Intent-based audience buying helps advertisers target audiences that are more likely to buy a specific product or service. Audience buying used to offer advertisers customer segments based on demographic information. Today, the practice relies on general web browsing data to infer consumer intent and optimize campaign targeting. This is because this type of 3rd-party data is easy to collect and buy.
The browsing-based method of audience buying uses algorithms to group online users with similar browsing behavior into customer segments, and then targets these segments with digital ads. To gather this information, ad trackers are placed on destination sites across the web. However, these trackers are typically only added to the front pages of a site, and never to the internal pages of a conversion funnel. The downside of this is that only general web browsing activity can be obtained and used to infer intent. There’s no way of differentiating information seekers from actual buyers.
For example, consumers that visited automotive websites would be grouped into the same customer segment as consumers that visited consumer review or auto price comparison websites. Similarly, consumers that browsed through multiple property listings, or just visited a real estate website, would be grouped into a segment with inferred interest in real estate.
This segmentation helps advertisers reach more relevant audiences. However, these audience segments are not as effective as they could be because they include window shoppers. In practice, the browsing behavior that’s captured for window shoppers is nearly identical to the browsing behavior for eventual buyers.
Conversion behavior to determine intent
To truly increase the effectiveness of audience buying so that advertisers will be able to reach individuals with intent to purchase, conversion behavior need to be used to create lookalike intent-based audiences. Every time a consumer fills out a form, buys a product, subscribes to a service or completes a measurable conversion event, they signal a deeper level of intent. Using conversion data in the calculation of intent essentially filters out the window shoppers.
Conversion data is much harder to get than general browsing data because few companies have it. Advertisers already struggle with tagging their own sites to measure conversions, and are unable to measure the conversions taking place on the sites of affiliates, not to mention competitors. However, Jumpshot has the 2nd-party data needed to create intent-based audiences.
Our data is based on the clickstream activity of our 100-million consumer panel. Clickstream analysis makes it possible to define an audience segment on past conversion behavior, path-to-purchase activity, and even shopping cart abandonments, and then create a lookalike audience based on unique behavioral trends that differentiate this segment from the general population.
Unique behavioral trends can be detected by analyzing the full online footprint of a segment of customers that have converted, and comparing their search, browsing and shopping behavior to the general population. Behavioral trends found within the segment’s clickstream are then used as triggers to identify a larger base of consumers that exhibit similar online behavior, but have yet to become converted customers. The resulting lookalike audience is much more targeted than the audiences currently available, as it is based on deeper levels of intent than general browsing behavior.
Bottom line: Audience buying has evolved from relying on general information to assumptions based on general browsing behavior. This was a good first step, but we’re not there yet. To truly optimize targeting to reach relevant prospects when they show interest in your products or services, (or even your competitor’s products and services), actual converting activity needs to be added to the equation.