Stop Talking About “The Algorithm” If You Didn’t Write the Code
You don’t know what “the algorithm” actually does, so stop telling people how to beat it
Many people seem to believe that social media websites are powered by a magical, super-special file called The Algorithm™, as if only a single, complicated equation in a heavily-guarded, secret document determines whether your posts get the organic reach you desire.
- Your latest post didn’t reach enough people? It must be The Algorithm.
- Your YouTube video didn’t go viral this time? Must be The Algorithm again.
- “Comment on this post so the Facebook algorithm knows to show my posts in your feed!”
- You notice some changes to a social media website’s user interface, so they must have also changed The Algorithm.
People seem to think algorithms are something they’re not.
A+B=C is an algorithm. An algorithm is just a set of rules (or a process) that is followed when performing calculations or other operations. An algorithm is not a magical, special file that software engineers maliciously change to keep you on your toes. It’s not a deeply complicated equation that singlehandedly determines your fate on social media.
Consider a visual example instead of a mathematical example. This is a diagnostic algorithm used to determine next steps after identifying iron deficiency anemia. This algorithm has outputs that vary depending on a chain of inputs. Software algorithms are similar.
“The Algorithm” that people obsessively hypothesize about is not a single algorithm, but many algorithms. Each one of a platform’s algorithms can range from a simple equation (if X does Y, show Z), to a small set of rules that are triggered by user interactions, to an extensive set of procedures that are triggered by an infinite number of potential inputs.
Changes to “The Algorithm” could be minor (e.g. if X does Y, show Z1 and Z3, but not Z2), or more complicated (e.g. change the numerical weight assigned to variables A, B, C, D, and E and then take action F, G, H, I, or J depending on the aggregated results).
Here is another example of an algorithm:
This should go without saying, but unless you wrote the code (or saw the code with your own eyes and it was beautifully commented and you interpreted it correctly), you don’t actually know what the algorithms in a company’s code base do. Period. End of story. You do not know. Neither do I.
People who wouldn’t be able to identify an algorithm if one slapped them across the face should stop telling people how to “beat” a platform’s algorithms. Even those who write algorithms all day long shouldn’t tell people how to “beat” algorithms that they didn’t write and haven’t studied. It’s counterproductive and it’s not rooted in evidence.
If there were evidence to support any of these “beat the algorithm” claims, each of the claims would include relevant citations.
The truth is that we do not know how a given platform determines which posts are seen on its homepage, or how a user’s interactions affect the content they are presented with. What we do know, however, is that it’s not a single algorithm, and that it’s not something that you’re going to “figure out” and “beat”.
Instead of trying to figure out how to “beat” algorithms (which, by definition, aren’t written to be beat), work on producing high-quality content. If “The Algorithm” functions as intended, it will do what the platform wants it to do, which might not be the same as what you want it to do. Sorry. If you don’t like the output, nobody is stopping you from changing the parts of the input that are within your control and seeing what you end up with — but don’t tell people that you’ve figured out how to “beat” a platform’s algorithm unless you’re citing evidence to back up your claims.