No crime, No chime, without online shopping, you'd just be wasting time.

'Recommended For You': Something You'd Probably Not Find Yourself.

We know that companies use algorithms to target ads and recommend products they think you'd like. Sometimes when I can't find a product online that I want to buy, I search for it on Google a few times, and more often than not, I eventually get ads with links to buy it soon enough! Okay, but how do you contextualise this? Let's use a psychological research point of view to put this into perspective.

Imagine the field of social psychology, a body of research that tries to understand and predict human social interactions and behaviour. And even then, psychologists can only analyse behaviour in front of them. The catch here is that, for many of us, our online interactions make up a significant part of social interaction. Perhaps behind those screens, our actual attitudes come out. One of the most prominent issues that come in the way of research is social desirability biases, in the sense that one will not willingly admit undesirable characteristics. You can understand why there needs to be technological progress in data collection or any method that efficiently collects all possible information.

Customer segmentation is the popular term used for the ability to divide customer bases into smaller categories, with growing similarities among those people. For example, instead of advertising Olivia Rodrigo's music to anyone and everyone, target it to people who listen to music with similar tones, phrases and genres. While social psychologists use culture, personality, and such visible characteristics as predictors, data science uses your online activity. By using the internet as participants in research, there can be richer data available: Reddit channels as your focus group or several clicks to measure your variable of interest. This is the use case of data science analyses.

Customer segmentation can be in so many ways: demographic, geographic, technographic, psychographic, behavioural, etc., or even any combination of these. Netflix uses some behavioural segmentation, meaning that they could track what types of shows you click on, which descriptions you read, how much you watch something, which actors you like, who seems to be following what, etc. Then, after consolidating this information, you get titles recommended to you. Even within family accounts, each individual's profile looks different, which is the extent to which Netflix tries to personalise.

Maybe combining the two areas of knowledge can answer questions like: Is there a correlation between an individual's ideas and the consistent viewing of action titles on Netflix? Are your attribution and thinking styles related to the type of music you listen to?

Go back and think of any time an app has given you a recommendation and wonder if you can respond to these questions!

As always, the silver line is that you can navigate through the wide variety of options you are given, and data science makes your life easier! You reduce your screen time, don't get fatigued looking through the internet, save valuable time and money, and have relevant desired material at your fingertips.