Auto-Target: Personalization reimagined by Adobe Target, Powered by Sensei
by Jason Hickey
posted on 03-20-2017
We’ve been talking about personalization for more than a decade now. So, why then does the conversation always feel roughly the same when so much has changed. In that timeframe, we’ve had advances both in algorithms and machine learning. Marketers can now take matters into their own hands, applying data science to personalize across the enterprise. Just look at our Visual Experience Composer as exhibit A for that point. The appetite for personalization hasn’t diminished — report after report show it’s a top priority for investment on every vertical.
So, why hasn’t the personalization conversation changed? Why do at least half of businesses limit their personalization to the homepage hero, carousel, marquee, primary banner, or secondary banner of their sites or apps? So much untapped potential.
We’re Excited to Introduce One-Click Personalization With Auto-Target.
At Adobe, we know the potential of personalization and have been on a mission to reinvent this space. We have been developing technology both to blur the lines between testing and personalization and to democratize personalization at scale across all businesses and in all industries. To blur those lines, we needed to remove the artificial barriers between A/B testing and personalization from within the activity setup workflow — and we did exactly that in our newest Adobe Target Premium feature: Auto-Target — one-click personalization available in the A/B testing workflow of Adobe Target Premium.
Now, more people in your company can do more types of personalization — all with just one click.
Auto-Target Enables Faster — and More — Personalization.
Here’s the path we took to develop Auto-Target as well as some of the reasons for the choices we made as we worked to develop this single-click personalization.
Just a decade ago, you might have used three or more vendors to do everything you needed in online testing and personalization — A/B testing, multivariate testing, rules-based targeting, algorithmic personalization, and even recommendations. Adobe Target Classic had these functions and more spread across a few user interfaces and workflows. Then, we launched Adobe Target Premium in 2013. The entire spectrum of testing and automation suddenly became available under one roof for the first time ever — with one common set of rules, governance, and workflows.
We’ve come a long way with Adobe Target, but we’re still working on removing the vestiges of outdated ways of thinking in our solution. For example, while Adobe Target Premium’s Automated Personalization feature drives significant lift for many of our customers, you select it in the user interface (UI) as the first step in setting up a Target activity. Only later do you specify the experiences that will be part of that activity. We wanted to give you — as a marketer or product owner — the option to fundamentally flip this around, defining the experiences first and then choosing what activity type to apply.
Auto-Target is this brand-new option. From within the A/B-activity setup workflow, you can now choose whether to use Manual Allocation (traditional A/B testing), Auto-Allocate (a multiarmed bandit), or the new Auto-Target after you’ve made your experience decisions. By selecting Auto-Target, you choose to apply our specialized personalization — based on random-forest machine learning — to the entire experience. We’ve placed this choice in the second step of a three-step workflow so that, with a single click, you can now experiment with and discover the power of personalization on your experiences.
By the way, we’ve noticed that, when we’ve made this feature available to customers, they’ve naturally decided to give it a try — its placement in the workflow sequence seems to coax customers into trying it. And they like it.
Auto-Target Enables More Types of Personalization in More Places.
Our one-click personalization is the feature that’s moving the ball forward with personalization. That’s pretty exciting. What’s more exciting though is learning all the great ways you might use this new personalization feature in Target. To really understand how Auto-Target offers a different way to personalize, you need to look at it in comparison with Automated Personalization. It’s different, and I’ll explain how.
We designed the activity-type Automated Personalization for full factorial personalization; meaning, each change you make on any given page multiplies the number of experiences being created for the activity. So, if you have three background images and two different sets of offer copy, you have six different combinations you can examine to determine which is best to deliver to a visitor. Automated Personalization was built for personalizing at the content level, not the experience level.
But, it’s 2017, and we’re well beyond simply personalizing content alone. Auto-Target, as part of the A/B workflow, does not have the multivariate strings attached that Automated Personalization does. With Auto-Target, if you change five things on a page, that doesn’t represent five different variables that multiply the possible experience combinations to test and personalize. Instead, those five changes produce one holistic experience that you can add as a personalization-experience option.
Why is that so important? With our custom code editor and Visual Experience Composer, the types of changes you can make are unlimited. Just a few years ago, you might have run an A/B test on your checkout flow (1-step versus 3-step, for instance) to find the best overall performer. When the test concluded, you’d deliver the better-performing experience to everyone. With Auto-Target, you keep both journeys and personalize the appropriate journey to the right person based on each person’s history, context, and user profile.
The beauty of Auto-Target is that, with just one click to turn it on, you can transform any A/B test in Adobe Target Premium into a personalized activity. In other words, every A/B test can have multiple winners — each one mapped to an individual visitor — that adapt over time based on additional actions the visitor takes.
Extend Machine Learning Beyond Web and Mobile With Form Composer.
We understand that personalization goes well beyond web and mobile sites. To meet your need to apply machine-driven personalization beyond the web, we added the Form Composer to Adobe Target, enabling you to use Auto-Target with native and hybrid mobile apps, set-top boxes, videogame consoles, Internet of Things (IOT) devices, and in-store kiosks.
When you think about personalization, it tends to involve both the marketer and the consumer. The marketer wants to deliver the best experience possible to the consumer and uses a machine-learning sidekick to get the job done. But, imagine a scenario in which your consumers are interacting with a human face of your brand and not a screen. Personalization at this level needs to involve a third voice — the brand ambassador. You can use Auto-Target in these non-digital scenarios, too. For example, a major telecommunications provider could use it to test, personalize, and optimize their call-center scripts to improve the consumer experience while building a more efficient customer-support team.
Personalize in More Ways and Places With Adobe Target.
If you’re an Adobe Target Premium user, Auto-Target allows you to truly address the goal of the experience business — to deliver exceptional experiences that cater to your customers and make them feel valued. It does this by letting you personalize experiences — not just content — and it extends that ability any place your visitor connects with your brand. We also believe that, when you see how easy it is to give it a try — you will. And, once you experience the results, we’re confident that you’ll start personalizing with Auto-Target everywhere. The conversation around personalization is finally about to change.
Topics: News, Personalization