MFA: More F-ing Ads.

This article is from the Marketecture Newsletter. If you like it, please subscribe for free.

Forbes, the venerable business magazine known for inflating the egos of billionaires and helping PR agencies justify their retainers, is now also known for something a little bit more sinister. It seems that just under the surface of Forbes.com’s generally crappy user experience lay an even crappier one at www3.forbes.com (no longer live).

Of course I’m talking about the bombshell WSJ article based on the unstoppable force known as Adalytics (full report) . The research found that Forbes ran a parallel website with the same content but with dramatically higher ad load designed to burn through programmatic ad budgets. If you think this wasn’t that bad, they turned a single article into a slide show showing 201 ads (!).

I wrote a banger of a tweet on this, and I’m not above embedding myself:

What have we learned?

Nothing? That’s a pretty good answer. We have learned some things:

“MFA” and “Quality Publishers” are overlapping circles, unfortunately. Not every publisher, of course, but enough to make reputation alone an insufficient basis for trust. Which is a shame, but it is a fact.

Paying for traffic is a corollary to bad ad placements. Adalytics claims 70% of the traffic to the www3 site was from Outbrain, Taboola, and the like. As a general rule, mainstream advertisers should never want to see their ads delivered to paid traffic. I wonder if this is something a publisher or an SSP could guarantee based on referral strings? Interested in others’ POV on this.

The industry’s defense mechanisms are starting to take hold but remain insufficient. I say this because it is clear in this case that Forbes took pains to avoid being caught. According to Adalytics, the www3 domain was misrepresented the prebid configuration. They prevented users from directly navigating to this site unless they went to a detailed URL. They also removed the publisher-side tags from IAS, DV, and MOAT on the bogus site.

Where were the verifiers?

A question that came up pretty quickly after the report emerged was the role of the verifiers, namely Integral Ad Science and DoubleVerify . This is an interesting question since it points to a chasm between what these companies promise versus what they actually do.

Here are some of the reasons the verifiers didn’t notice anything wrong:

• They both only released MFA-detecting products recently
• The ads were viewable
• The ads were not (necessarily) being seen by bots
• The ads were not being mis-targeted
• The ads were not being shown to inappropriate content

And the main reason the verifiers should have noticed something is simply that it is what advertisers think they are paying for. If you ask advertisers whether they would expect the verifiers to catch this, my bet is it would be an overwhelming yes.

Specifically, though, is serving ads to www3.yoursite while auctioning off the ads as www.yoursite a form of invalid traffic (IVT)? The MRC says yes. In their IVT definition doc (pdf) they list “domain laundering and falsified domain / site location” as an example of IVT.

And IAS seems to think so as well according to this blog post (screenshot of the embarrassing part below).

SSPs and DSPs

Adalytics claims that four SSPs were transacting the www3 content and misidentifying the sub-domain as www based on settings of the server-side prebid server. Not clear if this was done by Forbes or a vendor and who was aware, but overall not great.

Going down the chain, Adalytics found buyers using The Trade Desk and other DSPs that claim to be MFA-free on the www3 site, and further that these buyers were buying through the misidentified paths from the four SSPs. Not great, Bob. We should install a big red phone on the desk of SSP CEOs that connects directly to the supply teams at TTD — saves time between clawbacks.

Here’s a great quote

While on the subject of low quality ad experiences, this week AdX rolled out the IAB’s recommendation on properly labeling video experiences. AdExchanger did a bang-up job explaining this, to which I don’t have much more to offer. Except this anonymous quote from a senior executive who wanted to remain anonymous because of their involvement in this issue:

“up until at least a year ago, prob 80% of what was bought as In-stream in programmatic was actually outstream. that collapse of supply will be like the 1973 oil embargo. it's gonna be quite a shakeup.”

Time to start drinking

Don’t despair! Like most of the problems in ad tech, the quality issue is actually kind of easy to solve if you just stop caring about getting a lot of cheap reach. That, and look at your log files.

This article is from the Marketecture Newsletter. If you like it, please subscribe for free.