• Brainstorm with colleagues-ask them what terms they think your potential customers would use to find your stuff.
• Ask your customers and prospects what they use (or used) to find you on the web.
• Analyze your web traffic statistics. Your stats should show you what keywords/search engines your visitors are using now and, more importantly, show what you're missing.
• Look at the keywords that your competitors use (e.g. in their site title tags) if they rank highly on top search engines
• Use both Wordtracker (a pay service) and Overture (free tools available) to analyze the relative popularity of keywords that are actually being used, discover other possible keywords, and see what competitors are bidding and what their "ads" say

• Using no keywords at all on the right places of their site (their title tags and their html content toward the top of the page, links on their pages, and their site map).
• Assuming that your company name only is THE only important keyword-even worldwide brands need to use other keywords in addition to their name.
• Ditto for your "brand" product name(s), only. Use them, but think beyond them
• Assuming buzz words that you dreamed up to describe your business is the stuff that people type into search engines. Rarely.
• Assuming that there are only a couple of keywords that describe your stuff. Wrong. Likely, there are dozens and dozens or more (depending on the range and complexity of your biz)-with some being a lot more popular than others to web surfers. Bid for a bunch of them, not just one or two.
• Assuming that few people aren't searching for your stuff. Possibly, but you can literally check that out (using Wordtracker and Overture tools) before you make up your mind and make any investment at all in pay-per-click.
• Not writing good keyword/search engine "ads." You get very few words (no hype!) to help encourage the audience(s) you want to click into your site.
• Not bidding for one of the top 3 spots. Otherwise, you'll get a fraction of the hits on your site as a result.
• Thinking that if you don't have an e-commerce site that keyword work and pay-per-click are wasteful endeavors. With e-commerce pay-per-click is almost a no-brainer. If you're not using it, you're missing out. But even if you don't have stuff to sell directly online, you could be very surprised. Pay-per-click can also be a valuable tactic for businesses with good information commerce sites
• Not having a site with compelling content and/or things to sell-and a good way(s) to capture basic prospect information (email, phone, address). Otherwise what's the point?And a corollary:not immediately following up on leads that are generated via your website. Unless Sales uses those new leades and follows up by phone, email, etc., quickly, again what is the point?
• Thinking that paying a dollar or five or ten dollars per click, can't possibly make much economic sense. Oh really? Some people are aghast now that keywords in competitive "spaces" cost several dollars per click (and many good keywords, by the way, still cost only $0.15 a click). "$5 for someone to click into my website? No way," they say, "would we throw money at that." But what I've seen is that keyword bidding is the perfect free market economy. The online marketing pros out there know that their bids make good economic sense-or they wouldn't play-and some spend hundreds of thousands of dollars a year just on pay-per-click. Take your clue from others and then do your own initial ROI estimates and monthly budgeting (You can also use Overture's ROI tool).
• Not staying the course with your pay-per-click (making reasonable adjustments along the way) for at least 3-4 months with an e-commerce site or as long as your typical sales cycle--if yours is an information commerce site. You will generate more good prospects, but that does not eliminate your sales cycle.
• Not monitoring what's going on with your keyword bids, regularly. For some, you may need to look every few days-and others at least every couple weeks. Unless you change your bids and pay attention, your rankings will drop and/or your investment could sky rocket (you can set up various budget controls on both Google Adwords and Overture).
• Not using a pro to help you get started revamping your keywords, website content, and a pay-per-click campaigns. If you've got the time and motivation, you can learn by trial and error (and there are good references out there to buy), but there are experts who have experience and can get you better results, faster--while you focus on other things.

• You've got a good website (minimally with compelling content and a way(s) to capture prospect information)--required.
• There is some "reasonable" traffic for multiple keywords related to your company, products and services (that is audiences use the web to search for your stuff)-required, but the definition of "reasonable" will vary per size of business.
• You sell "niche" products or services--good, but not required.
• You sell globally, nationally, or in a large geography--good, but not required.
• You sell directly online and have e-commerce functionality--good, but not required.
• You sell in a space where competitors haven't caught on yet to pay-per-click. Bingo! There aren't too many out there, but if so, the keywords will be very cheap. You've got to play.