Beat the algorithm

It’s election season and impossible to escape round-the-clock coverage of everything the candidates do. If you’ve already decided on your choice for November, you might prefer to consume your content from a partisan source. Most online platforms claim neutrality and an unbiased presentation of content…except for the fact that your own actions (e.g. likes, shares, click-throughs, time spent) signal what you prefer, which results in a more relevant – and biased – feed. To see just how stark of a difference this can become, visit Blue Feed, Red Feed.

A few years ago while doing research for a client, I found some strangely dissonant reviews on TripAdvisor, like this:

TripAdvisor review

The reviewer had great things to say about the hotel and rated all six elements as “excellent,” yet the overall rating is one circle, which means “terrible.” If you look at the other one circle ratings, you can see language that’s more expected: “disappointing,” “ridiculously bad service,” “horrible experience.” If you want to understand the worst about this property, you’re going to see this review and maybe write it off, but maybe think twice about the others.

When individuals share content on social media, Facebook in particular, it’s filtered for relevance. When people share the latest political update in support of their candidate or in rejection of the other, the posts are always going to be “preaching to the choir.” Campaigns woke up to the power of digital and social media during the 2004 election and have refined strategies since then to fundraise and get out the vote.

So after 12 years, why are we still seeing what’s expected, rather than operatives who can game the system effectively? Where are the tactics like that one circle review that shows up where you’d least expect it – and where it’s not supposed to be? Maybe Facebook would simply filter out dissonant content as clickbait and does already. I try to follow a wide sample of sources, but can see that filters still limit discourse. Maybe there is no way to beat the algorithm. We helped create it and now it’s in control.