Content vs. Algorithms - An Experiment

The conversation today among many digital strategists and social media managers is the impact of social media algorithms on the performance/visibility of posts. As someone who has been on almost every social media platform since the beginning (except MySpace - that was before my time), I have seen and studied the evolution of social media in far greater detail than your typical media consumer or "seasoned" PR pro.

This past month, our agency encountered an interesting situation with a client. Following a huge industry trade show, the niche trade media we aggregate content from for our social media accounts went dark for several weeks. Most of the content being posted was sponsored or mentioned competitors and thus did not fit into our social media content guidelines for the client.

So what do you do when you need to write a full month of content for approval without enough recent content to fill the days? We did an experiment.

Our solution was to take a deep dive into the content that we wrote for our client over the past six months to see which posts performed best on Twitter and Facebook (our two key platforms) and used those posts to populate our next month's content calendar. We called this the "greatest of all time" or "GOAT" content.

With the constantly changing algorithms on social, it's difficult to tell if our GOAT content performed well because of the time and day it was posted or if it was because of the nature of the content. We took the first variable out of the equation by posting at our normal times, 10 AM-3 PM Monday-Saturday; but we were exclusively posting content that we already knew was popular with our audiences.

I would estimate that 60 percent of people reading this post will be shocked at the results of the experiment while only 40 percent will expect what our test demonstrated. So far this month - with just under two weeks left - the trend is obvious, but at the formal end of the month, I will disclose the results.

In the meantime, please comment below which statement you think best describes the outcome of our GOAT experiment:

A. Content performed consistently well on all days, at all times.

B. Content performed comparably to the previous month, heavily dependent on the day and time of the post.