This week's Whiteboard Friday features the return of Rand (woo hoo!) and his self declared biggest SEO mistakes. We screw up a lot here at SEOmoz (hell, they hired me), so we feel it is only appropriate to take this opportunity to share what we have learned in an effort to prevent you from making similar mistakes. SEO is complicated. The best we can do is practice, work hard and compare notes.

1. Reciprocal Links + Robots.txt NoFollow

Back before the formal SEOmoz days, Rand used to dabble a bit in some grayer areas of SEO. The first mistake he mentions is a tactic involving offering reciprocal links but blocking the outbound links via robots.txt/meta robots so that he could get all of the link value. This tactic didn't really work and he ended up having to do a lot more work to get in the good graces of the webmasters he had fooled. Head smack!

2. Buying Links for Clients

This tactic also took place before formal SEOmoz days. At the time, Rand spent client budgets on paid links. This is a bad idea because the value of the links can't be determined (was Google even counting them?). He later found out through Google employees that the links were not being counted and that they may actually be hurting the client's site ability to rank. Oops!

3. Recommending People Use H1 Tags with Keywords

This mistake is a little bit more subtle. For years, SEOmoz recommended including keywords in the H1 of pages. After we started doing formal machine learning correlation tests we found out that this tactic didn't actually help very much at all (including the keywords in normal text in bigger fonts worked essentially the same). This was a shame because it meant we wasted time and energy convincing our clients to update their H1s.

4. Recommending People Not To Use XML Sitemaps

When XML Sitemaps first debuted, Rand and SEOmoz recommended not using this. While the idea was sound in theory (having a XML Sitemap can make it difficult to spot information architecture problems) the observation ended up being outweighed by the impact we saw with the increased indexation rates of sites that employed this tool.

5. Incorrectly Redirecting Linkscape to Open Site Explorer

Recently we decided to 301 redirect all of the old Linkscape reports to our newer, better converting, Open Site Explorer reports in a 1-1 relationship. This was in theory a good idea but unfortunately including various tracking components on the redirect URLs resulted in us losing a significant amount of traffic. We later fixed this with rel=canonical but a lot of the damage was already done. Ouch!

Do you have any lessons you have learned after making some noteworthy mistakes? If so, we would love to hear what you learned in the comments below.