Five Unintended Consequences, Counter-intuitive Outcomes that Publishers May Not Have Thought Of

1. selling remnant inventory on exchanges didn't make publishers more money and it created the unintended consequence of buyers waiting till inventory became unsold, so they could buy it far more cheaply on exchanges (it's the same audience of the good publisher)

2. selling on viewable impressions meant publishers could get higher CPMs; but the counter-intuitive outcome was that only 40-70% of the inventory was viewable and the CPMs needed to be doubled just to break even (compared to doing nothing).

3. adding more ad units and creating more impressions were supposed to increase revenue; but instead, giving away inventory as "added value" actually devalued all inventory because it made the problem of unsold inventory even worse, and much of it was not deemed viewable anyway.

4. adding content-recommendation widgets to the site was supposed to increase pageviews and therefore ad revenue; but instead it allowed unsavory clickbait to sully good publishers' brands and worse, allow malvertising to easily hijack their users (without user interaction) and divert them to other unsavory or malware sites.

5. allowing 3rd party scripts on the site was supposed to help the publisher get more insights about their audience and charge more for that; but the unintended consequence was the publisher gave away their unique or proprietary audience; buyers could target audience members on other sites so they no longer needed to buy from the original publisher.

Things are not always what they seem. I know, having observed digital marketing since the very beginning and ad tech specifically over the recent handful of years. Most of the fraud goes on undetected and many participants are still high on the euphoria of big quantities (of impressions) and fast growth. So they overlook the unintended consequences of actions they took, thinking that those actions would lead to more revenues, when in fact it reduced revenue, increased cost, and increased risk (for them and for their audience).

I am not talking about the obvious, like ad fraud which diverted dollars into the pockets of bad guys. I am addressing the not-so-obvious unintended consequences and counter-intuitive outcomes that I've witnessed publishers suffer in recent years. Note that this does not affect all publishers equally; and some, not at all.

But hopefully, knowing what has happened to others, and what to look out for will help publishers be ready to take action or formulate strategies to deal with these scenarios if they do come up and impact them.

Let's examine five examples of this.
1. Selling remnant inventory on exchanges
Publishers thought they could make at least some money on their unsold or remnant inventory on exchanges and seeing whatever dollars they could get for it. But there's a few problems with that hypothesis. When publishers sell inventory on exchanges, the CPMs are way lower to begin with, compared to their own direct-sold inventory. On top of that, for every dollar sold, the publisher gets only 30 - 40 cents after all the middlemen's fees -- the so-called "ad tech tax." So the publisher literally gets pennies on the dollar for remnant inventory sold on exchanges. Jason Kint of the DCN has done research and written about this phenomenon; and other publishers have corroborated these figures.

Unintended consequences make this far worse for the publishers -- buyers waited, and did not buy direct, so they could buy it cheaper as remnant on exchanges.

If a buyer sees they can get inventory on exchanges for very cheap, compared to buying directly, what do you think they would do? More and more will buy the remnant on exchanges to get the low cost; some might even wait until the inventory goes unsold so that they can buy the same stuff for way cheaper on exchanges. So not only did the publisher not make much from selling anything on exchanges, they created an unintended behavior among buyers that caused even more inventory to go unsold -- they waited, and did not buy direct, so they could buy it cheaper as remnant. This also "creates price pressures to drive down overall CPMs, possibly affecting even direct."

Another unintended consequence is that bad guys can now easily spoof the good publishers' inventory on exchanges and steal ad dollars. They do this by passing the good publishers domain name -- e.g. -- in the bid request, so they can get bids for their ads. Note that the domain is falsified, but the sellerID is not; this way the bad guy still gets paid for the inventory, and the legitimate publisher is not, even though their domain name was used in the bid request. Basically bad guys have to pretend to be good domains; otherwise they would get no bids. But because most placement reports list domains and not the sellerIDs that got paid. So, the faked inventory is lumped in together in the same line item as the real inventory - listed as Buyers would have to reconcile it later by checking for the real publishers' sellerIDs, which they seldom do...

Read The Full Article


0 Comments Write your comment

    1. Loading...