Read the full post at emarketer.com
Authored by Ross Benes
Even when an advertiser’s programmatic bid fails to win an impression, they can still gather data from the auction to use across their campaigns. But doing so can be time-consuming and requires technical expertise, so many advertisers discard this data instead of utilizing it.
The amount of data that programmatic bidding creates has increased as header bidding has become more popular. Before header bidding’s rise, programmatic advertisers relied on a system called “waterfalling,” which sequentially passed bids from one ad exchange to the next. Header bidding allowed programmatic platforms to bid simultaneously on the same piece of inventory that was being offered across multiple exchanges. Three-fourths of the 1,000 most popular sites that sell programmatic ads use header bidding, according to Adzerk.
As the data that users generate has increased, advertisers have found themselves scrapping lots of data. In an August 2018 survey of 100 digital marketers worldwide conducted by Digital Element, 15% of respondents said they throw away at least half of their data.
Wesley Farris, director of partnerships at programmatic agency Digilant, spoke to eMarketer about how advertisers can make use of data exhaust.
What sort of data can a DSP store from a programmatic auction?
“Every time a DSP gets a bid request, there are a number of data variables that are passed on: the anonymous user ID, the domain, the time stamp, the location of the page, the creative size. There could be upward of 50-plus variables that are passed in the bid request, and the DSP stores that information in what I call ‘log files.'”
How can an advertiser use that information?
“You can use it for things like the customer journey. If you pull that information out, you can start to piece together a picture of how different channels and variables are affecting the campaign and put together that customer journey of how they engaged with your media. If you ingest that data across your search and social efforts, you can then more or less combine your search efforts with your social and display efforts and see how those are impacting one another.
“You can also use it for data science to find which variables within the bid stream are driving conversions or whatever key performance indicator [KPI] that’s being measured. You can use it for audience discovery, bucketing users by similar variables, and reaching them on a more granular level. The data is definitely used at the DSP level for optimizations and reporting insights.”