<< by on June 7th, 2007
Over the next few days and weeks, I’m going to be doing some in depth research into how Google’s new Universal Search is determining rankings. The first page element I decided to delve into is images.
Several queries using Universal Search bring up images either at the top or bottom of the regular index search results, and this seemed like it might be an easy way to get instant visibility at the top of a particular search. But I’ve found a few interesting things out about these images in my research:
- The three images that appear in the regular index through Universal Search NEVER match the first three results in Google Image Search. Therefore, it’s clear that the algorithm to rank images in Image Search differs from that of the Universal Search algorithm for image ranking.
- While Universal Search is designed to produce relevant results. However, in many cases, the images shown in Universal Search are rarely the most relevant images. As an example, a search for the term “dogs” in Google produces the following image results at the bottom of the search page:
This result seems perfectly normal and appropriate, until you click on the first result — the pencil drawing — by ADOLF HITLER. There are many great images of dogs in the Google Image Results, but apparently, Google’s Universal Search algorithm for images deems the pencil drawing by Hitler as the most important image result for dogs. This brings us to point number 3…
- Consistently, search after search, two of the top three images presented in Google Universal Search were from .org sites rather than .com sites. Many of us in SEO have known for some time that Google’s algorithm consistently gives greater weight and rank to .org websites than .com websites. Could this also be true for images in Universal Search? It’s looking like it may be that way. More testing will give us a more conclusive answer.
Keep posted. There will be more to come on our findings in the next few days and weeks.