December 2006

Who Wants What Your CM Puts Out? Bob Boiko President, Metatorial Services Inc. Affiliate Professor, University of Washington iSchool bob@metatorial.com As any good writer or speaker will confirm, if you do not live by the code of your audiences you will die by the sword of their neglect. Audiences consume your information. It is not so hard to begin to understand what they want and how to provide it for them. From this understanding you can begin to craft a deal whereby you give your audiences what they want and in return they help your organization meet its own goals. Introduction Customers are audiences for your information. To be customer-centric is to be audience-centric. I like the term audience a lot because it has just the right set of connotations. An audience can love or hate you. They can stand up and applaud or fall asleep. They pay good money (or other tender) to consume your information and they expect value in return. They are customers, but they are also users, members, constituents, opinions leaders, middle level managers, partner liaisons, and a million other kinds of people. They are the kinds of people who can help you reach your goals. Moreover, for our purposes, they are the kinds of people who will help you meet your goals if you can just get to know them enough to supply them with the right information. For you, a person, to know another person is one thing. For a computer to know a person is quite another. For you or me, a person isn't some set of data points; she's a complex being whom we intuitively "get." You say that you know someone if you can recognize her and can accurately predict what she may do, say, and want. No computer that I know of can "get" a person in this sense, so you better settle for something less. You can settle for coming up with a few isolated traits and using them to try to predict wants. And, amazingly, that approach basically works. In most of the circumstances that you face in a Content Management System (CMS), you can limit the number of potential wants that you serve. You can also draw wide enough distinctions between your users that determining who's who and what they probably want isn't too hard. As you learn to discern more traits more accurately, you can continue to get better at predicting more wants for more people. In a CMS, it is not enough to want to know and serve your audiences. You have to encode your knowledge into a set of user profiles and rules that can guide the creation and delivery of content. In this article, I'll help you get from the desire to know your audiences to the steps you'll have to take to serve audiences with a CMS. What is an Audience? Although many disciplines talk around or about audiences, I've never seen the concept nailed down enough to become specifically useful in the context of a CMS. It would be nice if someone had already figured out who the most important audiences are for your organization and how to serve them, but this is probably not the case. If your organization has any existing information on this topic, it is probably in the form of some type of marketing analysis. Although not used exclusively by commercial organizations, such an analysis usually defines the kinds of people you want to sell to and how they might be convinced to buy. This sort of analysis is good as far as it goes, but it does not go nearly far enough. A marketing analysis defines external audiences. A content management audience analysis defines segments within your organization (administrative assistants, or line managers, for example), in partner organizations (purchasing agents or partner liaisons for example), as well as outside the organization. In addition, the people in a content management analysis might help meet goals that are not strictly economic (like employee satisfaction, or increased community service). My approach to audience analysis borrows heavily from marketing. I believe in the notions of segmentation and profiling. To automate information delivery with a CMS, you need to tie groups of people together by using data about them. But my approach borrows heavily as well from communication and computer science: The written and oral communication discipline has the idea of an audience analysis, which tries to define what you need to know to "speak" with clarity and authority to a particular group of people. Content management is necessarily involved in communication. In fact, I firmly believe that a content management approach will come to dominate organizational communication in the 21st century. What I call information strategy is really communication strategy. At its base, it is no different than what every modern political campaign (and now every administration) attempts to do to understand and move the audiences it speaks to. Computer science has the notion of a user as a kind of person that an application must serve psychologically and ergonomically. Content management is almost always concerned with users because information is very often delivered to them via some user interface. Information is increasingly delivered interspersed with computer functionality and cannot be detached from it. When users navigate through information they use functionality as well. This functionality completely obliterates the distinction between user interface, information structure, and text. Every audience you deliver to has to be able to use the delivery systems and, more to the point, has to be able to use (not make use of but make its way through) the information. So, the idea of an audience that I pursue is a combination of the profiling and direct marketing approach of the marketing discipline, the audience analysis approach of the communication discipline, and the usability approach of computer science. That's enough on the idea of audiences. Let's turn now to their practicalities. Audiences are at the Base of a CM Strategy At the same time as you want to serve your audience, you also want to get something from them. The more that you know about people, the better you can anticipate their needs and provide them with just the right content. The better you serve them, the more they will help you get what you want as an organization. This is the basis of content strategy-deliver content that helps your audiences help you reach your goals. The more effectively you can gather and deliver such content the more you will be able to help your organization. This is the basis of CMS strategy-build systems that deliver strategic content. To get to content and CMS strategy, you can begin with simple value propositions. For example, suppose as a non-profit relief organization you form this value proposition: Potential members want a way to help fight world hunger. We want more members so we can fight world hunger. This value proposition is balanced. Your audience's need is in harmony with your need. Of course, it's not this simple. Potential members must be found, they must hear your message, and they must be convinced that their financial and time commitment to you will alleviate enough hunger to be worth it. This is where content and CMS come in. The content strategy defines the content that will demonstrate the balance to potential members. The CMS defines the system that can cost effectively push that content through the communication channels most likely to reach these people. The content strategy defines the audience "Potential Member" in a way that is directly useful. It states what the defining traits of potential members are, and under what specific circumstances they should receive what types of content. The CMS details the machine that will make it possible to recognize potential members when they arrive, for example, at your home page. It details how you can more and more efficiently tailor and package information given the specific traits of potential members that you have discerned. A full content strategy covers all audiences and content that most significantly impacts your goals. A full CMS covers the flow of all that significant information from wherever it is originated to the eyes and ears of all audiences. Serving vs. Exploiting Audiences The better you know audiences, the better you can serve them. On the other hand, the better you know audiences, the more you can manipulate them. This paradox plays out on both sides of the computer screen. Users expect the Web sites that they visit to be smart enough to anticipate their needs. They gravitate toward sites that seem to know them and remember their preferences. On the other hand, users are wary or even hostile toward sites that ask a lot of questions. The question immediately comes to mind: "What are they going to do with this information?" Direct marketers live by the creed of "Know thy audience." They collect as much information as possible on you and then carefully craft a message that they think you may respond to. Direct marketers live and die by the lists of targeted audiences that they create. Marketers walk that very thin line between serving their audiences and exploiting them. And, very interestingly, the line isn't a sharp one. Consider the same piece of junk mail sent to two neighbors. The mail is a flyer advertising a long-distance telephone plan. Neighbor A has a plan and is happy with it. She feels put upon and manipulated and says, "I hate all these advertisements trying to get me to buy something!" Neighbor B just moved in and has been researching long-distance phone plans all day. She looks with interest on the ad and says, "How fortuitous to get this today. I wish that every phone company had sent me one." So, while it is up to you to draw the line between service and exploitation, I believe that the key to staying on the right side is to be sure your value propositions are balanced. If you're willing to give as much value as you expect from your audiences in return, the relationship involves no exploitation. Defining and Selecting Audiences To begin to define your audiences, first think about what level of audience segmentation is feasible. You might start by asking, "What's the smallest number of audiences that we can divide our users into and still derive tangible business benefit?" Or you may ask, "What audience segments does everybody agree on today?" Regardless of how simple or ambitious your approach is, count on at least a few iterations of defining and refining audiences before you begin to act on your analysis. Start with whatever segmentation analysis your organization has already done. Expect to add significantly to the current analysis, adding internal and partner audiences and charting how every audience is affected by information. Get used to the fact that your audiences will change over time. It's not so much that your organization will choose to serve new kinds of people. It's more that your segments will tend to get smaller and smaller as you learn more about the differences within a group or that you may become ready to serve audiences that you were not initially prepared to serve. It's possible right from the start of your content strategy to drill way down into audience segments and sub segments, producing a very detailed taxonomy. But within the level of segmentation that is possible, what level of segmentation is desirable? If you can, should you create audiences of one? Should you strive to serve each person individually? Supposing that this task is even feasible-that is, that you can put in place the technology to accomplish it-you should still question whether it's desirable. What does having audience segments of one really do for you? Most argue that the benefit lies in increased loyalty and a better sense of service and trust. Maybe so, but it comes with a cost as well. Can you segment your information so thoroughly that it's different for each person who receives it? Can you create and maintain user profiles that are rich enough to differentiate every user? Do you want to forgo the capability to develop messages with wide appeal and leveragability over a large number of people? All in all it is not a foregone conclusion that you should serve very small or even individual audience segments. Given that you can arrive at some reasonable segmentation, your next task is to decide which audiences are most important to serve. Here are some guidelines to help you rank your audiences: Clearly, the audiences that help you reach important goals are important. Audiences that help you meet multiple goals are more important than those that serve only one goal. Audiences that are large and can only really be served by some system are more important than those that are small enough to get their information informally or directly from the creator. Audiences that are particularly ill-served now are a better choice than those that do not perceive many unfulfilled needs for information. Audiences that need information you can readily collect are better than those whose information will be hard to come by. With all of these factors, and some of your own, you should be able to prioritize audiences. Match audience priority with goal priority and you can create a graded list of who you most need to communicate with to meet the goals of your organization. Combining audiences and goals also gives you a strong base for looking at the information in your organization with a more critical eye. Audience Attitudes and CM Systems Learning the attitudes that members of each audience are likely to hold toward your organization and content is an important task for you. From knowledge of attitudes, you can craft the appropriate messaging for each audience. The messaging can be encoded in content types and authoring templates. The following constraints help you understand how the audience may react to your content: Credibility: According to Aristotle, the perceived credibility of the speaker is more important than what she says in determining whether she's convincing to the audience. How do you establish credibility with this audience? How much credibility do you have now with its members? Have you experienced any particular failures or successes in the past with these people? Current beliefs: What does this audience already know and believe about the subjects that your content addresses as well as about your organization? Are you reinforcing or trying to change existing attitudes? Do members have any particularly strong positive or negative beliefs that you need to take into account? Ask yourself, "What do people of this ilk trust, respect, like, know, and believe?" Argument: What do members of this audience consider good arguments and examples? Do they respond more to a logical or an emotional appeal? Must you cite certain sources or quote particular people for them? Can you leverage scenarios or examples that the audience has already heard of? Style: What tone and presentation style does this audience expect and respond to? Can you advance the expected style to a new level in a way that shows respect for the existing style and innovation (if, that is, the audience responds to innovation)? What vocabulary and usage does the audience expect and respect? Openness to giving data: How much personal data does this audience want to give? What profile collection methods do members most respect and support? Are they likely to be concerned if you buy information about them from outside sources (direct-marketing companies, for example)? Once you understand these constraints, you must make them active in your CMS. Here are the basic ways you might do so: Content type structure. You can choose to encode decisions you make right into the definition of your various types of content. For example, suppose scientists are your audience and that they find a certain set of journals most credible. You can make a citation of at least one of those journals a defined and validated part of the content types that you will produce for them. Metadata. You can tag your content based on what you know about your audiences. For example, you might tag articles with a style metadata element with values such as: "expert," "professional," "novice" and "general public." Then, depending on what type of audience you have (scientists would likely be experts) you can deliver articles that are written in a style that they will respond to. Personalization rules. With solid knowledge about your audiences, content type structures and metadata, you can construct rules that automatically rout content to audiences. For example, if you know the field in which a scientist is an expert in, and you have created a "style" metadata field, then you can create a rule that prominently offers them the expert articles in their field via email as soon as they are available. If officially encoding your audience knowledge is too constraining, you can create written guidelines that tell authors what to do without enforcing it. For example, on the form where authors submit articles for scientists, you can link to the respected journals and suggest that authors reference them. A bit more formally, you can provide templates for the body of the article that lays out how you expect it to be structured but does not enforce that structure. Bringing it all Together To have a really audience-centric CMS, it is not enough to have your authors know about and write directly to your customers. You have to schematize audience awareness and turn it into user profiles and personalization rules. You have to think deeply about who your customers are and create a complex picture of how your content relates to them. Then you have to turn that complex picture into simple rules that a CMS can implement and you can maintain. A CMS does nothing more than collect and deliver content to audiences. To build a CMS that is worth the considerable effort, that delivery had better help your organization in some significant way. A full audience analysis tells you which people are most able to provide this significant help. A full content strategy tells you what information will inspire them to help. A full CMS strategy tells you what system you will need to collect and deliver the strategic content. If you do this early work well, you will have all you need to secure resources and support for your project, and confidently direct its implementation. It's no easy task, but don't worry if you don't get it right the first time, because you will likely be doing it over and over for the rest of your career. Note: This article is excerpted from Bob's "Content Management Bible," as well as his upcoming eBook and analysis tool "Crafting Information Strategy," which will be available late fall 2006 at Laughingatthecio.com. More information can also be found at cmbible.com.