The next session I decided to attend is Search Term Research & Targeting, normally you would find Andy Beal and Dan Theis on this panel, but I bet this will be focused on European search term research.
First up was Christine Churchill from KeyRelevance, she is an admin over at Jill's HighRanking forum. She began to discuss what keyword research is and then described how too generic words are hard to compete on, and too specific words might not be searched on. She explained that logs files provide a good source of keywords for you to begin optimizing for. Instead of using company terms, try to think about how searchers will query your products, it often differs. She listed the adwords, overture and wordtracker tools as how she does keyword research, she cautions the audience that these numbers are inflated. She also recommends testing these keywords in a PPC campaign, its quick, and budgeted. SEO differs from PPC in that she targets more focused keywords for SEO whereas with PPC she goes broader.
Next up was Tomas Axelsson from Trellian to discuss there tool. I spoke with Tomas last night, he is an SEO who partnered with Trellian to resell the tool and represent them in Sweden. The tool he will discuss is Keyword Discovery (www.keyworddiscovery.com). He shows the keyword discover tool, then shows the KEI analysis, which was popularized by WordTracker. He then shows the real goodies within this tool, the season trend graphs - Trellian uses 12 months of historical data to graph this. In addition, they have a common misspelling tool, related terms tool, a keyword density report. He showed the regional breakdowns of searches based on country specific search, very useful here.
Tor Crockatt from Espotting was up next, who will focus on "user intent" as it relates to keyword research. She said, "always see keywords as question" - understanding the motivation behind the question. Multiple audience, meanings change over time, synonyms with differing user intent (vacation/holiday), related in theme but not in vertical (car/car insurance). Break down the meanings of the keywords you want to target. for example, "cheap web hosting" is 3 words with two meanings (cheap + web hosting). "Gebrauchtwagenleasing" is one German word with three meanings (used car leasing). Seven elements for keyword research; (1) Comparison/quality, (2) Adjective (price/product qualifiers), (3) intended use (4) product type (5) vendor (6) location and (7) action request.
Ola Svensson from Overture Nordics. Paid placement works because they drive traffic, quality traffic and converting traffic. He discusses the concept of "tail terms" with the overture keyword suggestion tool. He recommends segmenting through the buying cycle (information, shop, or purchase). Explore new opportunities/terms, experiment with new titles and descriptions, and optimize your efforts.
Q & A: Q: Why not show search terms with 0 searches in the keyword research database? A: They sometimes do, but often its not helpful for the Overture user. This happens with seasonality searches. They are moving to gathering more data towards storage of seasonal data. Espotting tries to predict search patterns before they hit main stream, most of it is seasonality but they change from year to year.
Q: Mike Grehan asks how skewed is the data based on the automated ranking tools being run 24/7 by SEO/SEMs? A: Espotting said they are not too bad, Overture kind of agrees, Christine agrees with Mike and says you need to look at the keywords on a conversion/roi level. Trellian said every keyword is some what equally inflated, so it balances out.