March 9th, 2006

The Coming Search Advertising Crash


I’m not sure if I’ll be the first to predict it, but search advertising is headed for a cliff. I can’t say with any certainty when–or whether–the crash will come, but the evidence is mounting.

Everyone is talking about the $90 million click fraud settlement, and the Google apologists are out in force, arguing that $90 million is a drop in the bucket and that the settlement reduces the likelihood of major legal liabilities in the future. I’m not a lawyer, so they may very well be right that Google is in good shape from a legal perspective. Of course, you can count on Google bear Henry Blodget to find a way to rag on Google, but even he’s missing the big problem here.

The issue is not whether advertisers can sue Google for any significant amount of money. The issue is whether advertisers will begin to lose faith in search advertising.

While everyone was getting their knickers in a twist over the settlement yesterday, I came across this article in the Financial Times–Search engines are not the only sites–which makes this disturbing observation (disturbing from the perspective of the Google fan club):

But, looked at another way, the search engines run by Google, Yahoo and MSN have a weirdly high share of internet advertising. Most people spend 95 per cent of their time onthe internet doing other thingsthan searching. They read newsand entertainment sites, send e-mail or instant messages, blogabout their beliefs or even displayindecent photographs of themselves.

So 5 per cent is the number that Google, Yahoo and Microsoft’s MSN Search should worry about: the amount of internet time devoted to searching. Perhaps it is even lower, since a lot of searches are so broad and unconnected with any impulse to spend money that no one will buy a linked advertisement. Whatever it is, an awful lot of advertising revenue flows from a single digit.

Advertising does not work this way in other forms of media. On television and radio and in newspapers and magazines, adverts are sold according to ratings and circulation. This means the number of people who read each edition of a magazine, or who view each showing of a programme. Publishers and broadcasters gain advertising by offering the chance to reach these people in a block.

If internet advertising followed suit, revenues would not be concentrated overwhelmingly on a small portion of internet traffic, but spread more evenly. In practice, however, it is hard to find sites with enough traffic (and therefore advertising potential) to attract large blocks of advertising. There is fierce competition to place brand advertisements on the few that do exist, such as the home pages of Yahoo or

This is media economics 101. Ad dollars follow audience. If search only represents 5% of online media time, it shouldn’t have 40% of the dollars, no matter how measurable search advertising is. Or as FT put it:

But can search engines maintain a 40 per cent share of advertising on a 5 per cent share of internet viewing indefinitely? That seems unlikely, no matter how smart their engineers and how valuable the signals their users send. As Mr Reyes bluntly indicated, Google and its peers have already made their easiest money.

I’ve cited other evidence that Google and search advertising are at risk:

Is Google Like a #1 Band on the Decline?
Is Search Advertising Reaching a Plateau?
The Short-Term Value of Google Advertising

Google and search advertising rode the hype train all the way up the mountain, and there’s a long way to fall. If advertisers lose faith in Google and the value of search advertising, there’s a rapidly growing and vibrant digital media universe that’s waiting to absorb those search ad dollars.

Google understands this risk, which is why they are working so hard to get some revenue eggs into non-search advertising baskets. Again, there are many reasons to be skeptical:

Google and Its Watchers Don’t Get Print Advertising
Google Tilts at Offline Advertising ROI
Google Chases the Declining Print Ad Business

When you live by the ROI sword you can die by it too. Will advertisers start to question the ROI of all the money they’ve dumped into search advertising in general and Google in particular? If they do, we all know how quickly the flow of ad dollars online can shift.


I wanted to respond to some of the comments below.

First, just because you can’t easily measure the ROI of brand advertising like you can transactional search advertising doesn’t mean the value is not there. And advertisers know this.

So the “search audience” is inherently more valuable than one, say, watching sports or reading

But surely the argument is that users who are searching are in a state of mind that it far more “open” to adverts than users of other services.

This is really the crux of it — do people have arms wide open to marketing while searching but have their back’s turned when engaging in other online activities? Other than rhetorically, it’s difficult to prove that all other forms of online ATTENTION are significantly less valuable than search attention.

On this point, Dave says:

How do we know it’s undervalued? All you’ve suggested so far is that search is 5% of online minutes, but gets 40% of the dollars. Are you suggesting that search ads should be no more valuable “per online minute” than other kinds of brand-building ads?

My response to Dave is that we just don’t know. It may well be that search attention IS more value than other online attention. The question is how MUCH more valuable? 5% of the attention getting 40% of the value is a significant imbalance — like Google’s stock price, you have to wonder whether the premium is justified.

Dave also says:

Also, do ads displayed on other websites through google’s adsense program fit into that 40%?

Great question, to which I don’t know the answer. But even if the stats were properly adjusted, I suspect there would still be a large imbalance.

I still come back to the issue of advertiser perceptions. Danny may be right the Google has locked in the money from the small business who are running their search marketing as a P&L. If that’s true, then Google may be facing a wall rather than a cliff, because at some point (if not already), they will have squeezed all the monetizable value out of search attention.

And if Phil is right that AdSense for content is what’s really at risk, Google may find themselves facing stiff competition in the effort to monetize non-search attention online.

If this weren’t the case, why would Google be chasing Old Media forms of ad targeting, e.g. by demographics?


Comment from Jakob Nielsen:

When on non-search sites, users do not look at the ads. So if you value attention and brand-building, you’re not getting it, because users are not allocating their attention to the ads.

You can see one example of an eyetracking plot from a study I am currently running. All of the pages we have analyzed so far look like this: almost no fixations in the ads. (More formal results to be reported later, after the study is done.)

The eyetracking plot is fascinating. But just because current advertising formats (including AdWords) do not attract attention on non-search sites, that does NOT mean it’s not possible to find ways to redirect that attention for marketing purposes.

Also from Jackob:

I agree with several other commentators on this thread that you also have to account for intent: Are people looking for something new? Are they researching or buying, or just browsing?

Sure it’s easier to market to people when they’re looking to buy something — that was the search marketing innovation — but effective marketing DID happen pre-search. Brand-building CAN happen even when people don’t have buying on their minds.

With all due respect, this huge faith in the status quo that is evident in all the comments below strikes me as a failure of imagination. It’s the same failure of imagination that kept online advertising in the doghouse until the Google AdWords revolution reinvented the rules of the game.

Why should we just walk away from the challenge of discovering a new way to effectively monetize the 95% of non-search attention?

When someone figures out a way to do this, THAT’S when the correction will happen.

  • Ted

    Grumpysecretery makes a good point. Newspapers will need to upgrade their search engines, and organize their content in a better way (the taxonomy approach might be good here) so people will start to think of them as a resource.

    As a counterpoint, however, why would I go to and search for sushi, when I can go to and search for 'sushi 90049'? Presumably this would include indexed LATimes pages as well as others. Is it the trust factor? If that is so, then Newspapers have a huge advantage, as there is still quite a wide gulf between how much most people trust search results and how much they trust a known local news source with the prestige of a LATimes. Interesting comment!

  • Interesting but the success of adwords is in the EASY way you can open an account and put your advertising. Same for Adsense. For big firms your are right but for millions of little firms, the success key is to be EASY SYSTEM !!! People prefer easy system that brings 0,01 than complicated system that may be can bring more.. Why ? Only because they want to spend maximum 5% of their time ;-)) to learn how the systme works ..

  • grumpysecretary

    Not a crash per se. Rather a shift. Advertisers, especially local advertisers will realize that a search result within the site of a trusted provider, such as a major newspaper, is much more valuable than a google result.

    The shift is coming. It would have been here already but newspaper editors have been very slow to see how valuable search results can be within their sites. Because in the end, most of what I want to find is local, and I'd rather turn to local trusted providers. And now that DART and others can serve rich media ads based upon key words searches within sites... things are gonna shift.

    Newspapers are going to gain a larger share of online ad buys for search results. And those buys from local and national advertisers will target search results within the newspaper sites. And as this happens, there will be competition between advertisers for those key words and thus the CPM will rise.

    The LA Times for instance, will be able to sell ads based upon a search within their site for "Sushi" for the same amount or more than a homepage ad placement. And their search data will tell them which keywords to market and to which advertisers. Soon there will be stand alone key word search results "rate cards"

    Really! Wait & see!

  • If figures matter :

    Point 1 :
    1 internet session out of 3 begins with the user (the "audience") performing a research on a search engine.

    Point 2:
    Datatracking allows things that other media do not :
    - the "media" (the search engine) knows what share of the audience watches a given page (one page per searched expression)
    - both the "media" and the advertiser get to know what share of the audience does notice AND click trough an advertising
    - the resulting clicktrough ratio is IN ITSELF an indication of the value of a given adspace (just like it would be for a magazine ad that could be identified as having led a potential customer to cross the door of a shop)

    Point 3 :
    What really gives an adspace its value is the amount of transactions it leads to, that :
    - cannot be tracked with precision for TV, Movie, Radio, Press ads
    - can be precisely monitored and thus enable Spends fine tuning DURING the advertisement campaign

    Last point :
    While search engines use represent only 5% of the time a user spends surfing on the internet, it does represent, for any given amount of adspends shared equaly between "traditionnal" (on site banners, or even rich media ads) and search engine marketing, more to far more than 40% of the transactions your advertising campaign will generate.

    My belief :
    Search Engine Marketing represents a volume of advertising share that is equivalent to the sellings it generates (I even think it's often under-estimated). It can still grow, while traditionnal onsites graphical advertisement will be mainly used to build brand awareness.

  • Ted

    Yellow Pages garner about $15b in advertising

    This is something that I surprisingly don't hear about enough. Google/Yahoo/MSN are all moving hard into that market. The Yellow Pages are disappearing (somewhat) and those dollars will flow into their online/mobile/in-car equivalents. That is why, even if Google plateaus with search ads, there is still at least one monster place where they can go.

    Throw in direct mail --> email and you've got another monster ad-eater.

blog comments powered by Disqus


Receive new posts by email

Recent Posts