Advanced ad targeting

How advertisers reach specific groups of users in Schibsted products and services

All advertising seeks to target the right products to the right audience, but most online ads aren’t matched to users as individuals; rather, they are matched to categories of likely interests or possible demographic characteristics.

Demographic Targeting

Age and gender targeting is made possible based on the information users give us, like when they create a Schibsted account, for example. Schibsted uses automated systems to analyse users’ behaviour against groups with certain demographic characteristics using techniques such as algorithms.

Algorithms extend the reach of online advertising campaigns that are more likely to appeal to specific demographic audiences. The automated algorithms create audiences similar to other Schibsted account users that have declared a specific gender and/or age-range in their profile. The similarity is based on user behaviour on Schibsted websites and apps during the last 28 days.

The most important key events reviewed between users with specific demographic characteristics and audiences 

Schibsted has no provided characteristics for are:

  • the parts of a Schibsted website or app users visited,
  • the keywords users typed in for search queries,
  • technical specifications like the type of device type and operating system.

Interest-based targeting

Interest-based targeting means the collection of online interaction data from a particular device or logged-in user over time and across multiple web domains or mobile apps. Schibsted collects data solely on Schibsted-owned or -managed websites and apps and then delivers advertising still only within those services. We do not purchase data or audiences from third parties in order to enrich our interest groups.

We believe our users are more likely to be interested in ads that reflect more recent visits and activity on our websites and apps. For this reason, interest categories are a rolling average of a users’ interest. This means they change over time based on a person’s most recent activity. Interest categories look back 28 days into the past. A user’s connection to an interest category decays over time, meaning that with each day that goes by, the interest category will slowly reduce its relevance for the user if no further activity refreshes the interest.

Categories fall across a broad range of topics such as travel, fashion, home remodeling or sporting interests; we don’t use categories that only include a small number of people.

Only a few key events are reviewed for these audiences:

  • Schibsted website or app sections visited,
  • the type of Schibsted article or classified ad visited,
  • the frequency of visits to a Schibsted website or app,
  • the keywords used for search queries on Schibsted websites or apps

When a certain threshold is reached, the user becomes eligible to receive ads that are associated with those interest categories. As the user’s activities with Schibsted services increases, the advertising systems should improve their ability to predict that user’s interests and enhance the relevance of the advertising he or she sees.

Apart from standard interest categories maintained by Schibsted, advertisers might ask to create custom interest categories. For example, they may wish to show ads only to an audience that has visited a specific section on a Schibsted site or has searched for specific keywords. This targeting is done by Schibsted, with no online activity history or interest category details being passed on to the advertisers.

Likely interest/prospect targeting

Audiences with patterns of similarities are likely to exhibit similar interests or buying behaviour. Predictive or prospect targeting uses techniques, such as algorithms, to create a large audience based on a smaller one with similar attributes.

The algorithms analyse the composition and characteristics of a core audience we would like to reach to find other users that show similar attributes or behaviours.

The most important key events reviewed between users with specific characteristics and predictive/prospect audiences are:

  • Profile account information such as postcode, gender or age,
  • Schibsted website or app sections visited,
  • keywords used for search queries on our services,
  • technical specifications such as device type or operating system

Matched audiences

Advertisers may create customised audiences based on their customer list of email addresses and/or phone numbers. In such cases, Schibsted will use automated technical platforms to match these customer lists with the email addresses or phone numbers in Schibsted accounts. If there’s a match, the technical platform will add the corresponding Schibsted account to the custom audience. The custom audience will be used solely for delivering advertising on behalf of the advertiser. Schibsted’s terms require advertisers to ensure that they have a lawful basis to upload and use the data you provided them in connection with matched audiences. We don’t use custom audiences for any other purpose.

 

Continue reading...