learn faster reduces both cost and risk, allowing marketers to experiment with more ideas and invest more effort in developing the ideas that prove effective. Had our customers used A/B testing, it would have taken, on average, 25 years for them to do what they’ve done with us to date.
One of our clients, Chime, an innovative banking app, uses Intellimize to increase new account signups from their homepage. In the first six weeks of using Intellimize, the digital marketing team at Chime tested nine different headlines, one new video, and three variations of body content. The result? 8% lift.
After analyzing the initial results, the team added new photography, headlines, and body content to their campaign. Over the next four weeks, they saw a 79% lift in new account signups. Overall, they tested 21 different ideas, which enabled 216 versions of their page, in ten weeks. Testing these ideas with A/B testing would have taken Chime more than nine years.
Thought leading marketers are using this increase in speed to test more ideas and deliver more insights, which together then deliver better results.
Predictive personalization automates most rote experiment management tasks, giving marketers more time to focus on understanding their prospects and creating personalized experiences that drive customers to take action.
Perkville, a referral and rewards program that helps companies drive revenue and loyalty, uses Intellimize to increase customer referrals for its clients’ businesses. The marketing team personalized calls to action, social elements, and the layout of their referral pages. They did no manual management of variations, test matrices, or audience cells, and were able to test 62 ideas that resulted in hundreds of thousands of possible versions of their referral pages. The result? 70% increase in referrals.
We’ve seen how the combination of faster testing, learning, and iterating with less rote work helps marketers drive impressive results. It’s an exciting time to be a marketer as machine learning tools like Intellimize are enabling us to focus more on understanding and delighting prospective customers. It’s been an amazing first year, and we’re looking forward to seeing the amazing and creative ideas our customers come up with in year two!