I would like to put forth a better reason for why investors were so quick to believe comScore's numbers. Namely that no one really believes the effectiveness of online advertising in the first place. Quick - everyone that has intentionally clicked through on an ad raise your hands.... One..Two... yeah that's about it.

I suggest most folks are like me in using Google - I ONLY click on a text ad when the normal search results have absolutely nothing promising in the first 2 pages. As a result, despite having been involved with at least 1/2 a dozen websites associated with businesses, I have never felt comfortable recommending that ANYONE buy an online ad from either Yahoo or Google. Why? Because I have yet to be convinced that such advertising works for much of anything. Sure some broad priceshopper links might work, but other stuff? Nah

So when comScore came out with a reason to disbelieve the Google ad numbers, everyone was quick to believe it.

And frankly so am I.

I'm a Google Advertising Professional. I have been, since shortly after the program came out in 2005. I've been cited by Google staff as one of the most knowledgeable people about the use of AdWords API. I've worked on significant AdWords accounts such as Skype and Thomas Cook. I say this to provide you with some credentials that I have some idea about AdWords. I'm also the most prolific poster to the AdWords Help Forum and I blog, mostly for other search marketeers, mostly about AdWords.

From my client accounts, Google is seeing reduced CTR. This is because Google has started using search history to choose adverts. The consequence is that whereas previously, all adverts used to have a traceable relationship to the search query, I can now see two thirds of the adverts with no relationship to the current search, or the previous search - but a bastard composite. For example, search for "Brazilian Tax Credit" and then search for "US Vacation" and you get adverts for Brazilian Holidays and US Tax. This is dropping CTR for my clients as they get exposed to less relevant searches, and it is increasing my Average Cost Per Click, as highly relevant adverts are now in the auction with high value but irrelevant adverts.

There may be backtracking at comScore, and elsewhere, but I'm pretty damn certain that clicks go down when page relevance is reduced. It's basic search engine economics. SE's with irrelevant results get lower click volumes and lose audience to SE's with more relevant results. QED.

That doesn't really counter the point I was making - namely that most everyone involved has serious qualms about the actuall effectiveness of online ads - especially text ads on a search engine. You'd have a tough time convincing me that its a better investment of my ad $$ for my new startup than sending high quality print flyers.

Methinks you're right on. I have never found any convincing evidence about the effectiveness of online advertising. The arguments are barely as convincing as "stadium/tv impressions" at sporting events ... which have zero value (unless one counts the ego stroking for corporate jocks).

Sorry, I thought that I'd answered that quite effectively. I've been a consultant to industries that spend millions of dollars per year in paid search to get online sales. If I couldn't track the clicks into sales, how long do you think it would go on?

Paid Search *can* work. Mostly when it doesn't it is because it is used incorrectly. Paid Search is not, for example, a good marketing tool for introducing a new business. PR would be more effective. That doesn't mean that paid search doesn't work at all - merely that it is often misused. We usually recommend a mixed strategy of organic search, community participation, CSR, PR and paid search. The precise mix depends on a lot of complex factors.

FWIW, I have some clients who achieve greater than 60% Click Through Rate (ratio of clicks to impressions) - this equals or beats the expected click rate on the first organic search result. Your personal usage is not representative of all market segments and all products. :)

Cheers, JeremyC.

Impressions aren't sales.

My girlfriend doesn't appear to distinguish between sponsored links and regular links. So long as the text appears indicative of her desired result, she'll click.

Now, that is for Google search results - she doesn't click on Google Ads, AFAICT, so comScore probably has some measure of truth to it...