Adobe Target gets more personalized thanks to AI
- Summary:
- Personalization is a tough nut to crack. Today, Adobe announced enhancements to Adobe Target, including AI, intending to take personalization forward. Barb Mosher Zinck shares her views.
The slow evolution of personalization
Not every website is personalized. Whether it’s the lack of technology or a lack of strategy, many still don’t offer a personalized experience.
Those that do started out small. Capturing audience data and personalizing by location, or some other geo or demographic. Then some brands starting monitoring clickstreams and personalization went a little further because the system understood what you looked at last time, or how you moved through the website and it adjusted certain widgets (or containers) on a web page. It’s a manual approach to personalization, but it works.
Personalization has mostly been by audience or segment. People are grouped based on common characteristics. No one has nailed the one-to-one approach for a few reasons. First, it’s hard to do. There are few tools that support that level of personalization. Second, the data needs to be there, and most organizations still struggle to pull together all the right data. And third, it’s expensive.
Adobe tackles the one-to-one approach with AI
Adobe Target is Adobe’s testing and optimization solution. It stands alone, and it’s integrated with other products in the Marketing Cloud. It has been through a lot of changes since it was Omniture over ten years ago.
Personalization is where Adobe has done some significant work with Target. It’s time to shift the focus from purely segment-based personalization to more one-to-one personalization and Kevin Lindsay, Director of Product Marketing at Adobe for Target, walked me through the big updates that will change how Target customers personalize their customer experiences.
Lindsay talked about the reasons why organizations want to personalize: drive engagement, increase conversions, improve ratings and reviews, and customer retention. We all want some degree of personalized experience because, for us, it shows the brand is paying attention to what we are interested in. However, Lindsay quoted studies that say 70% of consumers think personalization today is superficial.
To change that perception, Adobe Target has a few key updates. The biggest one, available today to Adobe Premium customers is One-Click Personalization.
When you start to implement a personalization strategy, you have to take into consideration not only the consumer, but the device, the channel, and the different types of data you can use to personalize. It’s a lot to figure out at scale manually.
One-Click Personalization with Target takes the work out of the hands of the marketer and puts it into the lap of Adobe Sensei, Adobe’s AI technology. The sample screen shot shows the “auto target for personalization” option in action:
In this implementation, marketers create a set of experiences with one control and let Sensei determine which one works best for each customer based on what it knows and learns about the customer. The marketer isn’t defining segments or telling Target what to look for to show a particular variation. They aren’t inputting the metrics or traffic allocations for the testing; they are letting Sensei figure that out automatically, improving over time as it learns more about the customer. The control experience (the backup as Lindsay called it) ensures that no experience will ever perform worse than it, so marketers can be sure they are delivering at a minimum an experience they know works for most of their customers.
One-Click Personalization works across all digital properties including the website, mobile web or mobile apps, emails, kiosks and more. It’s not about personalizing a particular section, or container, on a page, but about an overall experience. Lindsay said it’s creating an ongoing experiment, except you don’t have to do anything, the system does it all for you.
The marketer still needs to create a set of experiences, but now they have some flexibility. They can flex their creative skills and create a series of experiences, letting Target (and Sensei) determine which experiences work best for a customer.
Other personalization improvements in Target
Along with the new One-Click Personalization, Lindsay noted a couple of additional updates for personalized experiences.
Precision Content Targeting lets a marketer customize a particular container within a web page (or other experience) to provide specific content to defined audience or behavioral segments. Integration between Target and Analytics provides these “Experience Versions,” the data needed to show the correct content. This approach gives the marketer more control which is important when they know they need to show certain offers or content as part of an agreement or campaign.
Offer Recommendations is powered by Adobe Sensei as well. Lindsay said it’s based on NLP (natural language processing) research where interactions are interpreted and used to suggest visitor intent. This leads to a more personalized item or product offer for each visitor.
Recommendations isn’t available yet but is expected in a Fall Beta. However, it’s an interesting approach to identifying the best recommendations to visitors because it looks at what the visitor is doing on the website to understanding what the best offers may be to show.
“Think of every digital interaction a consumer has with a brand, be it reading a blog post or watching a video, as part of a dialog. Now think of the sum of these actions as a query or a meaningful expression of intent. At the machine level, Adobe Sensei is interpreting these queries and figuring out how to respond. Adobe Target takes it from there by displaying recommended products or content—ones that are highly relevant to a consumer based on what we’ve learned about them throughout this dialog.”
The new recommendations engine won’t be just for Adobe Target. The expectation is that it will be made available as a standalone solution in the Adobe offering, and used by other solutions in the Marketing Cloud.
My take - the next phase of personalization
When I was scrolling through my LinkedIn feed the other day, I came across a post that asked why we were still using AI for simple things when there were so many other possibilities. One of those possibilities was personalization.
The current approach to personalization using segments or audiences, defined manually by marketers is okay, and organizations that aren’t personalizing at all need to start seriously thinking about how to get something in place to be competitive. But it’s just the start.
AI lets organizations get to that next level of personalization, one that doesn’t require the marketer to plan variations or define experiences by segment. AI leverages machine learning to know what a customer is interested in and wants to see at an individual level. Adobe’s One-Click Personalization and Offer Recommendations is the first step in improving personalized experiences. But we have a long way to go, and we need to start seeing other implementations in experience tools.