Is there such a thing as good MFA?
After being demonized in the ANA’s transparency report last summer – which found that 15% of global ad spend goes to made-for-advertising (MFA) sites, often unintentionally – agencies and ad tech vendors immediately started launching anti-MFA solutions.
The MFA label connotes spammy sites with undesirable audiences that many advertisers want to avoid.
But that’s not necessarily the case for all buyers, said Jack Smith, chief innovation officer at DoubleVerify, which added a feature to its platform on Tuesday that classifies MFA sites based on a tiered system.
Just because a website gets labeled as MFA due to a heavy ad load or a high percentage of paid traffic, Smith said, doesn’t automatically mean the content is junk, its audience is worthless or that it doesn’t perform depending on an advertiser’s goals.
A tiered approach
Rather than outright blocking any website that could fall under the MFA umbrella, DV uses its existing publisher integrations to determine – through pre-bid and post-bid measurement – whether a site should be classified into an MFA tier.
It then allows advertisers to filter out certain impressions before they bid on them, said CEO Mark Zagorski.
DV looks for all the classic markers of MFA inventory, including a high ad density relative to on-page content, a high dependence on paid traffic with little to no organic traffic, low ad viewability compared to industry benchmarks and a tendency to reproduce content across pages. These criteria are largely in line with those used by the ANA, Jounce Media and others.
But DV also weighs other MFA markers, such as overuse of sticky video ads, site designs that rely on endless scrolling and similar tricks to refresh ad impressions as well as an overreliance on clickbait, AI-generated content or content that doesn’t fit with a site’s overarching theme, Smith said.
DV’s solution categorizes sites as having either a high, medium or low rate of MFA characteristics. (Think of it as the MFA version of “The Good, the Bad and the Ugly.”)
DV uses a mix of AI and human review to place sites into these categories. As a general rule, if a site meets seven distinct MFA criteria, it’s placed in the MFA High bucket. Five criteria gets a site classified as Medium and under five puts it in the Low bucket. However, DV’s criteria and classification methods are still evolving.
A beta version of the MFA classification feature first launched in September. The updated version, now generally available, includes feedback from brands, many of which requested a tiered approach.
“Large, sophisticated advertisers want a lot of control and a lot of nuance,” Smith said.
To that end, DV also introduced features that more closely examine the relationship between the ad experience and the on-page content and differences between desktop and mobile versions of pages.
Nuance needed
An appreciation of nuance distinguishes DV’s approach from other MFA solutions on the market, according to Smith.
For example, The Trade Desk’s DSP reportedly excludes MFA sites by default. Meanwhile, SSPs including Kargo, Sharethrough and PubMatic have been launching supposedly MFA-free private marketplaces or working with partners like Jounce to declare their entire publisher networks clear of MFA.
But brands deserve the ability to tailor their ad buys, Zagorski said. It’s up to brands to use the tools DV provides to block whatever inventory they deem unsuitable, he said.
Of course, there are some things almost no advertiser wants, such as running on sites that buy bot traffic. DV’s platform already excludes all bot traffic from ad buys by default, Smith said.
However, many MFA websites also pay to attract real human visitors. A paid human impression can still be valuable if it drives performance for the advertiser, he said, so it doesn’t make sense to block all paid traffic instead of blocking only invalid bot traffic.
DV also doesn’t categorize entire sites as MFA. For example, a site’s homepage ad experience can differ greatly from the experience on individual pages, which is why DV’s MFA solution weighs different site sections individually. Advertisers can decide to run ads in certain sections of a site while excluding others on that same site. That way, they don’t miss out on potentially valuable impressions.
“We’re trying to get away from these blunt-force tools that are circa 2010,” Smith said, “and modernize to the level of nuance that most advertisers want.”
