Viewability is Important


Reading AdExchanger some days just aggravates you. I have read articles where people talk about how advertisers shouldn’t worry about viewability if clickthrough is good. I have read articles about how frequency combined with viewability is really important, so viewability alone is a red herring.

This is a red herring!

Let’s be real. Two facts are true:

  1. Viewability is hard. The tech is wonky: It frequently doesn’t work, it isn’t applicable across every unit and every buy, it can be expensive because standards are weak to non-existent. When people ask me to implement some random viewability thing (because there are like 10), it aggravates me. Huge testing overhead, specious outcomes, impacts on deliverability, and impact on margins. Also, let me say it again: Specious outcomes. False positives are ruinous to publishers, yet are all too common.
  2. Viewability is important to advertisers. To imply that it is okay if the ad is not seen by a consumer is to tell an advertiser you like it when they waste money. Consumers have to see the ad for it to matter. I have seen the studies where multi-modal attribution shows that below the fold ads that are never seen somehow affect performance and I have to call bullshit on that. IT WAS NEVER SEEN. That was probably not a two-way test where they look to see if it didn’t affect performance. That is just truth speaking to data.

Strong clickthroughs without viewability is bots.

High frequency without viewability is waste.

Jay Friedman suggests that publishers move to rotating ad units that are viewable. Hate it. Hate it. Hate it.

I don’t want to weigh old school metaphors too heavily in the future of tech, but the current model is a newspaper metaphor. Ads are balanced around the content. If ads rotate, that is materially different than what the advertiser expects. Frankly, this is no different than viewability: We expect our ad to be viewed by the consumer while that consumer is on that page (or part of that page). Rotating units is another gimmick that publishers deploy to increase the amount of inventory they can sell while decreasing the quality of the product. If the New York Times introduces 20 second rotations, then the LA Times introduces 15. Then bad sites introduce 10 second rotation and all of the sudden there is a quality problem. And of course, all of this inventory gets dumped into an exchange without any of this context and advertisers are left wondering why this highly viewable inventory has terrible performance. I do not believe that we want to go down the path of aggressive ad rotation. We will be right back where we started. Also, I think ad rotation is not overly dissimilar from in-banner video. This must be very carefully considered for its impact on the consumer experience.

I want publishers to be able to survive, but I want them to survive by creating a quality product for consumers.

I do think the challenge that publishers and the industry face is that they fail to recognize that advertisers treat viewability more as a state of mind: They want their ads to be viewable! We look at viewability – heck, I look at viewability – and see a wave of technology challenges, vendors with mediocre projects, and false positives wrecking deliverability when we feel GREAT about our inventory quality. Managing the forest and the trees is the true calling of ad tech in this area.


Leave a Reply