Tapping into the Power of Automated Personalization
This is the first post in a three-part series on Adobe Target’s automated personalization. In this post, I’ll discuss why automated personalization can help generate more incremental revenue than A/B testing and traditional rules-based targeting. The next posts will dive deeper into the personalization models and in-depth reporting.
With buyers coming from all directions and all devices—influenced along the way by friends, motivations, moods, and budgets—delivering a personalized experience with only rules-based targeting is no longer enough.
Playing by the Rules: When Does It Make Sense?
Traditional rules-based targeting involves clear-cut scenarios—you have to know who to target and which content variations to show or test. For example, you might want to display a specific promotion to all shoppers except those who are coming from an affiliate coupon site, or only to visitors who have entered a certain search term.
Rules-based targeting can be very effective when you know your high-value segments and are focused on optimizing their experience. First, you need to understand the opportunity for each segment you want to target: are there enough potential buyers to impact your bottom line? Then, you need to manually set up the rules themselves. Fortunately, for customers of the Adobe Marketing Cloud, sharing audiences from Adobe Analytics to Adobe Target is as simple as checking a box.
But with customers’ interests changing by the minute, it’s becoming more challenging to place them in categories and ensure their experience is updated along with them. And there are still many “unknown” customers who are entering sites anonymously. How can you target them with the most relevant content?
Although rules-based targeting has its place, it’s no match for the scale and dynamic personalization of today’s big data. That’s where automated personalization comes in. My colleague Jonas Dahl just published a great article on how it works. In this post, I’ll show how it offers unique benefits above and beyond rules-based targeting.
Automated Personalization with Adobe Target
Adobe Target’s automated personalization is a powerful machine-learning tool that sifts through Big Data, identifies patterns, and adapts to visitor behavior over time. Whereas rules are fixed and rigid, automated personalization is fluid and dynamic, seamlessly changing and growing to accommodate customers’ behaviors and experiences. It’s not a mysterious “black box”—it’s a smart, sophisticated tool that gleans insights and provides automatic audience discovery, revealing the top-most predictable variables driving conversion.
Gleaning Big Insights from Big Data
A typical campaign generates thousands of data points—browsing paths, search keywords, device type, purchase history, geolocations, product ratings, and much more. It would be impossible for any human to analyze all of that information, glean usable insights from it, and stay abreast of constantly changing trends. Automated personalization feeds the data into a personalization model, automatically chooses the best offer, and delivers a targeted experience in real time.
As Big Data gets even bigger, we’ve gone beyond just changing content based on A/B testing. Yes, you still need smart and creative people to craft campaigns, but machine-based tools are essential to engaging a broad, diverse customer base and manipulating the droves of data available to today’s marketers.
Targeting on a Massive Scale
Automated personalization is effective for any type of site, in any industry. For example, a popular financial services company uses Target’s self-learning models to serve up dynamic offers like mortgages, credit cards, and online bill pay—all based on each individual’s previous browsing paths, account status, search terms, and other factors. A new homeowner who has recently secured his first mortgage will have a very different online experience than a grandmother approaching retirement.
A key benefit of automated personalization is that you can target on a massive scale. With rules-based targeting, you’re limited in the number of segments you can test against. Segments and campaigns have to be set up manually, and the segments must be large enough to impact revenue. But with automated personalization, the model is powerful enough to process staggering amounts of data and build segments on the fly based on predictive traits, so you can achieve a deeper, more personal understanding of all visitors.
Already using Adobe Analytics for your reporting needs, or Adobe Audience Manager as your data management platform? Because Target is part of the Adobe Marketing Cloud, it integrates seamlessly with Analytics to include Web behavioral data, such as past browsing and purchasing behavior. The segments you’ve already identified in Analytics are automatically folded into the model.
The tool can also leverage third-party data, such as income status and recent offsite purchase categories. And for those who are using Adobe Audience Manager as a data management platform, these segments are preloaded into automated personalization. All preexisting data is built into the modeling system to influence real-time decisions about which content is displayed to each individual.
Easy to Use
Despite the complex calculations and decisioning going on behind the scenes, Target’s automated personalization is extremely user-friendly and intuitive for marketers. From selecting an audience to creating experiences in a guided workflow, the entire process is easy and intuitive—because, after all, that’s the goal of automation.
In essence, automated personalization reduces the load on marketers and the need for dedicated data scientists. Automated personalization campaigns are meant to run for extended periods of time, which means marketers no longer have to go through recurring implementations to formulate a testing hypothesis, analyze statistical significance, closely monitor the timing and conflicts of the campaign, and gather and act upon insights. With automated personalization, all of those time-consuming, expensive elements go away. The tool does the heavy lifting in real time. Instead of worrying about countless data points and considerations, marketers can hit the ground running. The tool tackles both optimization and targeting at the same time, leaving marketers free to focus on developing relevant, quality content.
Customers never stop changing—and our automated personalization models never stop learning. The end result is fully optimized content, personalized to meet the demands and preferences of its audience.
How Does It Work?
It’s clear that automated personalization works. According to the Adobe 2014 Digital Marketing Optimization Survey, automation increases conversion rates by up to 3.6 percent. But what’s going on under the hood of this self-learning machine? In my next post, I’ll take a deeper dive into the models used in Target’s automated personalization.