Marketers should reduce the number of media sellers they work with, including supply-side platforms, rethink the use of private marketplaces and have direct contracts with all primary partners, according to the final version of a programmatic media supply chain study from the Association of National Advertisers.

This is the third wave of reports from the ANA on the subject since June and completes the study. The final report delves deeper into details that include how an overabundance of supply-side platforms (SSPs) can waste money, why some private marketplaces are little better than open exchanges because they mix in low-quality inventory, and which marketers and platforms did and didn’t provide data for the research.

Some of the industry’s biggest players, including Google, Amazon, The Trade Desk and Yahoo, did not participate, though Google data was available from agencies and advertisers, according to the report.

Prior releases by the ANA pointed out that marketers are wasting as much as 23% of their digital spending on the open web through duplicative services, buying on too many websites and pouring money into low-quality “made for advertising” (MFA) sites among other things. The ANA also reported in August the results of work by the investigative team at Kroll, which found fuller transparency was possible, but often at a high cost, while detailing questions marketers should ask agencies and others.

More news: 6 takeaways from ANA Masters of Marketing

Savings of $22 billion possible

In all, the final report concludes marketers could cut as much as $22 billion from the $88 billion spent annually on advertising across the open web by increasing efficiency and eliminating waste. Doing that requires using inclusion lists, as opposed to playing whack-a-mole with exclusion lists, to dramatically reduce the 44,000 sites used by marketers in the study. The report also recommends using an optimal mix of open and private marketplaces rather than simply leaning on the latter, having direct contracts with DSPs, SSPs, verification companies and others in the marketplace, reducing the number of SSPs used, and better managing measurement and fraud detection.

In all, the final report says 29 cents of every digital dollar in programmatic buying on the open web goes to DSP and SSP fees, and up to 35 cents to non-viewable, invalid traffic (e.g. bots and other fraud), non-measurable and low-quality MFA traffic.

Need to control SSPs

Among the deeper dives in the final ANA report is on managing SSPs, whose proliferation can reduce transparency into costs, increase complexity and result in marketers bidding against themselves.

The Trade Desk makes more than 350 SSPs available to advertisers, said Tom Triscari, CEO of consultancy Lemonade Projects and part of the team behind the study, during a preview of the final report during the ANA Masters of Marketing Conference in October. At the high end, one marketer participant in the study was working through 53 SSPs to buy digital ads.

The ideal number, according to the report, is in the five to seven range, which can deliver nearly 100% of all needed inventory. Being selective also can give marketers leverage to get SSPs to provide transparency into exactly where money goes in the supply chain. That was key to one marketer, which did not participate in the study, reducing its supply chain costs, according to an executive familiar with the matter.

But the report also cited Jounce Media, which found marketers need to reach $50 million to $100 million in annual spending before they get sufficient leverage to effectively negotiate directly with SSPs.

Problems with private marketplaces

While private marketplaces often are seen as preferable and more insulated against fraud than open ones, a strategy of using only PMPs can raise costs without necessarily guaranteeing better quality, according to the final ANA report.

“Some PMP inventory is high quality and may be worth the premium price,” according to the report. “However, the days of simply assuming all PMP inventory is worth the premium are behind us.”

The study found that 14% of PMP spending analyzed contained MFA inventory, similar to the 15% proportion in open exchanges. PMPs with more than 500 domains delivered quality similar to OMP deals but at double the cost.

The report cites input from Triscari comparing some PMPs with the collateralized debt obligations that led to the 2008-2009 financial crisis by combining a relatively small proportion of high-risk mortgage debt with mostly high-quality debt, resulting in a high overall credit rating that masked risk. Some PMPs pull in MFA and other low-quality inventory to reduce average CPMs for inventory that’s mostly high quality, according to the report.

“With proper checks, controls and optimizations in place, OMP inventory can drive comparable quality at a lower cost,” according to the report.

Participants and non-participants identified

The report also for the first time identifies participants in the study, including 20 of the 21 marketers, which included Nissan, State Farm, Walgreens, Molson Coors, Nascar, Discover, Dell, Kenvue, Kimberly-Clark Corp. and Mondelēz.

DSPs participating included Adobe Advertising and Adform, but not some of the biggest, such as Amazon, The Trade Desk and Yahoo, according to the report. While Google DV360 was not an active participant, the study was able to obtain Google log level data from agencies and advertisers.

SSPs participating included Magnite, Microsoft Advertising, OpenX and Index Exchange. FreeWheel, Google AdX, Pubmatic, TripleLift and Yahoo did not participate.

DoubleVerify, Integral Ad Science and Oracle (Moat) were participants as verification companies.