Monday, May 21, 2007

How much do Search Engines remember old keywords and links?

This article isn't about Real Estate necessarily, as the ideas discussed can really apply to any web page, and optimizing it for search engines, but I'll use examples and lessons that I've learned from working on my folks Real Estate website to illustrate my theory.

When reading about Search Engine Optimization - methods for making web pages do better in search results - I've come across alot of comments that indicate that you should refresh, and otherwise update your content frequently. In response to this, I began to update the page titles for my folks listings on a regular basis, changing the keyword "house" to "home", "acres" to "acreage", and "lake front" to "water front" for example. Now I figured also that, since the search engines don't crawl and process at exactly the same time, that it might increase the odds of coming up in results for a wider variety of keywords. For example, if Yahoo crawls a listing when it says "Lakefront Home 30 Minutes South Of Jackson, MI", and then I change the title to "Water Front Property 60 Minutes West of Ann Arbor Michigan", and Google crawls the page, then between two search engines I've just associated the page for that listing with the following key words:

Lakefront
Home
South
Jackson, MI
Water
Property
West
Ann Arbor Michigan

Cache Keyword Smearing

Now have you ever noticed when you get search engine results in Google and you click on the cache instead of the actual website link, it shows you keywords on the page highlighted, but also words that don't even appear on the page, but rather are present in links that point to the page. Well that got me thinking that if Google is keeping track of keywords for that page that aren't even on the page, then aren't keywords that used to be on the page also data about the page that isn't on the page? In other words, just because I take the term "Real Estate" out of the title, possibly that history still helps Google weight the page - at least for a while - a little stronger for the term "Real Estate", even after the word is gone, especially when the page still says "Real Estate" in three other places.

Admittedly this "sticky history" idea is just a theory of mine, but as well as my folks humble website has done locally against many bigger players, it's one of my theories as to why. Since I haven't seen any reference to this idea out there yet, I'll call this Cache Keyword Smearing. I may be behind as I haven't had much time to read lately, so if anyone has read any other articles on it let me know.


Natural Link Smearing

Now if you like to learn about SEO (Search Engine Optimization) like I do, then your probably familiar with the term "Natural Link", which means that you put a hyperlink behind some text, but also typically the text of a paragraph, or some other larger stream of thought or communication. When you link that way, search engines - assuming everything is above board - tend to associate those keywords that are linked with the web page that they point to. So, if for example, I say I have one sponsor on one of my websites that has a really cool tool for helping people with arthritis open pill bottles , then the keywords tool, helping, people, arthritis, open, pill, and bottles all become associated with that site in any search engine that finds that inbound link.

Well back to Real Estate. I created an RSS feed for our listings, and this is what made me realize that I needed a "title" since title and article body tend to be two key fields in rss, and atom for that matter. For those people not familiar, rss and atom are two xml based standards for sharing data across the Internet. Anyway, I decided while I was at it that I might as well create two "titles" instead of one, since it wouldn't be much harder. Then I could use one title on the actual listing page, and in the page title, and use the other title for the rss feed. Now when most rss sites display an rss item, its typical (and courtesy) to hyperlink the title back to the actual page on the original site. So now I have some natural linked keywords out there on some rss feed sites with extra keywords that may or may not overlap the keywords in the page title. Finally, I go in to the listings every so often and swap the titles once in a while. I think of this as Natural Link Smearing.

What does a search engine do when title keywords and link leywords keep disappearing and reappearing? Would a search engine like Google actually remember the old title and link keywords and continue to weight your page on them to some extent, even after it's read the new keywords? What does it do when it sees these link and title keywords being swapped? Only the folks at Google know for sure, but with all the talk about the sand box - and the delayed effect getting stuff into Google, why would the logic work any different at getting stuff out?

No comments: