Blog Post:There is an interesting relationship between content and personalization. As I said in a previous post, content is king, but it needs a serious sidekick to deliver maximum value. Likewise, personalization systems are incredibly content- and data-hungry, constantly devouring every article, video, offer, and touchpoint you put in their paths. It is a powerful, ongoing, and value-rich cycle. So, now the big question is “What do you do when you are lacking the content or the data — or even both — to deliver relevance at scale? What do you do when those rich, high-value foundational pieces are missing? Understanding the Personalization Engine A personalization engine is, essentially, a series of algorithms and modeling processes all tied to a targeting system. This system makes constant decisions to deliver specific content against specific criteria — in other words, visitors and the data they represent. Does this person qualify for this experience? Does he or she meet the criteria for this journey? This decision making can be as simple or complex as you want it to be. It can be a binary decision: new visitors see this banner, while repeat visitors see something entirely different, for example. These processes work with just a few pieces of data and content and can even work with a limited audience. It is not resource heavy, because a human marketer can process the data and synthesize it fairly easily. Beyond that and, chances are, you will need an automated-personalization sidekick. Only a machine partner can make decisions quickly and granularly with a degree of statistical confidence — in other words, decisions you can take to the bank. Where the Bottleneck Lies Given how much data and content a personalization platform needs, delivering relevance can be a real bottleneck for organizations. How can a marketer tap into the power of personalization, especially when content, data, and/or traffic is lacking? And from there, at what point should a human drive the ship, and when should a machine take over? The first thing to consider is whether your decision making is simple enough for a human marketer to tackle or so complex it needs a machine. Is the decision based on simple rules like, “always show people from San Francisco this welcome hero,” or more complex (but still manageable) rules like, “always show this hero to men from San Francisco who shop on Saturdays during the winter and who have purchased before”? Where is the line drawn when rules upon rules upon rules add so many layers that you need help with the heavy lifting? Wherever that line falls, Adobe Target can be deployed to do that vital modeling and decision making; but even then, the problem is not solved. Why? Because we are right back to the content conundrum. My colleague, Loni Stark, talks often about the notion of content velocity and how it is essential that you continue delivering the right images, offers, and other pieces quickly and seamlessly, while keeping pace with the mounting demands of increasingly hungry personalization platforms. It is not just a consideration; it is a real hurdle for data-driven marketers, as is the flipside of this: being smarter and more strategic with the content you have. Figuring out how to do more, more, more without more, more, more is not an easy thing to do. However, that is where data science comes in, helping you to put the right content in front of the right people, ensuring greater value for both the consumer experiencing it and for your organization. Once personalization is part of the conversation, it is possible you will not need quite as much content. Maybe now, you do not have to show 20 different scenarios or hero images, but you still need content that is going to work hard and resonate in a big way. Where Your Organization Falls All of that said, many organizations are not quite here yet — in a place where content velocity, machine learning, and complex personalization algorithms come into play. Many companies are still trying to create recommendation experiences, for example. It takes most businesses time to successfully clear these specific personalization hurdles. And here is the thing: simple personalization is great, too. You do not need to boil the ocean right away. Instead, step by step, grow your optimization and personalization strategy, gaining more traction and more wins along the way. That is one of the great things about personalization: even a little moves the needle. Understand the content you can deliver — X offers in Y timeframe, for instance — and dig in to the data you have. Try delivering a different experience for tablet users versus desktop users, or ensuring that mobile visitors always get a geo-specific message. See whether the tests you are rolling out help you achieve business goals. Determine what is working and what is not. At Adobe, we have been trying to simplify content creation for all levels of personalization processes. By leveraging the power of our Visual Experience Composer — part of Adobe Target — marketers can be more self-sufficient and even do more with less. Instead of creating more content by constantly returning to your designers or creative teams and saying, “I just need one more ad that says ‘vacation’ instead of ‘holiday’,” marketers can tackle many of these experiences themselves. It allows you to be extremely nimble, changing pieces of the consumer experience with just a few quick clicks. And again, if that is all you are doing — swapping out ‘holiday’ for ‘vacation’ or delivering this experience for new visitors versus that experience for returning ones — that is a good first step. If, to any degree, you are using data to personalize experiences, chances are that you are going to do better than you would without delivering those varying touchpoints. Personalization is about fine-tuning experiences, but it also requires significant data and content to keep feeding the platform. Understand what you have, and what you can realistically produce, and then chart a course from there. Even if content and data are lacking, you can still deliver some degree of personalization and then refine more over time. Some is better than none, though! So, do not let a lack of content or data hold you back from dipping your toe in the personalization waters. Author: Date Created:April 11, 2016 Date Published: Headline:Personalization When Content and Data Are Lacking Social Counts: Keywords: Publisher:Adobe Image:https://blogs.adobe.com/digitalmarketing/wp-content/uploads/2016/04/AdobeStock_71713685-1-e1460001154613.jpeg

There is an interesting relationship between content and personalization. As I said in a previous post, content is king, but it needs a serious sidekick to deliver maximum value. Likewise, personalization systems are incredibly content- and data-hungry, constantly devouring every article, video, offer, and touchpoint you put in their paths. It is a powerful, ongoing, and value-rich cycle.

So, now the big question is “What do you do when you are lacking the content or the data — or even both — to deliver relevance at scale? What do you do when those rich, high-value foundational pieces are missing?

Understanding the Personalization Engine
A personalization engine is, essentially, a series of algorithms and modeling processes all tied to a targeting system. This system makes constant decisions to deliver specific content against specific criteria — in other words, visitors and the data they represent. Does this person qualify for this experience? Does he or she meet the criteria for this journey?

This decision making can be as simple or complex as you want it to be. It can be a binary decision: new visitors see this banner, while repeat visitors see something entirely different, for example. These processes work with just a few pieces of data and content and can even work with a limited audience. It is not resource heavy, because a human marketer can process the data and synthesize it fairly easily. Beyond that and, chances are, you will need an automated-personalization sidekick. Only a machine partner can make decisions quickly and granularly with a degree of statistical confidence — in other words, decisions you can take to the bank.

Where the Bottleneck Lies
Given how much data and content a personalization platform needs, delivering relevance can be a real bottleneck for organizations. How can a marketer tap into the power of personalization, especially when content, data, and/or traffic is lacking? And from there, at what point should a human drive the ship, and when should a machine take over?

The first thing to consider is whether your decision making is simple enough for a human marketer to tackle or so complex it needs a machine. Is the decision based on simple rules like, “always show people from San Francisco this welcome hero,” or more complex (but still manageable) rules like, “always show this hero to men from San Francisco who shop on Saturdays during the winter and who have purchased before”? Where is the line drawn when rules upon rules upon rules add so many layers that you need help with the heavy lifting? Wherever that line falls, Adobe Target can be deployed to do that vital modeling and decision making; but even then, the problem is not solved. Why? Because we are right back to the content conundrum.

My colleague, Loni Stark, talks often about the notion of content velocity and how it is essential that you continue delivering the right images, offers, and other pieces quickly and seamlessly, while keeping pace with the mounting demands of increasingly hungry personalization platforms. It is not just a consideration; it is a real hurdle for data-driven marketers, as is the flipside of this: being smarter and more strategic with the content you have. Figuring out how to do more, more, more without more, more, more is not an easy thing to do. However, that is where data science comes in, helping you to put the right content in front of the right people, ensuring greater value for both the consumer experiencing it and for your organization. Once personalization is part of the conversation, it is possible you will not need quite as much content. Maybe now, you do not have to show 20 different scenarios or hero images, but you still need content that is going to work hard and resonate in a big way.

Where Your Organization Falls
All of that said, many organizations are not quite here yet — in a place where content velocity, machine learning, and complex personalization algorithms come into play. Many companies are still trying to create recommendation experiences, for example. It takes most businesses time to successfully clear these specific personalization hurdles.

And here is the thing: simple personalization is great, too. You do not need to boil the ocean right away. Instead, step by step, grow your optimization and personalization strategy, gaining more traction and more wins along the way. That is one of the great things about personalization: even a little moves the needle. Understand the content you can deliver — X offers in Y timeframe, for instance — and dig in to the data you have. Try delivering a different experience for tablet users versus desktop users, or ensuring that mobile visitors always get a geo-specific message. See whether the tests you are rolling out help you achieve business goals. Determine what is working and what is not.

At Adobe, we have been trying to simplify content creation for all levels of personalization processes. By leveraging the power of our Visual Experience Composer — part of Adobe Target — marketers can be more self-sufficient and even do more with less. Instead of creating more content by constantly returning to your designers or creative teams and saying, “I just need one more ad that says ‘vacation’ instead of ‘holiday’,” marketers can tackle many of these experiences themselves. It allows you to be extremely nimble, changing pieces of the consumer experience with just a few quick clicks.

And again, if that is all you are doing — swapping out ‘holiday’ for ‘vacation’ or delivering this experience for new visitors versus that experience for returning ones — that is a good first step. If, to any degree, you are using data to personalize experiences, chances are that you are going to do better than you would without delivering those varying touchpoints. Personalization is about fine-tuning experiences, but it also requires significant data and content to keep feeding the platform. Understand what you have, and what you can realistically produce, and then chart a course from there. Even if content and data are lacking, you can still deliver some degree of personalization and then refine more over time. Some is better than none, though! So, do not let a lack of content or data hold you back from dipping your toe in the personalization waters.