In her recent post, Rebecca discusses the issue of being forced, or choosing to, optimise for one search engine over others. From the comments I noticed something I have noticed previously - the commentators refer to the potential of the main three search sites only in terms of volume of traffic (market share).

I suppose this is due to many of the commentators' experience (greater than mine, I am sure) being of business models that are a simple numbers game - they work on percentages. If they get X visitors, Y percent of them will transact in some way that will earn themselves or their clients revenue.

However, I wonder if focusing solely on volume is always best. I work on a site for which we are presently having to rely heavily on PPC advertising to draw traffic. This is mainly due to us being almost non-existent on Yahoo and MSN and not being indexed in Google. Because we are paying for every visitor/click, I am spending a great deal of my time analysing which keywords are converting to sales from where and at what cost.

From my analysis I can see that for particular keywords, the cost per sale (CPS) on Google conversion is sometimes far higher than Yahoo or MSN - the CPCs are higher and sales conversion rates are often lower. This is such that we can make money per sale from Yahoo and MSN and lose it all (and more) trying to make the same sale on Google.

I won't pretend to know why the CPS varies like this, but I would guess that it's partly down to user behaviour on Google. Or perhaps it's because we get so many more Google users that we are seeing a higher volume of speculative clicks that won't convert to sales? It is difficult to draw conclusions because the numbers are not very high, and they certainly wouldn't be statistically significant. But it makes me wonder if we had equal visitors from Google, Yahoo, and MSN, which would convert the most sales? I am beginning to suspect that Google might be the worst, but that is only a suspicion at best.

How does all this relate to SEO?

Well, if you knew that Yahoo or MSN users had a significantly higher transaction rate, what would be the tipping point at which you would change your approach and optimise for them instead? Would you sacrifice Google volume for Yahoo or MSN reliability?