July 25, 2016

Display Ad Horseshit

Here at The Ad Contrarian Worldwide Headquarters, we're in a never-ending battle with the forces of evil who are trying to get you to piss away your money on worthless, wasteful, creepy and abusive "interactive" advertising.

Last week's post called "Tons Of Data And Not An Ounce Of Sense" established for all eternity that no one on planet Earth interacts with "interactive" advertising. 

As expected, the dimwits who know nothing about advertising have sent me lots of nonsense about how display ads shouldn't be measured by clicks, but by how effective they are at brand building and engagement and all kinds of other imaginary hogwash. Apparently, these goobers believe this crap.

It's time to put this bullshit to rest.

The following data comes to us courtesy of Lumen, a research company in the UK that since January has been conducting what it calls "the world's first eye-tracking panel." 

What they do is kinda technical but simple to understand. They have a panel of over 300 people. They've done 35 controlled studies and studied 28,000 minutes of "natural browsing" against more than 3,000 different online ads. They follow peoples' eyes to see what they look at. Let's have a look at what they've uncovered. (I want to be clear, the research is theirs, but the conclusions are mine.)
The first column is impressions. Impressions are the crap you're paying for. In any truthful world an impression would be defined as one person seeing one ad. But in the corrupt world of online advertising an impression has nothing to do with either people or ads. According to the IAB, an impression is 
"...a measurement of responses from a Web server to a page request from the user browser"
Got that? No humans, no ads. Servers and browsers.

The second column shows us that only 54% of "impressions" are "viewable." This means that even if people wanted to see your crappy ad, half the time they can't.

It may not load in time. It may be "below the fold" where it can't be seen. It may be the result of fraudsters who send a pixel that registers as a "response from a web server to a page request from the user browser," or it may be some other incomprehensible technological or criminal reason above my pay grade.

But the point is, about half of all online ads you pay for are not even visible.

If you're asking yourself, why in the the fucking world would I pay for an ad that's not visible? I have bad news for you. You'll never be a CMO.

Now we get to the third column. This column shows us that of "viewable" ads only 65% are actually viewed. Doing the math, this means that only 35% of the total number of "impressions" we bought are actually seen by anyone.

But wait there's more wonderful news. Of the 35% of ads that are actually seen, only 25% of them are noticed for at least a second. In other words, 75% of the time ads that are technically "viewed" are really only glanced over for less than a second.

This brings us down to 9% of our so-called impressions being seen for a second or more.

Can it get worse? Sure it can, this is the web. It always gets worse.

Now we have to factor in the bots. Remember, Lumen only tests humans. But as we surmised last time, at least a third of the time an "impression" is not a human at all, but a malignant software thread that is pretending to be a human.

So, when we're done with all the shady online horseshit, what we're left with is this -- at the end of the line, of the 100 "impressions" you paid for, maybe there are 6 real people glancing at your ad for a second or more.

If that's your idea of money well-spent on brand building or engagement you, my friend, are what is known in my hometown of Brooklyn as a fucking moron.

No comments: