On the Web, Context is Vital

Supply subject area context; avoid institutional context.

In a recent blog post, On the Web, context kills, speed saves, Gerry McGovern states:

A key difference between web writing and writing for print is that on the Web you need to avoid context and instead focus on instructional, how-to, task-based content.

Since one of my seven principles of Every Page is Page One is that an EPPO topic establishes its context, this quote made me think that either Gerry McGovern had gone bonkers or he was using the word context in a very different way than I use it. Turns out that it is the latter. (Sorry if you were hoping for fight here.) I actually agree with almost everything that McGovern says (I’ll get to the part I don’t agree with a little later). In fact, I have said many of the same things before, but in a different way.

Still, it’s worth examining McGovern’s argument here, if only because context is one of the most important concepts in communication, and (therefore?) subject to very different interpretations. Looking at McGovern’s argument will hopefully help make it clear what I mean when I say that an EPPO topic must establish its context.

It is axiomatic in communications that you must establish the context of what you are talking about. If you simply say, “Push the red button”, readers are left to wonder: Which red button? When should I press it? What happens if I press it? Is it my job to press it, or someone else’s? Do I need authorization to press it? Should I press other buttons first? Should I press more buttons afterwards? After I press it, should I stand still or run for my life? Without this contextual information, the instruction “Press the red button,” is not empowering, but paralyzing. If I cannot establish the reason and context for acting, I am powerless to act.

But this is not the kind of context McGovern is talking about. He’s actually talking about this kind of thing:

You need to find the nearest government children’s center. “Welcome to our page which we are no longer updating. The government is committed to the welfare and education of children. Our team has been established based on certain legislation, which we are about to tell you about in great detail. We are a high quality, well trained workforce and we would also like to tell you about how we work.”

This is the institutional context of the information. It is about the organization that provided the information, rather than about the location of the information in its subject area. It does not tell you when to press the red button; it tells you who manufactured the device with the button, who designed the button, and who chose that particular shade of red. But readers, for the most part, don’t care about any of this. As McGovern notes:

But on the Web, people just want to do stuff as quickly and easily as possible. Engaging with most organizational websites is only slightly more interesting than interacting with your dentist.

There is one point on which I do have to disagree with McGovern, though. He writes:

The words people search with set the context. Once they’re at your website, the context has already been established as far as they’re concerned.

Yes, the user’s search terms set the context of what they are seeking, but they don’t establish the context of what they have found. In a great many cases (especially in technical content) it can be very difficult for the reader to establish if the page they are looking at matches the context they were searching for. As I have noted before, findability does not end when the reader reaches a page with the sought information. The reader is following an information scent, and will only give each individual page a sniff before moving on to the next one. If a page does not smell like what they are looking for at first sniff, the reader will generally move on before ever discovering that the information the seek is in fact on that page. Establishing the context of the page in its subject matter is vital to the reader recognizing that the context of the page matches the context of their search.

I have noted previously that we are not always even aware of whose site we are on when we find information through a  web search. We are focused on our problem and the need for a solution, not the organizational context of the information provider. And since we are more likely to recognize the scent of the information we are looking for on a page that begins by establishing subject matter context rather than institutional context, it is likely that the pages where we are most likely to find our answers are also the ones where we are less likely to recognize where we are.

This is, of course, a problem for the people providing the information. Many of them are not providing this information out of pure altruism or an unwavering dedication to the proposition that information wants to be free. Most content creators are hoping to lead the reader on to other transactions with them, either by leading them directly to a purchase, or by at least raising their brand awareness or increasing esteem for the brand. I write this blog, in large part, because I want to lead you into hiring me (or my company, Analecta Communications Inc.) to provide content creation, content engineering, or content strategy services. If you don’t notice where this blog comes from, it has no chance to lead you into becoming a customer of mine.

Indeed, there can come a point in the reader’s interaction were they do want to establish the institutional context of the material they are reading. If they have found a repair procedure for a product they own, they may then want to figure out who wrote the procedure, to raise their confidence that the procedure won’t further harm their product. If the content has successfully interested the reader in doing business with the person who wrote it, the reader will then, of course, want to know exactly who that person or company is.

A website does, therefore, have to establish its institutional context somewhere. It doesn’t necessarily have to lead with it, as many sites do, but it has to be there somewhere. The problem is that search engines do not discriminate well between information pages and institutional context pages (nor can they know with certainty which you are searching for at any given moment). Thus, as McGovern notes, readers often end up on institutional context pages that don’t help them solve their immediate problem.

Often, general information that describes an area such as planning gets found in searches when people are actually trying to make a planning application. General marketing material about the launch of a product can be found years later when someone is actually trying to get a price for that product, or find a technical description of it.

This brings us to another important characteristic of Every Page is Page One topics: rich linking. As I have argued before, linking is the last mile of findability. Search may land the person looking for a planning application on a page that describes the planning department. We should anticipate that this kind of search near miss is going to happen a lot. To accommodate it, we should make sure that the planning department page links to the planning application page (and vice versa). And we should make this link a prominent one, not one of a dozen general links buried in the footer, but part of the context setting material for the planning department page, because that is where you have the few second to successfully engage and redirect the reader before they give up looking and start firing off strongly worded letters to their city councilor about how hard it is to get a planning application.

Context, therefore, is vital on the Web. But it is subject matter context that matters. Context provides not only the necessary information scent to allow the reader to recognize a page as the one they are looking for, but also the ability to guide the reader onward to the right content when the search engine has dropped them a little shy of their goal.

, , , , ,

6 Responses to On the Web, Context is Vital

  1. Joe Pairman 2013/12/11 at 22:43 #

    Great post, Mark. Context is so important — but appropriate context in appropriate places, as you say. And that’s an interesting point about institutional context in the sense of creating business. You didn’t talk about content marketing as such, but I think what you said applies to that area too. The challenge for content marketers is to present institutional context in a skillful and honest manner. Sometimes it does naturally belong within the body of a content page, but more often, it’s awkwardly tacked on, in a manner that convinces nobody.

    • Mark Baker 2013/12/16 at 17:11 #

      Thanks for the comment, Joe,

      I do believe that all content is marketing content (and that content created solely for “content marketing” purposes is the weakest form of marketing content). Certainly technical communications, in all its forms and media, is marketing content because it can have a profound affect on the customer’s inclination to buy, buy again, buy more, and recommend a product to others. I wish more tech writers would recognize this, rather than behaving as if marketing was beneath their dignity.

      The problem with organizational context, it seems to me, comes down to the question of generating leads. It seems like every week some content marketing guru is castigating businesses for doing lead capture in their content, arguing, quite rightly, that it annoys readers.

      In an ideal world, we would be content to allow readers to organically move from subject context to institutional context when they were ready to buy. In the real world, though, more actual sales get made by the companies who find a way to control the conversation. Potential customers may not want to give up their information, they may not want to become a lead, and then be sold to, but the companies that aggressively collect and follow up leads do seem to actually make sales.

      At the commercial level, then, the art may not be to avoid ever annoying any reader, but to be strategic about how and when to collect the data that turns them into a lead, even a the cost of annoying them.

  2. Jonatan Lundin 2013/12/12 at 05:27 #

    Hi Mark,
    Context is vital. I look at context perhaps from another angle. Users are seeking information, they display an information-seeking behavior (my research aims at describing and explaining this behavior where I focus on professional practitioners of industrial devices).

    When a user has found a page, s/he starts a relevance judgment process to evaluate the page it terms of its relevance to the perceived information need. Information science has struggled to define the relevance concept, but for sure it is very central. The information retrieval research field talks about pertinence which is slightly different from relevance.

    A page must therefore signal its context since it helps the user to judge the relevance. As technical communicators, you can use several ways to signal the context. To signal where a page is located in a table of content is a poor way of signal context (something we have discussed before). We both agree that subject matter classification, linking etc are much better.

    A powerful context is given by combining these two: Show the subject matter classification (as user evaluateable tags etc) and allow the user to navigate to related pages by selecting related subject matter classifications.

    • Mark Baker 2013/12/16 at 17:20 #

      Thanks for the comment Jonatan

      I find the most persuasive theory of user selection behavior to be information foraging. I like the way it emphasizes the ease of moving on to another information patch as a key factor in how long a reader will spend assessing a page. I also like the fact that information scent is not specific to any one characteristic of a page. It discourages us from thinking that any one technique is always correct or will always perform better than another.

      Classification is a good example of this. In some contexts and for some readers, it works very well. In other contexts and for other readers, it can be baffling and repelling. Classification only gives off a good information scent if the reader is already familiar with the classification scheme, so you always have to ask what classification scheme your intended audience knows before you commit to a classification as a means of navigation and context setting.

      This also reminds us that findability and context are not universals. The same information that smells sweet and familiar to one reader may smell rank and strange to another. One always has to ask, findable by whom?

  3. Michael 2014/01/17 at 10:41 #

    The level of context depends on the type of question and the skill level of the questioner.

    A lot of questions in application development are “what value does parameter x in function y take?” or “problem with program x and action y”.

    The reader can then interpret the answer quickly and leave the site. See http://stackoverflow.com/ for examples of these questions and the terseness of the “best” answer.

    Non-technical equivalents would be googling “where is x in city y?” or “what time does x close?”.

    These two user types differ considerably in the sophistication of the question, but both have a specific need and expect a specific response. They might stay on the destination site longer, but they did not intend to.

    • Mark Baker 2014/01/17 at 18:54 #

      Thanks for the comment, Michael,

      Indeed the level of context required does depend on the nature of the question. In some cases, the amount of context required takes up far more space than the answer itself. The answer might be “7”, but to establish what that value is an answer too might require a lot of text in order for the reader to be certain it is the answer they are looking for and to act on it with confidence. One often sees this in StackOverflow, where much of the discussion is sometimes devoted to clarifying the context while only a few words are necessary to convey the answer once the context has been established.