Views on Instagram stories used to be predictable and recently they’ve become much more dynamic/volatile with the number of story views bouncing all over the place, which is causing people and companies to question what they’re doing wrong.
However, with volatility we eventually find stability, and see where things are headed. For the most part it’s not companies doing something wrong, but instead them freaking out about algorithm changes they have no control over, do not fully understand the impact of, and need guidance to put them at ease while planning for what’s next.
Looking at Instagram stories, the view count flows in waves almost as if Instagram is running tests to push stories to your followers at certain intervals to weigh engagement. And from an AI perspective, we see how the feedback from such engagement (training deep learning models) helps with curating the best stories in the future.
If you landed on this post and read up until now you’ve likely seen the conversations brewing around inauthentic behaviors/content leading to a drop in Instagram story views, wild ideas about new algorithms being rolled out, and the list goes on.
But what we’re seeing in the real world is a scientific approach to testing theories on behaviors then learning and building from the insights achieved to produce the best user experiences possible.
We recently ran tests across a variety of enterprise companies we’re guiding and found patterns that are ingenious on Instagram’s part. Kudos to the team for taking this data-centric approach.
Similarly, this relates to Google’s recent focus on improving search experiences by answering questions people have on a large scale, the newfound value of product reviews that impacted recent algorithm updates, and so much more.
Like a seesaw, the more weight/strength you achieve the more the universe tips in your favor. We’ll be sharing more of our research soon.