iMedia Connection

How Google destroyed the ad model

Tom Hespos

Nobody seems to remember the old ecosystem, much less why it went extinct. That's problematic if we want to avoid extinction ourselves.

I'm talking about the search landscape. There are quite a few of us old-timers left around who remember when search advertising was a display product. Advertisers bought keyword-triggered banners on a CPM basis from the likes of Lycos, Infoseek, and HotBot. More often than not, all those big "portal deals" that made headlines in the mid-1990s contained a rather sizeable component of search inventory, and you could buy whatever keywords you wanted -- even if they were completely irrelevant to your business and what you were selling.

Most ad sales reps didn't care if your keywords were relevant or not. They made money on the ad display, not the click. So if you were selling tent pegs but wanted to buy keyword "baseball," they didn't put up too much of a fight. The ads were, in many cases, completely irrelevant to searchers and their mindset. The ad model was destroying the product.

Meanwhile, on the organic side, gaming the system became the norm. SEO was less about content and linking strategies, and more about meta tags and submission tools. Search engines weren't incentivized to spend money on maintaining the editorial integrity of the product, mostly because they were making money selling keyword-triggered banner campaigns for double-digit CPMs. 

So the model we had at the time was a candle being burned on both ends, with irrelevant ads and irrelevant search results eroding the utility of the product, until users just gave up in frustration.

I think that anyone who witnessed this first-hand should have a special appreciation for the genius of Google's ad model. That's right, I said genius.

If irrelevance killed the old model, relevance was to be the core of the Google model. Not only would relevance be the ultimate driver of useful search results for the end user, it would be baked into the ad model as well.

Under the newer model, advertisers who didn't want to spend time thinking about the mindset of the searcher would still be able to advertise -- they would just be penalized heavily. This penalty would come partly from opportunity cost and partly from paying more for the irrelevant traffic they received.

You see, Google's economic interests are directly aligned with how relevant paid search results are to the end user. So ads are subject to a Darwinian process wherein if they don't generate revenue for Google, they either get canceled by Google or the advertiser ends up paying more on a per-click basis for fewer clicks. This lines up directly with how relevant end users think an ad is.

So if you're active in search but don't live and breathe relevance, you're almost certainly missing opportunities to expand into search terms that ARE relevant. At the same time, you're probably paying more on a per-click basis for the traffic you do get from paid search. So you're economically incentivized to understand search behaviors, test for relevance, and get rid of terms you're not relevant to (even if you do believe that you're relevant to them).

This model not only caused a stratospheric ascent for the search category in general, but it thoroughly destroyed the old model. It destroyed it to the point that it's possible to work in search marketing today without ever having heard of HotBot or Excite or anything having to do with the old model.

Once you fully understand and appreciate the nuances that brought about this cataclysmic change, you begin to realize that Google has an interesting challenge when it comes to growing its business.

Under the old model, sales reps working for search engines could sell advertising and have plenty of inventory available for advertisers who wanted to scale up their buys. Under the new model, how do sales reps balance relevance and volume?

For advertisers who are not very experienced in search, you can build suggestion tools that give inexperienced advertisers plenty of search terms to bid on and experiment with. But what about those advertisers who have done the testing themselves and can't spend more without becoming less relevant to searchers? In such cases, a sales rep looking for a larger dollar commitment is actually at odds with Google's business model itself.

Sure, Google will likely always need people to help grease the wheels. It needs people who can handle billing, assist with complex analysis, and help find new ways to show search advertising's relevance to an advertiser's business. All along the way, educating new advertisers about the nuances of the ad model is a key to unlocking ad budgets.

But trying to push irrelevant advertising on experienced search advertisers doesn't make sense. If I were directing sales at Google, I'd place more importance on making sure that the top advertising spenders in the country were adequately educated about how search works, and less importance on getting experienced advertisers to spend more. 

I'd also share the right methodology for testing, ensuring that advertisers fully understand what happens when they think their products are relevant to a term when they're clearly not. I would also emphasize showing how search fits with the rest of the marketing mix.

But I most certainly wouldn't have my sales reps out of alignment with my business model. That's for sure.

Tom Hespos is the chairman and president of Underscore Marketing and blogs at Hespos.com.

On Twitter? Follow Tom at @THespos1 or @_MarketingLLC. Follow iMedia Connection at @iMediaTweet.