The Contextual Advertising Technology Blues

Have you ever heard venture capitalists (VCs) referred to as lemmings? You know, those little rodents who run blindly in large numbers off a cliff. Though the lemming’s bad rap is based on myth rather than fact, it has certainly become good lore. And though most VCs would certainly object to the association, their lemming-like behavior has more data to support it than those rodents’ suicidal tendencies.

The latest cliff-diving exercise may be the quest to build a better contextual advertising mousetrap. Deals are chasing money, and vice versa, based on the contention there’s an opportunity to improve the relevance, and hence increase the yield, of online advertising.

Sound like a good idea? It’s hard not to find it compelling, especially when presented against a backdrop of Yahoo’s and Google’s conclusive demonstrations of how to create billions of dollars in shareholder value through keyword advertising. Nevertheless, a good idea isn’t enough, and a lot more lemmings chasing these deals will die than take flight, let alone soar.

Last week, AD:TECH was here in San Francisco. Bless the organizers for realizing we’re in the age of engagement. I had lunch with an acquaintance attending the show. He works for an East Coast company that does “better” online media placement than DoubleClick. The company’s definition of better is better customer targeting. It can supposedly be really smart about how it targets Web visitors based on individual profiles and behavior. Big Brother aside, it sounds compelling enough. What advertiser wouldn’t like to know Joe Web Visitor is 34, drives a Lexus, and has a 700 credit rating. Surely, that’s been online advertising’s Holy Grail for a long time.

The problem is even if my friend’s company can demonstrate better yields, unseating DoubleClick for media placement isn’t easy. DoubleClick has a lock on its customers, who often sign long-term contracts, receive nice volume discounts, and have many other incentives not to switch. There are several players in the space, and between them they’ve locked up the bulk of the online media market. As with so many other markets, better isn’t good enough unless what exists is very broken. And online advertising isn’t very broken.

What does this have to do with lemmings? In recent conversations with entrepreneurs and investors, I’ve come across a number of companies on a mission to create a better contextual advertising mousetrap. Some very smart technologists and experienced online advertising executives are promoting compelling-looking solutions with interesting business angles. Some say they’ll do better contextual ad placement based on semantic parsing of ad and site content. Others will track customer interactions and “get to know them better,” hence allowing more targeted ads.

I had a VC tell me the other day his company had just seen a startup demonstrate a 300 percent improvement over Google’s contextual ads. Don’t ask me how! Its business case varies on the theme that since Google and Yahoo make so much money and the price of popular keywords continues to increase, even a relatively modest improvement in yields, certainly a 300 percent improvement, will result in enormous value creation.

Meanwhile,’s search engine,, Google, and Yahoo are all trying to convince users they should register when searching the Web. In a round of announcements over the past few weeks, Yahoo introduced “My Web” and Google something it calls “My Search History.” These all tout big benefits to the end user. The boost to the bottom line may turn out to be equally significant through the ability to build detailed profiles and histories of what people search for and look at.

Better Customer Insight = Better Context = Higher Ad Yields = Higher Margins

What I don’t understand is how “better” technology among myriad contenders trying to improve contextual advertising will be enough to make the slightest difference. The big guys have the customers, they have the advertisers. In short, they have lock-in. They have smart technologists, too. What am I missing?

My prediction: A few of the “better” contextual advertising startups with really hot technology and valuation expectations that aren’t too inflated may be acquired by Google or Yahoo The rest? They’ll jump off the cliff and die. Better is only better if what I’m doing today is broken.

Incidentally, I also predict another 100 million, or 500 million (who’s counting?), in VC investment dollars will be flushed down the toilet. That process is always accompanied with the proverbial “giant sucking sound.”

Related reading

Overhead view of a row of four business people interviewing a young male applicant.