Skimlinks captures 15 billion shopping intent signals from across its network of 1.5 million publisher domains on a daily basis. We apply machine learning to this colossal amount of data, enriched by our product intelligence layer - which understands the taxonomy and metadata of 100 million product references and links - to build high-converting audience segments based on the products and brands that users are likely to purchase. Agencies and brands can access these segments via a number of programmatic platforms such as DSPs or DMPs to run more effective display, social and video campaigns.
The patterns we identify are observed across multiple publishers and merchants
The depth and breadth of data and patterns observed across the network allows Skimlinks to more accurately predict purchase behaviors. For example, the combination of one user visiting ten pages across five different websites might be the pattern our machine learning algorithms determine indicates an interest in buying a high-end handbag in the next week.
Our audiences do not exist in the raw data of any single publisher
We have a rule of 50. Which means that segments can only be created where data is sourced from at least 50 publisher websites. This means that, on average, any one publisher contributes less than 2% of the signal information needed to classify a user.
We classify each user according to 20,000 attributes. These attributes are created by a process called ‘machine learning’. Machine learning is where powerful clusters of computers read through petabytes of information to discover patterns and important combinations. Each user is scored 0-1000 based on the likelihood of having each of the attributes.
It is this processing that makes Audiences by Skimlinks different from other publisher data management platforms. We combine signals from the entire retail ecosystem to identify patterns not visible when looking at a single website. Even the largest websites in the world do not have enough information to make accurate conclusions about the browsing and shopping behaviors of all their users.