My friend Alex was surprised to learn that a buddy’s Yom Kippur Break Fast this year is apparently “sponsored by Pork.” At least that’s what it says on the banner ad Evite placed on the RSVP page. As my mother-in-law always says, you have to keep an eye on those reformed Jews.
eMarketer predicts that spending on desktop ad banners will peak in 2014 at $39.35 billion, and then will slide south in subsequent years as digital ad dollars move to mobile. By 2017 ad spending on non-mobile banner ads will be down around 2012 levels.
A major reason for the shift toward mobile is simple: With more than half of US mobile users now on smartphones, and time spent with mobile devices increasing each year, many digital publishers are looking to shift ad revenues to mobile. Smartphones and tablet devices also account for a growing portion of US retail ecommerce sales, further contributing to advertisers’ desire to shift dollars away from desktop.
But I have to wonder if there’s something else. Since the beginning, the banner-industrial complex* has been measuring its success by clicks, even though we’ve all seen the data telling us a single-digit percentage (8% back in 2009) of Internet users are responsible for that vast majority (85% in 2009) of all clicks on banners. Those heavy clickers, meanwhile, tend not to come from the demos that most advertisers intend to reach.
And it gets worse. The latest numbers from Comscore show that 46% of desktop banners are never seen by human eyes. On ad exchanges like Google’s AdX, according to Piper Jaffrey, that number might be as high as 80 to 90%.
Were it not for the industry’s own bad behavior, I expect that eMarketer chart would have a longer up-and-to-the-right curve to it. Let’s not blame it all Apple and Samsung for making such cool mobile gadgets.
(*I think Brian Morrissey came up with that phrase.)
What happens when a publisher fails to create quality advertising products designed for its unique content experience (native advertising, broadly defined)?
It concludes, correctly, that it won’t get much of a premium by having human beings sell its not-very-well targeted 300×250 banners. So it offloads its inventory to ad exchanges and networks. Since the exchanges and networks don’t deliver a great CPM for most impressions, the publisher is forced to add more banner placements just to maintain revenue-per-thousand-pageviews levels it achieved two years earlier. Not a pretty picture, no matter how you look at it. The money still isn’t great, and the user experience is horrifying.
Here’s a screenshot from the San Jose Merc website on February 7 snapped by John Battelle. From his post on Searchblog:
Six or more of [ads] in this screenshot, and three more below the fold. There’s a Verizon site wrapper (on either side of the page), an expandable top banner, and three medium rectangle units crammed in there. Not one of them is what you might call a ‘quality’ ad — at least by most standards. (Do you think Verizon is happy that their site takeover is overrun by social media buttons and competing with belly flab, diabetes, Frys’ Electronics and travel pitches?) If you bother to scroll down (who would?) there are three more pitches waiting for you there.
And check out the number of beacons and trackers on the right, in purple. That’s Ghostery, which I run on my browser to see who’s laying down data traps. Man, Merc, that’s a lot o’ data. Are you doing anything with it?
Look, I’ve built my career around ad-supported media, and I continue to believe that advertising (in some shape or form) will support digital publishing. But if there’s one thing you learn early — somewhere around your very first day in the business — advertising does not work if there aren’t consumers on the other end to look at the ads. So if you ad strategy erodes your audience, or merely burns out their eyeballs by way of toxic design, you’ll soon be left without a business at all.
From Andy Ellenthal’s post at Digiday. He’s the CEO of semantic ad tech company Peer39.
“Considering how much attention, from the press and venture capitalists, is paid to audience-based ad buying, you’d think it would inevitably rule the roost — at the expense of old-fashioned content-based buying. You’d be wrong.
“The hype is a natural outcome of the fact that audience data has never had this scale and accessibility before. Buyers can now match their targeting criteria against huge pools if impressions. The concept is super cool: Displaying your message only to the exact group of desired consumers?
“But the reality reminds me of the difference between a retailer’s merchandising and marketing. Merchandising’s job is to sell once a consumer is inside the store, but marketing needs to drive them there in the first place. There will always be more prospective buyers and influencers outside the store, than shoppers with wallet in hand. If you only merchandised, at some point soon the store would be a very quiet and lonely place.”
UPDATE: Mark Chu Cheong’s take:
Great suggestions by Battelle (Searchblog) on improving online ad targeting — both for those who feel stalked and for the stalkers.
“As I’ve said a million times, marketing is a conversation. And retargeted ads are part of that conversation. I’d like to suggest that retargeted ads acknowledge, with a simple graphic in a consistent place, that they are in fact a retargeted ad, and offer the consumer a chance to tell the advertiser ‘Thanks, but for now I’m not interested.’ Then the ad goes away, and a new one would show up.
Facebook already does something similar, as Battelle points out. So do the story-list ads on Digg (“DiggAds“).
“And when a consumer says ‘no thanks,’ as any good salesperson knows, that’s an opportunity to learn. No rarely means no forever. Marketing is a conversation, one with more than one exchange. Just because the first one isn’t a sale, doesn’t mean the next one (or the one after that) can’t be. Especially if you have the good graces to know when to pull back into the wings for a while.”
TechCrunch points out that my two most recent posts — Steve Jobs’s medical leave and deceptive ads served by Yahoo’s Right Media — have a connection. At least within the flawed logic of the ad-targeting machines used by ad networks and context targeting engines. The screenshot below, from TechCrunch, shows a Kaplan University ad (“Jobs Become Obsolete. Talent Doesn’t.”) alongside a WashingtonPost.com story on the Steve Jobs situation. An added wrinkle: The Washington Post owns Kaplan.
Om talks to Yahoo about the dearth of premium online CPMs: The web needs more quality content and something that makes the content unique. For example, the GigaOM sites!
I couldn’t agree more, Om.
Except when you say GigaOM competes with the ad networks. The GigaOM sites — sold individually to brands that want to associate with the premium, exclusive content in the rarefied context that is reading a site like GigaOM or New Tee Vee — live at the top of the media-pricing pyramid. That’s not to say that unsold inventory on those sites can’t be sold as part of a blind ad network like AdSense or Blue Lithium. Publishers such as GigaOM, FM, or ESPN offer one kind of value (building brand value through association with quality), and ad networks offer another kind of value (low-cost clicks that answer a direct-response need). The two can use inventory from the same sites symbiotically, if both entities understand the value they bring to the marketplace.
(Disclosure: The GigaOM sites are part of the FM family so I’m biased when I say you should buy ads on them.)
Another one for the contextual mis-targeting file, courtesy of ValleyWag:
“We’ve got a call into Walmart, but our guess is that through ad network LinkShare’s affiliate marking program, hundreds if not thousands of Web site publishers put Walmart banner ads on their sites in hopes of referring shoppers and earning a slice of revenue from whatever they buy on Walmart.com. It would be very difficult to thoroughly vet each publisher. But if there’s ever been a need for a clear example as to why Madison Avenue interactive agencies do not trust their clients to ad networks that claim extensive reach above all else, there is no more.”
Last year AOL paid $275 million to buy Tacoda. Now, according to Venture Beat, AOL is dropping the brand and rolling the technology into Platform A’s Ad.com unit.
“iThis is a shocking move for some, because Ad.com doesn’t target much at all, and offers ads of $1 or less per a thousand views — and is generally considered a ‘bottom-feeder’ by some in the industry.”
Times are tough at ValueClick, too