It's not really the systems or the artificial intelligence supporting the bidding, buying and serving in question, but the data pushed through the system. 

Dirty data is one of the big challenges that companies have run up against since starting to work with multiple companies to serve one ad.

Bob Liodice, ANA CEO, wrote in the summary that most of the challenges stem from three data-related themes.

He said marketers are not fully skilled in optimizing the management of their data, they are not able to fully retrieve their brand performance data because of unfortunate contractual limitations with their respective partners in the digital media supply chain, and they suffer from an imbalance of available data in the supply chain.

All this leaves marketers with sub-optimal information to make the most effective marketing decisions for their brands.

The average campaign among study participants runs on 44,000 websites, although 63% of impressions came from the 500 top websites, by volume of impressions.

In 2018, automaker General Motors used inclusion lists to run ads on 4,000 and 15,000 websites. There was no negative impact on performance when compared with campaigns run on 800,000 websites.

The ANA study closely tracked the flow of dollars through the programmatic supply chain and found that only $0.36 of every dollar that enters a demand-side platform (DSP) reaches a consumer.

Where does it go?

• Around $0.29 goes toward fees to ad-tech intermediaries
• Another $0.35 goes to low-quality media, including invalid traffic (IVT) and made-for-advertising (MFA), non-viewable and non-measurable inventory
• While most supply-side platforms (SSPs) took less than 5% of the ad spend, one SSP in the study took close to 45 percent of media spend