The Art of the Possible with Adobe Target: When Humans and the Machine Work Hand in Hand
We often use “the art of the possible” and “Adobe Target” in the same breath. That’s because Adobe Target lets brands leverage artificial intelligence (AI) powered by Adobe Sensei to more deeply personalize the experiences they deliver their customers. Adobe Target also delivers that better personalization to new digital touchpoints that previously were the stuff of science fiction movies, but today surround our everyday lives in the Internet of Things. What’s amazing is that it doesn’t take an army to create these types of experiences. With Adobe Target, setting up exceptional experiences is easier and faster than ever, and delivering personalization at scale is achievable and sustainable.
With the stage set, I’d like to share our latest Adobe Target features that showcase how we are making the art of the possible a reality. These new features are all about how we’re letting humans collaborate with “the machine” in exciting new ways.
Breaking open the black box
A major barrier to adopting AI for personalization has consistently been the inability to learn what influenced the decisions of the AI model. What customer attributes did the model find most important and predictive in delivering a given experience or offer? And what specific audiences did it surface and identify as important for delivering a specific experience or offer? Even if conversion rates or revenue skyrocket — as they often can with AI — without that understanding there’s no way to learn from the activity, iterate on it, and improve the experience.
New Personalization Insights reports for Auto-Target and Automated Personalization, two key personalization capabilities powered by Adobe Sensei, provide that all-important understanding. Now marketers can see the top customer attributes that influenced the AI model in delivering its experiences. They can also see significant audiences that the AI model determined were distinctive enough to warrant receiving a specific experience.
A giant leap forward with Adobe Target Recommendations
Recommendations of products and content, such as videos and customer support articles, are one of the most powerful ways marketers can increase sales and customer engagement and consumption. They’re also a powerful tool in continuing the “conversation” with a visitor or customer at their next interaction based on their category affinity and other profile data. Almost every business can and should be leveraging recommendations in refining its personalization strategy. Yet scaling personalized recommendations to a variety of different audiences presents a challenge.
To help marketers overcome that challenge, Adobe Target now helps refine recommendations with AI-powered decisioning in three key ways:
- First, marketers can now leverage Adobe Sensei to automatically choose the best Adobe Target algorithm to use to deliver personalized recommendations to a broader spectrum of audiences. For example, a home improvement company can use AI to tailor recommendations to each important audience—its primary home builder audience, but also secondary audiences like DIYers—all within a single campaign.
- Secondly, Adobe Target now offers a new recommendations algorithm, powered by Adobe Sensei, that automatically factors in customer preferences, such as geography or favorite color, that are most predictive of a desired behavior, like a purchase. The algorithm layers this customer profile information on top of all the data the model currently considers, like what other people bought or viewed, to further refine and deliver highly personalized recommendations to each customer.
- Finally, marketers can now use recommendations as an offer type in an Adobe Target activity rather than as a standalone activity. This provides the ability to test an experience with a recommendations “tray” of suggestions against one with no recommendations—and also against one with two trays of recommendations that use different algorithms. The result is much more dynamic recommendations, including the ability to easily apply Adobe AI features like Auto-Target or Automated Personalization to recommendations.
Leverage your own data science to target customers
Many businesses today have their own internal data scientists who develop models and output valuable business-specific knowledge like propensity and customer churn scores. The value of internal data science can be further extended by applying this knowledge when personalizing customer experiences and offers. Now Adobe Target helps dynamically compare numerical values, such as these scores, to create and target audiences.
Using this capability is pretty straightforward. For example, a brand could target customers with an offer related to living room furniture if they have a higher propensity score for purchasing living room furniture than kitchen furniture.
Personalizing customer experiences with voice interactions
In a recent survey, Adobe found that 64 percent of consumers think they’ll use voice assistants more and more in the next few years. Adobe Target can personalize almost any experience, whether it’s conveyed through images, copy, video — or voice.
For example after a person has viewed an ad fora bank, the bank can use Adobe Target to suggest a credit card offer to a customer via Alexa, based on pre-qualification and geolocation. Or a company that connects customers with doctors, dentists, or other healthcare professionals, could target a customer with a highly personalized response when asked, “Alexa, suggest nearby pediatric dentists who have a five-star rating.”
Personalizing shopping experiences that use augmented reality (AR)
What’s more telling about customers’ feelings about their experiences with a brand than the face they make in response to each experience? And what if you could capture those facial expressions as a stream of data about the customer and then use that data to dynamically deliver in-the-moment personalized recommendations?
Adobe Research is looking into using Adobe Sensei to offer dynamic product recommendations within an augmented reality (AR) setting based on a customer’s facial expressions. For example, a customer may try on sunglasses in a store that uses AR. If the customer furrows her eyebrows after trying on round-framed sunglasses, Adobe Target could suggest aviator-style sunglasses, and could continue to make suggestions in response to the customer’s reaction until the customer zeros in on and purchases the perfect pair.
Who would have thought…
I’ve worked with Adobe Target and in optimization and personalization almost since this area of marketing began. I continue to marvel at what is possible in this space — things that even two years ago might have caused me to laugh out loud if someone suggested we could do them. And yet, here we are, deep into using AI to personalize. Making it not only possible to tap into the power of incredibly sophisticated AI, but actually making it easy and understandable to do so.