Moving Past Last Click Attribution

Market researchers everywhere agree: more and more advertising money is moving into programmatic media. Zenith Media predicts that 65% of all media will be traded programmatically in 2019, and what can be bought programmatically continues to grow, with expansions into Connected TV and audio. Yet, the amount of consumers that actually click on ads is decreasing, especially as additional ad formats to display banners gain traction with advertisers. A consumer can’t click on a Connected TV ad, and someone who is driving and listening to Spotify is not going to stop paying attention to the road to click on an ad they heard (hopefully).

In fact, according to comScore, only 16% of users click on ads, and a large percentage of clicks on mobile are accidental. Even more, this focus on clicks leads to more fraudulent traffic for programmatic campaigns as AI will optimize to a click-heavy site that might be fraudulent or not a positive consumer experience.

Last click attribution tends to favor search engine marketing, as this is the end of the marketing funnel. The consumer has been educated about the product through other channels, and they search with the intent to buy. With a last click model, this means other channels of media get much less credit than they deserve, and appear to perform worse in terms of revenue generated.

Dynamic creative helps programmatic display advertising out in terms of clicks, as it rotates through previous items viewed by the consumer, which make the ad more relevant and likely to result in a click, but it still is not on the same level as search engine marketing and will perform worse in terms of direct response.

As our industry moves more into less click-heavy media, and consumers are clicking on display ads less and less, we need a better way of measurement than simply last click attribution. There are several measurement partners in the advertising ecosystem that work to help break down how much credit each channel gets. C3, Conversion Logic, and many more offer distinct ways to attribute revenue and conversions across channels using algorithms and AI.

There are other ways to value metrics other than a last click attribution. An advertiser can discount post view conversions or decrease the post view window to less than a post click attribution window if they are looking at reporting from the platform they are serving from.

Decreasing the attribution window allows marketers to take the conversions that are closest to the viewing of the advertisement and attribute those to the campaign, which is a more accurate method than discounting the conversions. Another method is an in-depth analysis of the marketing mix of a brand, which enables advertisers to determine how much revenue is truly able to be attributed to certain channels. This can lead to the creation of multipliers that will allow a marketer to get a proxy amount for revenue driven by channel using this marketing mix modeling.

There are also other attribution models for marketers to choose from. One of them, time decay attribution, gives each interaction in the marketing funnel weight, but as time goes on in the consumer’s journey, the earlier interactions are weighted less and the last interaction is weighted the heaviest. Another model, position based attribution, assigns the most weight to the first and last interactions a user had, and less to the middle interactions, allowing the most attribution to be given to the introduction to the brand and the ad that drove them to convert.

Our industry is rapidly changing as new forms of media emerge daily, and measurement and attribution must evolve with it. Programmatic display, video, audio, and other channels are growing, and last click attribution will stunt their growth for direct-response focused brands unless a more fair attribution model is adopted.