Search Query Keyword Insight
eMarketer's newsletter today says that to a recent report from Oneupweb search engine user clicks tend to present higher conversation rates the longer the keyword phrase.
Quote:
Looking at data from high-traffic keyword phrases for sites that were optimized by Oneupweb, the search engine optimization firm found that conversion rates grew the longer the keyword string, peaking at four words. The results were the same in all three months studied.
More information here.
- Y! MyWeb
Holy Conversions Batman!
Man, if I could get the kind of conversion rates they quote in that research, I would be able to retire by the end of the year! I don't know what their client is selling, but I find those conversion rates a bit hard to believe. The average conversion rate for most sites that I am aware of is around 2-4%, which is about what I do. But they are quoting conversion rates from 4.75% - 38.28%!!!!
I must be selling the wrong stuff, either that or my website blows! ;)
Got those sorts of conversions at christmas
But yeah, ordinarily it is less than 10%. I hear some folks though do double figure conversions on one-page-wonders using JVs so perhaps they have really good targetted traffic?
What is a conversion?
I think one of the problems here is what is a conversion? If you look at this company's clients (and eMarketer is one) what do each of those consider a conversion?
Does eMarketer consider a conversion someone who visits their site or someone who buys one of their $1,000 reports? Does Universal/Motown consider a conversion someone who buys a record or a DVD?
The number of words will always give a more relevant result. One of my clients who has a news site for new industrial products does see this on an ongoing basis. If someone is looking for a 1/3 hp stepper motor and the user is looking for that they have a better chance of getting it than typing in "AC Motor" "motor" or even "stepper motor". They also have to be careful to find the motor on it's own and not a dishwasher which has the 1/3 hp stepper motor as one of it's components and listed in the description.
All in all, although the report was interesting it was kind of stating the bloody obvious.
Conversion rate theory
Unfortunately, there is no any theory that would help us to understand how plural/single, number of words, or words order would affect the conversion rates.
Let's consider an example: which keyword would delivers better conversion rates:
"Hotel Tokyo"
"Hotels Tokyo"
"Hotels Ginza"
In one study we found that Hotels Ginza has 200% conversion rates compare to Hotels Tokyo. In other study it was 50%...
There are many other questions that nobody can answer today:
How robust is conversion rate over the time? What is the normal fluctuation? How much political events, seasons, weather, etc. affect conversion rate?
I think that we are at the very early stage of learning process: we are collecting knowledge and mythologies. I hope to see sometime in the future a decent University study related to conversion rates theory.
Hope that somebody will get Nobel price for making conversion rate prediction… similar that happens in a theory of stock options thirty years ago.