ArtFire SEO RoundUp March 12th
Posted by ArtFireSEO on 03/16/2010 at 16:37:27
comScore Search Market Shares: The February search market share numbers are in from comScore. Google gained back some of what they lost in January. As in previous months, Yahoo lost market share and Bing grew to make a combined MicroHoo search market share similar to previous months. In related news, the popular Yahoo Site Explorer feature that many search engine marketers assumed would be lost with the Microsoft and Yahoo search alliance has at least been discussed by the two companies. Site Explorer allows users to take a look at the links (from Yahoo’s index) to any site.
Yahoo Updates: Not much substance to this update but I’m keeping an eye out for what changes may have been made. Yahoo’s search blog announced updates to its indexing and ranking algorithms. No hints as to what those changes are, but if we see anything of interest I’ll make sure it makes its way into a future SEO roundup.
SEO Tip: There wasn't a lot of Search Engine News this week, so I wanted to share an SEO tip related to optimization testing. Many of you are probably familiar with the terms, “A/B” (changing one variable) and “multivariate testing” (changing multiple variables) when talking about optimizing pages for search or conversation. I often say that the single most important thing you can do for your business, as it relates to search, is to test SEO changes to find what works best for you. I haven’t, however, done a great job of suggesting HOW to test, and for that I apologize.
If you are going to run an SEO test for your items, I would suggest that you only do A/B testing. A/B testing is where you have your base point (what you’re doing right now) and you test a single change. Maybe you try changing one keyword in your title, or you use a synonym of a word. You could change the name of a color to a different yet very similar color (color names are very subjective), or you could try changing one paragraph in your description. Try to keep the change to one variable and leave the rest of the listing the same. Once your test is over, if you have significantly different results you can instantly pinpoint what made the difference (the one change that was made).
Multivariate testing (testing multiple different variables all at once) can be very tricky. If you have two very similar items and your test is different photo angles, different title structures, different keywords, along with different ways of describing them, and you find that one items gets more traffic/interest your left with a question: Was there one variable that made the difference or was it a combination?
One final word on listing tests; if possible, run your tests between similar items. Similar products with similar prices will give you better test results because the actual differences in the listings are the most important differences between the items. That’s not to say that you should only run a test on two items. The more items you’re testing the better the results, however, running a test between necklaces and bracelets won’t help you as much as testing half your necklaces against the other half and likewise with the bracelets.