This is a guest post from Ted Rheingold, who will be speaking at Startonomics about word-of-mouth marketing on Oct 2. He blogs about his business at blog.dogster.com.

Between September ‘07 and August ‘08, our monthly search traffic to Dogster.com went from 196k to 461k visits.

With web stats it’s very easy to maximize your existing search traffic by simply analyzing what content is driving traffic and expound upon that content. But how do you know what topics would drive search traffic to your site that you don’t already have?

Your options are limited, but there’s enough to get started. The newest and best free options are offered naturally enough by Google. Google Trends lets you see how often any word or phrase is being searched on Google. If your site is about pine trees you might want to say it’s about conifers as that term is searched on more frequently. Even better is the Google Adwords Keyword Tool which displays search counts for each term and also, cleverly, shows you related search terms you may not have considered. A third place to find new keywords that could drive traffic is Quantcast where you can see other sites that your visitors go to. Then, you can see the popular search terms for those other sites. Finally, for a fee, Compete.com Search Analytics will report how your site ranks for searches on particular keywords. This knowledge can be gleaned from doing your own manual searches, or using a tool like the nifty Rank Checker, but you’ll have to keep track of all the data points over time.

So, in using the tools above you have come up with a list of new keywords and phrases that you think could drive traffic to your site. Historically, the only way to really know how well a new keyword would perform was to make great content around these keywords on structured pages that highlight the keyword in ways that search engines like. Once these pages are live it becomes a waiting game. First, wait for the search engines to index those pages and then wait for those pages wait to slowly grow until they hit their natural rank, or wait for your white hat tricks to get it an even higher rank. This can be quite an investment in time and money. It’s reasonable to have to wait 3-6 months to see how well a new page on your site is going to do. This is especially true if you are in a deeply competitive space. And of course this gets even more complicated as over time your business changes, your customers change, the competition changes and the search engines change.

Here is where Search Engine Marketing (SEM) comes in. As you know SEM is when you pay to have your site promoted on search engine results pages. The net effect is that your link or ad appears on the top of the very first result page. The normal use of SEM is to pay to get people to go to your site and your goal is to find the terms that drive traffic to your site that have the lowest purchase price from the search provider. But it’s also a handy tool to learn how prospective customers will respond to your site when they search on certain keywords. If you have a piano store you can test how interested people are in the terms ‘grand piano’, ‚’baby grand’, and ’sheet music’ without having to wait the 6 months to see if people are interested. Thus you can quickly buy this knowledge and skip producing special pages and waiting for them to get top ranked (if they do at all) to know if people like clicking on them. In fact, you can run through hundreds of phrases in a very short period to very quickly find out what terms are most likely to get your site checked out when people search on them. Why sit around and guess what is going to work best when you can simply find out?

In Q2 2007, we started aggressively using AdWords and YPN to market our websites Dogster.com and Catster.com. Our goal was to get new registrants for under $4 (our APRU at the time.) After trying out a massive spreadsheet of terms we found terms related to adoption, rescue, health questions and informational resources could drive new members to our service for under a $2 acquisition cost. We very quickly maximized our spending on such terms, but found we couldn’t create the deep, rich, related content that would keep these new members and visitors happy. So we immediately cut-off our SEM campaigns and went from being purely community sites with typical community-created content to licensing, authoring and re-packaging community-created content areas that offer expert and authoritative content on all related topics. We bought whole books and put them online. We made a directory of 100k pet friendly business and locales. We created Dog Answers and Cat Answers where each page is a specific question with helpful answers. We wrote descriptions of all major dog breeds and cat breeds and hybrid breeds. Much of these pages flew up the search rankings because they were the good content that we knew people were looking for and thus would get highly ranked by the search engines. We saved a lot of time by knowing for certain what searchers wanted.

In summary, use tools like Google Trends to help create a list of terms/topics/phrases to test. Then use an SEM service such as AdWords to test how well these terms correlate to your site and at what cost you can acquire those users. Take the highest performing keywords and make sure your site has a great content each keyword (and for retention, a great user experience). Watch organic search traffic grow and consider continuing to use SEM if the acquisition costs work for you.

[Note: if you're wondering why our search traffic cut in half in June '08, that's what happens when a robots.txt file that blocks all search spiders gets inadvertently pushed from your staging to live server =! ]

Heard enough yet? Join us at Startonomics