Estimating potential traffic before construction of a site.
I am making this post in response to the excellent post SEO and Feng Shu.
In the thread eduardo lorenzo mentions about the need to be able to judge how traffic will increase from better SEO etc.
Much has been written about judging how many more backlinks are required to get into top 10 rank for certain terms. But little has been written about estimating increased traffic.
Ussually you wish to do this before construction a new site, to judge economic viability, or when improving an existing site.
There are 3 methods I use
Wordze , tool which get SE data from various ISP. It is a very good service, but unfortunately it is prone to regionalism (as in where the ISPs are located from where they get thier data).
overture yahoo’s SE word tool
These 2 tools are fairly simple to use.
Adwords Traffic Estimator Sandbox , how can you use that for traffic estimation you may wonder. If you think about it the data for the traffic must be coming from google itself. Therefore it will be fairly accurate.
Lets look at some data
Keywords Estimated Clicks / Day
buy viagra 358 448
order viagra 107 133
buy altace no data
Also very handy is you can get traffic estimate from regions 
For the term “buy viagra” for the number 1-3 position you are looking at ~400 clicks a day. But bear in mind this is for “advertised results”, organic results will get a higher CTR, probably 50% more, so in this case 600 clicks a day.
“order viagra” is only good for like ~150 clicks per day. So if possible you are better off to optimize for “buy viagra” instead of “order viagra”.
Knowing that pharma affiliate converts at approx 1:40 you are looking at 12 sales a day if you rank for that term. Knowing how much you make per sale you can estimate income
If you have existing site, using this you can see which words you should optimize for. Same principle applies for long tail, it better to optimize for “buy viagra in newyork” then lets say “order viagra in newyork” since the root “buy viagra” will get more search then order viagra.
Making a website to sell altace probably is a waste of time
Before going into new niche do your homework, check out if niche is viable.
No point making nice website if no one searches for it.