Beating Algorithms Is A Literal Game
Rob Davison 04 November, 2016 at 12:11
It’s no secret that our digital experiences are defined by algorithms. From search results and Facebook newsfeeds to YouTube video suggestions and junk mail folders, algorithmic codes elevate, deprecate, coordinate and regulate content for us everyday. We rarely know the inner workings of these proprietary tools, the details of which are protected by their owners with covertness to rival that of the Knights Templar. They are prized intellectual property; the better the algorithm, the more relevant the digital experience. Or, at least, so the thinking goes.
Even though the collecting, sorting, and collating of content is driven by the algorithm, the process of surfacing content to a user also relies upon the quality of care put into crafting the content itself. An asset designed to be found will perform infinitely better than one that relies upon discovery by happenstance.
Content marketers are left to deal with a simple challenge: how can one create content that stands a better chance of performing in our algorithmic world? For starters, we need to think – and create – literally.
Many marketers are first introduced to algorithms by search engine optimization and the basic SEO rules of creating literal matches. If you want to rank well in the Google algorithm for the phrase “maroon stiletto shoe” don’t title your content “red high heels.” Algorithms love literal connections. We call this Vernacular Planning. If one writes, speaks and designs in the vernacular of the user, the content will automatically have a better chance at connecting.
A quick spin through the Internet will reveal how few marketers actually make time for Vernacular Planning, despite it being the lowest of low hanging fruit. It should be considered table stakes as it does much more than help in search: Facebook, Instagram and all of the other social content networks use similar algorithms to decide what content surfaces in someone’s feed.
In simpler times, preparing content would have meant optimizing copy and writing basic metadata (probably nothing more than a title, description and tags for a video or image). It’s not so easy today. Metadata (text called out in specific code to alert the algorithm of things to look out for) can include everything from technical data about the formatting of the asset itself to richer and more descriptive fields describing the content of the asset. Time stamps, location data, tagged friends and other deep fields compliment the more direct social network aspects of who you like, follow or subscribe to.
How are these literal connections found? Some are common sense, while others lurk in the digital footprints we leave behind. Ogilvy has developed an audience intelligence service, which looks under the surface to connect words and actions to user intent. The result is a data matrix of vernacular habits, engagement behaviors and community connectivity that we use to develop a roadmap to literal relevance.
At this point you might be asking, “can’t I just use paid media to circumvent all of this?” Yes and no. You can implement paid search or paid social tactics to increase reach, but if you want to do so efficiently then the very same data that makes your organic content more valuable also makes your paid spend work harder. Paid media is not a way to circumvent the algorithms, though it will help your content become more visible within the rules of the sites you participate in.
Despite our industry’s penchant to try to “beat” algorithms, we never really achieve that. Success comes from preparing your content to do as well as it possibly can within the confines of an algorithm. All the search engines and social content sites periodically update their algorithms, so even if you feel that you have mastered one or two, that sense of accomplishment is fleeting. The only guaranteed path to the top is through the data mining and research I have described. If you get that right, the rest will take care of itself.