One early morning in my student days, I had the munchies. It was near the end of the semester, so there wasn’t much in the cupboard. I poured a can of beans into a pan, and thought I’d spice them up with some curry powder. Then I thought “spicy beans — that’s almost Mexican” so put in some chocolate powder to get a kind of mole sauce effect. I put them on a low heat, then got distracted by something on the TV for half an hour. When I got back to my beans, they’d disintegrated into a stodgy mess. I managed to swallow one spoonful before giving it up as a bad job and going to bed hungry.
Lots of web content seems like that, especially technical content. Bits of information that were perfectly good ingredients on their own are mixed together so you’re not sure where one subject ends and another starts. Sometimes the information itself seems lost in a starchy mass of words.
A good stew, on the other hand, keeps distinct flavors in the chunks of vegetables or meat. The cook knows not to put everything in the pot at the same time, and that different ingredients require different treatment. For example, you should never put raw courgette/zucchini chunks straight into a broth. Squashes need heat to bring out the flavor, so you should sauté them first over a high flame.
Well-thought-out content sets are very much like a good stew. For sure, there’s a certain consistency throughout — the sauce that provides a clear theme, for example a set of content about a particular product, for a particular audience. But each chunk has a distinct role. Carrots bring sweetness; potatoes, substance. In the same way, carefully designed information is served in purposeful pages whose structure reflects their intent. Within a given subject area, for example, you may define one overview page and several more specific, deeper dives into aspects of the subject matter. This pattern is followed consistently across all subject areas, and templates can ensure a common structure to pages of each role.
These pages — the chunks in the stew — need to be substantial and satisfying. It’s not that you eat one and it’s enough. Quite often when looking for information or just idly browsing, you’ll read through quite a number of pages on a site. But it’s annoying and dissatisfying when the information is broken into tiny fragments (baby food!), perhaps so the site can show you more ads, or to work better with the authors’ content management system. Each page needs to answer a question or help you build up a mental picture. A page of content should have a clear role, and only when that role is fulfilled, should the page finish.
These roles don’t arise organically, but in response to rigorous examination. Occasionally, a worthwhile recipe is created by accident. More often, uncontrolled creativity results in a jumbled mess, like my mole curry split baked beans. Proper recipe development is the process of information architecture and taxonomy. Examine your information needs, question the purpose of all your current content, clearly define the concepts you’ll be writing about, and then agree on the rhetorical and technical structures you’ll use to organize those tasty chunks of information.
A related thought
Of course, the web is not just one dish but millions, prepared by a similar number of cooks. For organizations whose product is not their content, one might wonder whether it is worthwhile to manually prepare content still. It seems that for any question about a product, someone, somewhere has written about it. But the Internet is not an evenly distributed gruel. Rather, to borrow a phrase from David Weinberger, “The Internet is like most oatmeal: sticky and lumpy” (in Too Big to Know, pub. 2012.) Coverage is uneven, and some lumps, it must be said, are tastier than others.
There is a site called Allrecipes.com where the recipes come from site users, and the ones with the highest ratings are presented at the top. I can tell when someone has used one of these recipes, because the dishes always have too much bacon, cheese, sugar, or all three. There are two useful parallels we can draw with other content. One is that ratings, or usage analytics, don’t tell the whole story. For example, a page may be rated comparatively highly because the subject matter is inherently easier to understand than that of other pages. Or, it may have low ratings, not because the content is no good, but because users are coming to the page erroneously due to some quirk with keywords and the search function. For more thoughts on using analytics judiciously, see my slides on How just a little data analysis can improve your content.
The other lesson is that recipes, like most content, can be improved with restraint and some editorial control. There was a time when the web was almost a counter-culture phenomenon, and part of that was the breaking down of traditional sources of textual authority: the book author, the editor, and the publisher. The great thing about the web was that anyone could publish anything, and this has been taken by some to mean that as a matter of principle we should avoid the old controls on content. But in fact, like most innovations, the new ways of communicating have not replaced the old ones, and the old ones have grown to embrace the new technology. Long-form writing is back; thoughtful, edited content is still appreciated; and search engines and readers alike recognise that people who make products can still be considered authorities on those products.