In the digital advertising ecosystem, effective targeting is crucial for the success of campaigns. The shift to first-party data and advanced AI technologies has revolutionised digital media buying. One of the highlighted approaches is lookalike modelling, or modelling based on shared characteristics using AI. This approach allows brands to target their advertising campaigns more precisely to the right audiences with a much higher likelihood of liking a brand. In this blog post, I will explain how digital media buying is done using first-party data and lookalike modelling using AI and show a practical example.
Segmentation starts with the collection of first-party data
Segmentation starts with the collection of first-party data about existing customers and website visitors. The first step in the process is a thorough review and collection of various types of information, including on-site behaviour, content preferences, past purchases and geographical location. This collected data is then processed and analysed using data management platforms to identify key patterns and common characteristics of different customer segments. This process results in clear and precise customer groups based on their specific needs, preferences and behaviours. This is the foundation for effective customer segmentation and identification of target audiences, which enables more targeted and impactful marketing communications.
Developing lookalike models using artificial intelligence
By using artificial intelligence to process the datasets collected in the DMP, we can efficiently develop lookalike models that identify potential customers who are demographically or behaviourally similar to existing customers or share their interests. Artificial intelligence allows the analysis and comparison of first-party data with other data available through online media or programmatic marketplaces. It identifies groups of individuals who share similar characteristics and media behaviours with existing customers. In this way, the identified target audience can be extended across media users, who can be reached with ads within the programmatic media ecosystem.
Real-time optimisation of advertising campaigns to maximise effectiveness
Using lookalike modelling and artificial intelligence, brands can effectively optimise their campaigns. They can adjust the messages, offers and even the platforms on which they advertise to better align with the profiles of their target audiences. They programmatically serve ads only to specific users, regardless of the medium where they are buying ad space. This allows for more precise targeting, more effective ad campaigns and better ROI. For example, Lisca wanted to increase online sales in the Croatian market. Using first-party data and AI lookalike modelling, they developed a strategy to identify and target new potential customers who are demographically and behaviourally similar to their existing customers. Implementing this strategy led to an eightfold increase in online sales and provided valuable insights into their target audience.
Progressive brands recognise the value of first-party data
The effectiveness of buying ad space in digital media using first-party data is confirmed by another practical example and several successful advertising campaigns. For example, Mömax doubled its sales by using first-party data to buy ad space in digital media. By combining iPROM Retail and a private DMP, Mömax effectively promoted its product range, increased sales and achieved a higher return on advertising investment compared to the previous year.
Microsoft wanted to boost direct sales of the new Microsoft Surface laptop through three Slovenian retailers. Using iPROM’s Private DMP platform, first-party data and advanced advertising solutions, the campaign for Microsoft Surface increased laptop sales by 20%.
Similarly, SKB banka improved the effectiveness of its digital advertising by using iPROM’s Private DMP to leverage first-party data about its digital audiences. By targeting ad space buying based on proprietary or first-party data, the bank increased direct response advertising by 72 percent and reduced the effective cost per user reach of the relevant audience (eCPC) by 25 percent.
Buying digital media based on first-party data and using lookalike modelling and artificial intelligence enables brands to execute more targeted and effective advertising campaigns. This not only significantly improves the effectiveness of advertising, but also helps brands better understand their audience and adapt their strategies according to their needs for long-term success in a competitive digital environment. At a time when agility and innovative technology use are paramount, the use of first-party data combined with lookalike modelling using artificial intelligence plays a fundamental role in developing advanced, forward-looking advertising strategies.