ChasNote

Bill Wise, CEO of Did-it,”a New York-based firm specializing in auctioned media management,” according to his byline in Mediapost, says there’s no reason brand advertising campaigns on TV shouldn’t adopt the same self-service auction approach as eBay or Google (Mediapost).

“The TV networks, meanwhile, are not happy. An article in last week’s Wall Street Journal nicely covered the networks’ displeasure. Members of the TV world put forth a number of arguments why they think the system won’t work“

The TV execs Wise sites talk about custom integration, product placement, and campaign packaging — campaign features that are difficult to automate — as reasons the approach won’t work for TV. There’s another: Finding appropriate context in a fragmented media world. As customers fan out across a more diverse and diffuse media landscape, marketers need scale; marketers need partners or platforms that provide media-buying efficiency that matches the old days of spending $10 million across a handful of networks. Ad networks and syndicated paid search are both good at scale. But brand advertisers, to a much greater extent than direct-response or paid-search marketers, care about context, and even the best contextual ad-serving technologies still miss the mark too often (see ChasNote 8/12/06) to kill off the humans just yet.

Piers at PSFK observed that Google’s AdSense algorithm pushed to his site ads for low rent, get-rich-quick schemes instead of ads that matched his content (corporate identity, brand marketing and media-business trends) or the audience profile (entrepreneurs and marketing execs):

“Google Adsense is a fiendishly clever system that predicts the perfect ad for the reader based on content and some other dark-magic. Anyway, the system must have gone wrong. Instead of ads for fast cars, boutique hotels, fine wines and expensive watches, we’re getting ‘make money fast’, ‘make money blogging’ and ‘email marketing software to spam your way to success’ ads.”

His post reminded me of a call I received a few months ago from the Ad Council. They were concerned that their client, Girl Scouts of America, who bought ads through a behavioral-targeting ad network FM uses to backfill unsold banner impressions, ended up with ads on our adult-beverage site, Drink of the Week. Oops! Humans still have the bots beat when it comes to the subtle art of understanding context.

Revver has built a cool system for inserting ads at the end of online videos in a manner that allows advertisers to select flight dates and enables content producers to monetize their hits by rotating in new advertisers at the audience for a particular clip grows (see AdAge). Clickable ads run at the end of each video. With the old approach — attaching a single advertiser to a single video clip — an unexpected viral success gave the advertiser a bargain but the content producer ended up underpaid for his or her work. Revver’s technology can even the playing field.
But I worry the model may still underpay the content creators. One, Revver takes 50% of the revenue, even when a viewer finds the video by way of the content producer’s own site; the content producer makes the content AND finds the audience, and only keeps 50% of the money? Two, advertisers are paying only when viewers click on their ads. This works fine today, while the technology is novel and the click-through rates are high. Say an advertiser pays $1 per click and 4% of viewers click through: that’s a $40 effective CPM! But if CTRs on ads in online videos follow the path of ads on websites, in email newsletters, on blogs and in RSS feeds — which all started in the single digits — they may settle somewhere below 1%. At 0.5% click-through, that effective CPM drops to $5, $2.50 to the content creator for every thousand viewers.

Tom Hespos latest MediaPost column hits the mark on ad targeting so well, I wish I had written it. (My attempt: ChasNote 4/13/05.) Because online advertising has better targeting capabilities than offline media, media buyers take advantage of online’s micro-targeting and lose the benefit of reaching prospective paying customers who don’t fit the “ideal customer profile” outlined in the RFP.

“Let’s say we’re putting together a test launch of a widget in Philadelphia. To advertise that widget, we decide to go with some spot radio, so we plan to purchase 15 GRPs a week in Philly against our demographic target–Men 18-49. We do a little spot cable, too. and we also look to advertise the widget online across six content sites, using inventory that is both geotargeted to the Philadelphia DMA and demo-targeted to M18-49…. One big difference is that the Internet advertising is a lot more focused on the target. Remember that, in advertising on the radio and on cable TV, our ads are guaranteed against our demo target of M18-49. But in advertising in those environments, we also reach a ton of women, kids, senior citizens and other folks we might not consider to be part of our target market. Assuming that the widget can be used or purchased by people outside the demographic and geographic target, some of those people from outside the targeting parameters of the traditional plan are going to buy the product…. The Internet’s targetability looks like a surgical scalpel in comparison….. The dynamics of how broadcast and online are purchased have led quite a few advertisers to carve out a distinct role for online in the media mix — online becomes the focused medium that reaches the target and the target only. At the same time, broadcast inherits heavy-lifting chores, in large part due to the limitations on its targetability. Seems strange, doesn’t it? One medium does a great job of reaching a focused audience and is rewarded for it by being relegated to the role formerly played by direct mail.”

Wired News on Google’s — and our whole industry’s — problem with click fraud:

“Google’s $6 billion-a-year advertising business is at risk because it can’t be sure that anyone is looking at its ads. The problem is called click fraud…..

But the overarching problem is both hard to solve and important: How do you tell if there’s an actual person sitting in front of a computer screen? How do you tell that the person is paying attention, hasn’t automated his responses, and isn’t being assisted by friends? Authentication systems are big business, whether based on something you know (passwords), something you have (tokens) or something you are (biometrics). But none of those systems can secure you against someone who walks away and lets another person sit down at the keyboard, or a computer that’s infected with a Trojan…..

Standard testing doesn’t work online, because the tester can’t be sure that the test taker doesn’t have his book open, or a friend standing over his shoulder helping him. The solution in both cases is a proctor, of course, but that’s not always practical and obviates the benefits of internet testing.”

Bring back the humans! Look, I’m a big fan of technological innovation when it both creates efficiencies (higher productivity, lower costs) AND it does the job as well or better than a person could. Like Excel or email, for example. But too often we hand over the reigns to the bots because they’re way cheaper and only a little worse than the staffers who used to do the job. Think of that automated voice system (Simon?) you talk to when you call United’s reservation line: It’s pretty good when you need find your exact departure time, but it’s terrible if you’re trying to book a trip to New York with a stop-over in Denver to drop your kids with their grandparents. Advertising — especially when the goals are brand-oriented instead of transactional — isn’t ready for a fully automated system.

In his SearchInsider column for MediaPost today, Bill Wise discusses Google’s new daypart targeting for keyword ad campaigns: “This is good news for search: it is bound to greatly heighten advertisers’ precision, and eliminate campaign waste.”

Now I know I’m out of synch with most traditional marketers when I say this, but I just don’t get why anyone should care about dayparting online — especially with respect to CPC marketing. If your prospective customers are doing their research or shopping at 2 in the morning, why would you turn off your ads then just because more of your prospects shop during lunchtime?! Instead tell your IT team to expect spikes in certain dayparts, but keep the ads in front of relevant prospects whenever they are reading relevant content.

Dayparting on TV makes sense to me, at least in the pre-Tivo, pre-iPod days. The good shows air in primetime, and that’s where you want your ads. Since the crap, re-runs and infomercials are on late night, daypart your ads to avoid the weak programming that doesn’t attract your audience. Nowadays if, say, 24 is the best program on TV with the most desirable audience, associate your ads with that programming and don’t worry when those great viewers tune in. Keep your ads away from junk content, no matter when people view it.

The same approach should apply to online marketing: Seek out the quality publications and sites, avoid the rest, and don’t worry when your customers log on. By adding daypart targeting, perhaps Google is admitting that advertisers across their sprawling AdSense network of sites have begun to complain that their ads aren’t consistently landing in quality environments or reaching high-value customers.

Chitika, the contextual shopping / reviews widget that publishers can post to their sites to provide their readers a search tool for related content, may be a bit racier than its syndication partners expected. From JenSense:

“images included in the ad units happen to be nearly-naked women gracing the covers of Hustler and Genesis magazines. And there is absolutely no supporting text on the page that would influence these types of ads. It is strictly ‘magazines subscriptions’ as the targeted keyword phrase that is triggering these adult ads.”

This is too much! John Jantsch at Duct Tape Marketing reports to eating Pringles potato chips with ads on them.

Battelle posted news (Searchblog) that, earlier this week, he launched his first AdWords / AdSense program promoting a real service, FM’s new media planning interface. Perhaps it’s no surprise, his readers lit up the comments line. What did surprise me is how much his readers — the gurus and wizards of the search marketing world — still DON’T know about how Google’s ad programs work:

“I have wondered for a long time why Google doesn’t provide more transparency with its click data. Your content click reporting issue occurs with site targeting as well. Not sure if they are consciously hiding the information from advertisers or just don’t have the tech to push this data out.”

The person who wrote that comment did put a winking emoticon after that last line. Of course Google “has the tech” to push out that data; they apparently don’t want to.

Last night at 9pm FM announced the beta launch of a seach engine for marketers who want sort the leading weblog sites and place ad campaigns (in real time) based on content categories, audience demographics or available ad inventory: Federated Media blog.

For those of you who called me a media dinosaur after my post on the extreme visions of frictionless media buying (ChasNote 4/14/06), here’s a step in the right direction!