ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

The worst deals in digital advertising

The worst deals in digital advertising Andrew Ettinger

As a junior media negotiator, I tended to think of media buying as a zero-sum game; one side had to win and the other had to lose. As a more seasoned professional though, I have learned that the best deals in digital advertising are win-win scenarios. Both sides, both the seller and buyer, come out ahead.

However, there are some deals in advertising that are just terrible. Some are bad for one side, while other deals are actually bad for both sides. A terrible deal is more than just a high CPM. Rather, a bad deal in digital advertising is one where either side feels they did not get what they wanted and, quite possibly, paid a lot for that inconvenience.

The worst deals in digital advertising

The worst deals tend to be overly complex; one side does not really understand what they agreed to. While some complicated deals make for great ad campaigns, others just become hopelessly mired in confusion. The worst deals are confusing, costly, and complicated. Understanding what you bought, how to measure it, and what the data mean is the key to avoiding bad advertising deals.


Blind ad networks are a bad deal. In today's media savvy marketplace, not knowing where you are advertising is a bad idea. There is so much bot fraud that placing your bets with unknown websites is surely bound to lead to trouble. Moreover, today's buying world includes fully transparent demand side platforms (DSPs). These are platforms that allow buyers to understand exactly where their ads are running. Compare this to five years ago when ad networks ruled but no one knew where their ads ran. Some networks would show a list of affiliate sites, but many did not. Buyers assumed their ads would run on major sites and not the smaller, unknown ones. Unfortunately this was the power of suggestion at its worst. Usually the inverse occurred and ads ran on the bottom 10,000 sites rather than the top 100.

Eventually DSPs evolved and buyers got smart. Now buyers can bid on the exact inventory that they want and institute black lists to exclude bad sites, sites with subpar content, or ones that are known for fraudulent activity. Moreover, advertisers can apply different data sets (first party, third party) in real time and monitor the results. Programmatic DSPs have trumped blind networks as a solution to many advertisers' planning needs.


Recently I downloaded a new game called "The Line" on my iPhone. The game is fun and, best of all, free. At first I couldn't figure out how it supported itself, how it made money. Soon I realized that my thumb, acting as the joystick at the bottom of the screen, was actually blocking the ads. If you are an iAd advertiser, this is a bad deal. Advertising in app, in this particular game, is hard to do well. Ads that support free games are often overlooked because people are busy playing the game. Here the ad is not very intrusive to the game experience mostly because it cannot really be seen. In the best of circumstances, advertising in game is difficult because the ads must be complementary to the experience. Integrating product into the game is far better than intrusive ads that stick out like a sore thumb or, even worse, are covered over by a sore thumb.

Viewability issues extend to desktop advertising. Make sure you know where your ads are running and that they can be seen. Few things are worse in advertising than a great ad that can't be seen. If an ad is served below the fold, did it really get served? Of course it did, but no one saw it and you still get the privilege of paying for it! Ads below the fold should be much cheaper than above the fold ads because there is much less likelihood that someone will see them.

Use viewability tools like Moat or DoubleVerify to understand how much of your advertising is visible. Viewable impressions will eventually be the common currency amongst buyers and sellers. According to the IAB, an ad is visible only if 50 percent or more of the pixels are in view for a minimum of one second. Think about it; only half of the ad must be in view for just a second to be technically visible. Given that many ads have 15 seconds of animation, one second is not a whole lot of time. Even if your ad is technically visible, was it really viewed?

Consider this: One major CPG has an "any second" rule about creative. For any second that a creative might be viewed, you must be able to identify the message of the ad. For instance, an ad for crackers will always feature the signature red color of the box and pictures that show how crunchy the cracker is. Here, viewability and people's attention spans are factored in to the very nature of the creative execution.


Sponsoring great content is a tricky thing. Great content is key to garnering people's attention, which translates to ad impressions. After all, people are looking on the web for interesting things to read, for content to consume. However, depending on the campaign goals, it is possible to overpay for great content. Simple regression analysis will help a planner determine the maximum CPM that the campaign can support. If a campaign has an $8 CPA metric, it could be relatively easy to figure out what the max CPM should be given expected CTRs and historical conversions. A bad deal would be sponsoring great content at a price that isn't sustainable. There is a fundamental tension here -- content needs to be supported and ad campaigns typically want efficient media. Being around great content is only so good until a point. Then it's not good at all, especially if it isn't bringing the return that you need.

On the other hand, poor content is even worse. That is content with which no one engages. Bad content is a scourge to the online advertiser. Pay too much and the back end metrics don't make sense. Pay too little and wind up on sites that will hurt your brand image or, equally as bad, are scanned only by bots.

The most important thing to do with content sponsorship is understand your limitations. Know how much you can afford. Have a candid conversation with the other side over the phone, either seller or buyer. Some content just can't be sold for less than the asking price because it is so expensive to produce or attracts such an important audience. Understanding why, the rationale behind the price, will immeasurably help with your decision. If the content really is that great and you simply must have it measurable (ROI be damned), then at least you have a solid rationale to justify your decision.


Recently I received an interesting proposal from a really forward thinking website. Their proposal was super innovative. Unfortunately it was so complex that no one could easily or wholly explain it to me! Such deals are bad for both advertisers and publishers because they make expectations impossible to manage.

Both sides, buyer and seller, need to be clear about the deliverables. Without understanding those details, it would be impossible for me to sell it through to my client. First to market opportunities are great but can come with risk. Since no one has ever done them before, they are by nature daring. This is a good thing; it can set your client apart and stand out from the clutter. However, if I can't explain it to my client, then it's essentially worthless. Moreover, as the advertiser's representative, it is the buyer's job to hedge any risk in the marketplace. Sometimes this is financial risk (bad economic deals) and sometimes it is content driven. Every deal has to be sold twice, once to the buyer and once to the payer (the client). A bad deal is one where the buyer doesn't really understand the proposal.

Demographic targeting

Nielsen Online Campaign Ratings (OCR) is an interesting new model for advertisers. It is yet again changing the face of digital advertising and leading us to more precise targeting. Nielsen's OCR and comScore's Validated Campaign Essentials (VCE) allow advertisers to go beyond behavioral clues and measure demographic information. Moreover, it allows smart advertisers to pay only if the person who saw the ad fits the age and gender of the target. Essentially, old school advertising metrics like GRPs and demographic ratings have been introduced into this new medium. 

This is a great deal for advertisers but potentially a bad one for publishers. Publishers could lose money based on mismatched demographic targeting. One advantage for all parties is that if pubs can get better at targeting, then more TV dollars will go digital. Digital is rightly held to a higher standard than TV, and if online can prove it's reaching a more pure target, then naturally more dollars will follow. However, it's not always win-win for both sides. A bad targeting scheme could end up losing money for the publishers. Online advertising is still not perfectly precise and may equal a financial loss if the publisher isn't careful. Online ratings hold publishers accountable, which is a good thing. This is great for advertisers since this guarantees they will reach their target audience.

Andrew Ettinger works in advertising in New York City.

On Twitter? Follow iMedia Connection at @iMediaTweet.

"Businessman with thumb down" image via Shutterstock.


to leave comments.

Commenter: Lynn Salton

2014, July 31

Treating media buying / selling as a zero sum game is a fast track to a brief career. I have always said that advertising and promotions, should be a win-win-win scenario. It should be a win for the buyer and the seller but it should also be a win for the intended target audience too. If any of the three stakeholders lose... a repeat transaction is unlikely.