Blog Post:Last month, artificial intelligence, neural networking and data science took a collective leap forward when an AI program beat an expert player at the game of Go. Go had been widely considered the final frontier, a seemingly impossible hurdle with borderline-infinite complexity. Yet here we are. Go is on the verge of being solved. I’ve never played it personally, but the announcement still gave me goosebumps. Computing achievements like this one push us closer and closer to the highest heights in every field imaginable. It’s thrilling, isn’t it? But this is far from the only bright spot on the landscape—it’s also a thrilling time to be an Adobe Target customer, too. The organization is deeply invested in advancing data science rigor, with a keen eye on taking complex data sets and working machine magic to make marketers’ work easier and more impactful. Our approach to data science is simple: increase the power, depth and intelligence of our offerings, while making these benefits even more intuitive via a process we’ve dubbed Guided Optimization—more to come on that later. So what do these updates and advances mean for you? Plenty. We’ve made even more incredible strides in personalizing content plus “just for you” recommendations and site experiences that take the power of machine learning and data science and put it all firmly in the palm of your hand, in new and intuitive ways, even. Sound good? Well then, get ready because this is going to be an exciting year. Here’s what to look for—and leverage—starting right now. New lifetime Value (LTV) Algorithm to Optimize the Whole Journey J. Wellington Wimpy of Popeye fame once declared he would “gladly pay you Tuesday for a hamburger today.” Dubious intentions of evading payment aside, the concept of delayed but still increased rewards is always an interesting one, especially from a marketing perspective. Would you rather I give you one hamburger today or five hamburgers in a week? While your answer may be influenced by how hungry you are right now, there’s a better than average chance you’ll take the five hamburgers, right? Adobe Target’s new lifetime value (LTV) algorithm isn’t too far off from Wimpy’s hamburger request. In addition to maximizing the highest immediate reward, Target now enables organizations to optimize against all possible future interactions—the LTV of a customer, regardless of the number of interactions. Another way to think about this LTV process? This algorithm makes content decisions based on recorded interactions, events, and profile data. By aligning those inputs with potential trajectories and rewards associated, we get a comprehensive model for LTV. It’s incredibly beneficial for digital properties with a high quantity of returning visitors—the higher the ratio of return visitors, the better the algorithm performs. Voila! Deliver Personalized Recommendations to Visitors that are “Just for Them” Next is an exciting new approach to personalized visitor recommendations. Recommendations are an entirely different beast. Separate from content and, even, the user experiences described above, recommendations are specific things—products, services, offers, videos, articles—that you want to show visitors to entice valuable actions. With Adobe Target’s powerful new recommendations capabilities, marketers will be able to surface highly compelling and personalized products, media, or materials to customers, ensuring more spot-on experiences and increasingly high value next steps emerging from those touch points. These context-aware, real-time decisioning algorithms are heavily catered to the individual and based on existing personas. One algorithm—the “Just for You” algorithm—will be an integral piece of Adobe Primetime, delivering next-generation personalized TV recommendations. “Just for You” leverages a multitude of data points sourced from Adobe Analytics, building a personalized model for every web or app visitor. We’re also releasing a new parameterized profile and search-driven algorithm that uses stacked collaborative filtering to deliver recommendations based on profile data and user inputs like search queries, facet or filtering decisions. Imagine extending the current capabilities of a category affinity to an “everything affinity.” Imagine how today you have a favorite category, but tomorrow you’ll you’ll have a favorite color, brand, style, author, director, genre, size...the list goes on and on. And all of this will be factored into a model, adding incredible depth to 18 out-of-the-box algorithms inside of Adobe Target. Add Data Science to A/B/N Testing to Create Unique, Optimized Experiences Last year Target introduced automation into A/B/N activities with a feature we call “auto-allocate.” If you’re new to this lingo, A/B or A/B/N can both refer to multiple variants in an A/B test – A/B/C/D/E for instance – but executed in classical blind test. MVT is our term for full-factorial-optimization. While my colleague Drew Burns covered this in-depth last October, here’s the basic gist: Target now serves the highest-performing experiences to new visitors even more often, increasing your ability to gain valuable wins while, at the same time, making use of traffic for your tests. Adoption has been incredible and our customers can’t get enough. We’re eager to introduce personalized A/B testing, our latest and greatest leap into data-driven marketing automation, to our testing-hungry audience. This enhancement allows marketers to move beyond content and dynamically deliver entire site experiences to visitors using personalization algorithms. Whereas auto-allocate focused on finding the best holistic experience and ensuring it got a bigger piece of the traffic pie, personalized A/B testing takes things even further, intelligently delivering the best experience to each user based on their highly predictive but still anonymous profile. Think about the power here. Imagine your current personalization practices and processes—chances are you’re focused on giving customers highly relevant, highly personalized offers, messages or calls-to-action in prominent site sections. If you are, that’s great—but just imagine moving beyond those messages and into the underlying user experience associated with your site or app. What if you could personalize your navigation, button colors, or entire layout with the same comfort and familiarity level of Visual Experience Composer, which you already use to run A/B tests? What if you could tailor the entire objective of your site and edit, hide, show, rearrange, move, or resize entire components to deliver the most relevant experience possible? You’d love it, right? I wasn’t kidding—this is an exciting time to be an Adobe Target customer. You asked and we answered. We’re democratizing data and the experiences that come with it, making our tools, resources and processes easier to leverage and more powerful on integration, even if you aren’t a data scientist. The end result is a better personalization process. Because our goal is always relevance—being more relevant for you and your business so you can, in turn, deliver those incredibly spot-on experiences for your customers. And now, with these exciting changes, we’re able to do even more, and help you move that personalization needle further than you ever thought possible. Author: Date Created:March 21, 2016 Date Published: Headline:Adobe Target Applies Data Science Rigor and New Algorithms to Personalize Content, Recommendations and Site Experiences Social Counts: Keywords: Publisher:Adobe Image:https://blogs.adobe.com/digitalmarketing/wp-content/uploads/2016/03/AdobeStock_105102569-e1458590560221.jpeg

Last month, artificial intelligence, neural networking and data science took a collective leap forward when an AI program beat an expert player at the game of Go. Go had been widely considered the final frontier, a seemingly impossible hurdle with borderline-infinite complexity.

Yet here we are.

Go is on the verge of being solved. I’ve never played it personally, but the announcement still gave me goosebumps. Computing achievements like this one push us closer and closer to the highest heights in every field imaginable. It’s thrilling, isn’t it?

But this is far from the only bright spot on the landscape—it’s also a thrilling time to be an Adobe Target customer, too. The organization is deeply invested in advancing data science rigor, with a keen eye on taking complex data sets and working machine magic to make marketers’ work easier and more impactful. Our approach to data science is simple: increase the power, depth and intelligence of our offerings, while making these benefits even more intuitive via a process we’ve dubbed Guided Optimization—more to come on that later.

So what do these updates and advances mean for you? Plenty. We’ve made even more incredible strides in personalizing content plus “just for you” recommendations and site experiences that take the power of machine learning and data science and put it all firmly in the palm of your hand, in new and intuitive ways, even. Sound good? Well then, get ready because this is going to be an exciting year. Here’s what to look for—and leverage—starting right now.

New lifetime Value (LTV) Algorithm to Optimize the Whole Journey
J. Wellington Wimpy of Popeye fame once declared he would “gladly pay you Tuesday for a hamburger today.” Dubious intentions of evading payment aside, the concept of delayed but still increased rewards is always an interesting one, especially from a marketing perspective. Would you rather I give you one hamburger today or five hamburgers in a week? While your answer may be influenced by how hungry you are right now, there’s a better than average chance you’ll take the five hamburgers, right?

Adobe Target’s new lifetime value (LTV) algorithm isn’t too far off from Wimpy’s hamburger request. In addition to maximizing the highest immediate reward, Target now enables organizations to optimize against all possible future interactions—the LTV of a customer, regardless of the number of interactions.

Another way to think about this LTV process? This algorithm makes content decisions based on recorded interactions, events, and profile data. By aligning those inputs with potential trajectories and rewards associated, we get a comprehensive model for LTV. It’s incredibly beneficial for digital properties with a high quantity of returning visitors—the higher the ratio of return visitors, the better the algorithm performs. Voila!

Deliver Personalized Recommendations to Visitors that are “Just for Them”
Next is an exciting new approach to personalized visitor recommendations. Recommendations are an entirely different beast. Separate from content and, even, the user experiences described above, recommendations are specific things—products, services, offers, videos, articles—that you want to show visitors to entice valuable actions. With Adobe Target’s powerful new recommendations capabilities, marketers will be able to surface highly compelling and personalized products, media, or materials to customers, ensuring more spot-on experiences and increasingly high value next steps emerging from those touch points. These context-aware, real-time decisioning algorithms are heavily catered to the individual and based on existing personas. One algorithm—the “Just for You” algorithm—will be an integral piece of Adobe Primetime, delivering next-generation personalized TV recommendations. “Just for You” leverages a multitude of data points sourced from Adobe Analytics, building a personalized model for every web or app visitor.

We’re also releasing a new parameterized profile and search-driven algorithm that uses stacked collaborative filtering to deliver recommendations based on profile data and user inputs like search queries, facet or filtering decisions. Imagine extending the current capabilities of a category affinity to an “everything affinity.” Imagine how today you have a favorite category, but tomorrow you’ll you’ll have a favorite color, brand, style, author, director, genre, size…the list goes on and on. And all of this will be factored into a model, adding incredible depth to 18 out-of-the-box algorithms inside of Adobe Target.

Add Data Science to A/B/N Testing to Create Unique, Optimized Experiences
Last year Target introduced automation into A/B/N activities with a feature we call “auto-allocate.” If you’re new to this lingo, A/B or A/B/N can both refer to multiple variants in an A/B test – A/B/C/D/E for instance – but executed in classical blind test. MVT is our term for full-factorial-optimization.

While my colleague Drew Burns covered this in-depth last October, here’s the basic gist: Target now serves the highest-performing experiences to new visitors even more often, increasing your ability to gain valuable wins while, at the same time, making use of traffic for your tests. Adoption has been incredible and our customers can’t get enough.

We’re eager to introduce personalized A/B testing, our latest and greatest leap into data-driven marketing automation, to our testing-hungry audience. This enhancement allows marketers to move beyond content and dynamically deliver entire site experiences to visitors using personalization algorithms. Whereas auto-allocate focused on finding the best holistic experience and ensuring it got a bigger piece of the traffic pie, personalized A/B testing takes things even further, intelligently delivering the best experience to each user based on their highly predictive but still anonymous profile.

Think about the power here. Imagine your current personalization practices and processes—chances are you’re focused on giving customers highly relevant, highly personalized offers, messages or calls-to-action in prominent site sections. If you are, that’s great—but just imagine moving beyond those messages and into the underlying user experience associated with your site or app. What if you could personalize your navigation, button colors, or entire layout with the same comfort and familiarity level of Visual Experience Composer, which you already use to run A/B tests? What if you could tailor the entire objective of your site and edit, hide, show, rearrange, move, or resize entire components to deliver the most relevant experience possible? You’d love it, right?

I wasn’t kidding—this is an exciting time to be an Adobe Target customer.

You asked and we answered. We’re democratizing data and the experiences that come with it, making our tools, resources and processes easier to leverage and more powerful on integration, even if you aren’t a data scientist. The end result is a better personalization process. Because our goal is always relevance—being more relevant for you and your business so you can, in turn, deliver those incredibly spot-on experiences for your customers. And now, with these exciting changes, we’re able to do even more, and help you move that personalization needle further than you ever thought possible.