We're getting better and better at predicting the success of our search marketing efforts in advance.  I'm just geeky enough to find it fun to dust off my high school algebra skills and crunch numbers.  Here's the scenario that a prospective client recently challenged us with.

They said, "We've got $2500 to spend on clicks with PPC.  Our goal is a $100 cost per sale or less, and we typically close 70% of our leads."  That's it -- that's all the information we had to go on, but it was enough.  Here's how I was able to use that information to predict whether or not we could be successful:

• I put together a quick and dirty sample PPC campaign on Google.  First, I used Google AdWords' Keyword Tool to generate a list of about 180 relevant keywords.   (Note that a real campaign would contain thousands of keywords -- this is just a starter list to give us rough data.)
• I downloaded the list of keywords into Excel and loaded them in Google's Traffic Estimator, using the $2500/month budget.  This allowed me to play around with different bids (maximum costs per click, or CPCs) to see how results varied.  Raise the max CPC and the average position/ranking increases, which can generate more clicks.  But it can also generate fewer clicks because each click is more expensive, meaning the budget buys fewer clicks.  So, by lowering the max CPC, I was able to find a "sweet spot" of $1.50 max CPC that generated more clicks -- 1290 -- than did either lower or higher max CPCs.  (Note that the actual CPC paid is predicted to be $1.13, and that Google estimates are often off, but it's a start.)
• How many sales would we have to generate to meet our goals?  At a $2500 spend and a $100 cost per sale, we'd need to generate a minimum of 25 sales.
• How many leads would we have to generate to meet our goals?  At a 70% close rate, we'd need to generate a minimum of 36 leads.
• What kind of conversion rate would we need to generate 36 leads from those 1290 clicks?  Answer: 2.8%.  We know, based on past experience, that this is very do-able.
• If you're really paying attention and breaking out your calculator here, you may notice that a $2500 budget and $1.13 actual CPC should buy 2200 clicks, not 1290.  By adding more keywords to the campaign (beyond the preliminary list of 180), we should be able to do this.  To generate 36 leads from 2200 clicks, we'd only need a conversion rate of 1.6%.  This is even more do-able than 2.8%!

Bottom line, when a prospective client can give us just a few basic pieces of information, we can fairly quickly tell them whether or not our search marketing services can help them reach their goals.  They know what to expect, and we know what to shoot for.  It's a win-win, and it doesn't even require statistics or calculus!