SEO observations in 2010

17/03/10 posted by Michael Sharkey

The search engine landscape has changed quite a bit in the last 12 months with the introduction of Microsoft's Bing and Google's strong move towards social media integration with real time results.

So what overall effect does this have on your web site, bottom line and more importantly the ranking of your web site in search engines?

Crawl rates have been tweaked

It appears Google has changed the rate it crawls web sites and the depth of which it crawls. According to an interview with Google's Matt Cutts, the crawl rate is proportional to the page rank of the pages Google is trying to crawl. For example, Google is more likely to crawl, and crawl frequently, pages that have many inbound links or higher Page Rank.

The suggestion was made by Cutts that web site owners link to the most important pages from their home page to ensure those pages are crawled more often.

My personal observations have seen dramatic decreases in crawl rates on the sites we work on. In particular, our own web site which publishes fresh content daily has seen a significant reduction in the pages Google is indexing. We have overcome this problem by feeding out new content out to social networks such as Digg and Twitter. In return the most important parts of our site (the new bits) are being crawled again.

Cutts referred to server capacity as being a factor that determines crawl rate. The higher the server capacity the more likely Google will spend time crawling deeper into the site. In particular it should be noted that a dedicated server is still the best option for larger web sites.

Session IDs are bad!

Any smart developer has known session IDs in URLs suck. For everyone else session IDs are appended IDs on URLs e.g. http://www.website.com/ecommerce/something.apsx?session=34t65757hkjdzgskrugheatk4358u569857. They are ugly, horrible and make site URL structures extremely unaccessible.

Google doesn't like session IDs and recommends not using them. To quote Matt Cutts's interview, "Don't use them. In this day and age, most people should have a pretty good idea of ways to make a site that don't require Session IDs. At this point, most software makers should be thinking about that, not just from a search engine point of view, but from a usability point of view as well."

What the hell is rel=canonical?

Ever since rel=nofollow was introduced it seems Google has been moving away from their robot style days to actually allowing developers to assist them with the crawling process. To get everyone up to speed rel=nofollow is an attribute that a developer can use to stop Google crawling or giving weighting to another page.

rel=canonical is a different ball game. The tag helps sites that use ugly session IDs (see above) or has duplicate content through URL structure to specify that the content is duplicated through URL combinations and to preference one particular version of the content. Confused? I certainly was when I read it. It makes perfect sense though and is a chance for older web sites to lift any duplicate content penalties Google may have already put in place.

If I made no sense a full explanation is provided here.

Getting hard on link spam

For quite some time Google has said they were, "getting hard on link spam". I'm still not sure what that means because personally I haven't seen many improvements. To prove a point I set out late last year to run a test on our own web site. I set myself a budget and went around the web buying extremely clean one way links with my primary keyword as the anchor text. I got about 5 decent paid links from related sites. In Google's eyes this is black hat (a bad, punishable technique). The results however were quite un-punishable. In 4 weeks we went from not appearing to page 3.

Many sites in the market continued to purchase paid links from high ranking web sites to improve their rankings. I'm not sure how Google will ever overcome this and distinguish from genuine and non-genuine. It seems impossible and I highly recommend some paid links as part of your overall SEO strategy.

In summary

So what strategies should everyone be looking at implementing in 2010? My pick:

• More quality content: a page a day or bust
• Paid links: high quality paid links (shh don't tell Google)
• Better URL structure and hierarchy: down with session IDs
• Most important first: if you can't navigate to the most important parts of your site from your home page Google won't.
• Twitter, Digg, Reddit - don't forget to share your content

Get to it SEOers!