Finding patterns in the data

Learn about segmentation, profiling and data lookalikes

A big part of the way data is used by businesses that generate income through advertising involves finding patterns and making predictions. You might have heard of user segmentation and profiling, but what does all this mean for individuals and their personal data?

Segmentation and profiling

Let's take a look at a fictional example of what's known as segmentation or profiling. We'll look at a user who has a Schibsted account: Kari has told us she's a 24-year-old woman who lives in Oslo. Kari is OK with her data being used to show her relevant ads on Schibsted sites.



As Kari reads the news on Aftenposten, browses classified ads on FINN, and clicks various links on these Schibsted sites, her data profile begins to take shape: she enjoys hiking, traveling to sunny places and walking her dog. She likes the convenience of shopping online, so she sometimes clicks on ads for offers she likes the look of.


We can combine what we know from the information Kari has given us (her age, gender and the city she lives in) with what we've learned from her browsing activity to calculate that Kari fits into a user segment of young, physically active women who like online shopping and live in Oslo.


When we present Kari with an ad for a sale on hiking boots and she clicks on it, we can tell that we've given her a relevant ad. If she ignores all ads for makeup, because she doesn't really wear it, we learn that not all women in Kari's segment want to see that type of advertising. In this way, our understanding of her segment improves.


Data lookalikes

Now, let's take a look at something called data lookalikes with another example. Peter is a man in his mid-thirties who lives in Stockholm and is a total foodie. He loves going to new restaurants and he's always looking for ways he can improve his kitchen at home. He's not into social media and is a bit skeptical of data usage, so he tends to avoid logging in to sites unless he has to, but he accepts cookies on sites he wants to use.ts to use.


When Peter first goes to Aftonbladet to read the news, or goes to Blocket to look for new kitchen appliances, we don't know anything about him. But, as he clicks around, reads articles, types in search words and clicks on ads, we receive data that helps us learn what other content he might like. Though we don't know his gender or exact age, when we compare his activity to other similar data profiles, the patterns in his behaviour give us clues about the segments he might fit into.


Pretty soon it's clear that Peter likes food-related content, so he is matched with ads about restaurants in Stockholm and for local kitchenware suppliers that are having sales. We don't know his exact age, but the kinds of articles he reads on Aftonbladet match with other men in their thirties and forties, so we also match him with the kinds of ads we know those men tend to engage with.


A note on data security

Your personal data is safe with us. When reading through these individual examples, it might feel like there's someone sitting in front of a screen watching Kari's and Peter's every move online. But, when we say we match Kari or Peter with an ad, we're talking about an automated process.

We take the responsibility of storing and managing data very seriously. We work closely with our peers, data privacy experts and regulators to continuously review our security measures and practices to ensure that data is handled in line with your expectations as a user, and with the law.

To learn about the specific data retention periods and privacy measures taken by individual Schibsted websites or apps, visit the privacy section of that individual service.


Continue reading...