There's a lot of quantative measurement of the search market, including from Efficient Frontier, but as with all data points you should consider what the numbers mean, if they mean anything at all, before pronouncing judgement.

Take for example comScore's search engine rankings report for February, which declares that search queries are down across the board.  Jefferies & Co analyst Youssef Squali is quoted as saying, "Google's month-over-month 5 percent decline is a little surprising, but all of the major Web search names were down." The comScore data showed Google having 6.14bn US queries in January, down to 5.86bn in February.

February is described by comScore as 'seasonally soft'.  You or I would describe February more simply as 'shorter'.  Two days shorter than January, given that it's a leap year.  Yes, that's right – those search query numbers don't take into account that February is 6.5% shorter than January.  In that context, a drop of only 5% doesn't seem so bad does it?  In fact, it means that the average number of queries per day has increased in February.  Panic over!

Credit to Patrick Altoft at e-consultancy who has a sharp eye for this kind of nonsense.

I was astonished when I read an interview with Proctor & Gamble’s Marketing Director, Roisin Donnelly, in NMA. This was the part that really knocked me back:

“A big barrier [to spending more online] is measurement,” she said. “We know how to measure TV, we know what the ratings are. Online has been slower to set up measurement systems, even though ISBA is working on it now.”

It would be a lot more accurate to say:

“We have a wild stab in the dark when we measure TV, we know what the ratings are, give or take 75%”

TV measurement is laughable compared to the data that comes from online. I understand that there’s still nothing like the telly to deliver massive reach, but don’t pretend that the numbers that come from the BARB panel are in any way accurate.

Chris Lake from e-consultancy does a great job presenting the case for the defence.