Reciprocal Links and Robots.txt Nofollow
Before SEOMoz, Rand offered reciprocal links, but thought he’d be tricky. He blocked the outbound link using robots.txt/meta robots so he got all the link value. Too bad it didn’t really work out! Many webmasters figured it out, and were not happy.
Buying Links for Clients
This was before Google came out and said they don’t like paid links. However, Rand was spending client’s money on buying them until he heard through the grapevine that they weren’t being counted.
H1 Tags with Keywords
This tactic doesn’t help much at all according to formal machine learning correlation tests. This means SEOMoz wasted a lot of time, energy and client’s money convincing clients to update H1s unnecessarily.
Don’t Use XML Sitemaps
When they originally came out, there was concern that XML sitemaps made it difficult to spot information architecture problems. However, the benefits content getting indexed ended up outweighing the difficulties.
Redirect Linkscape to Open Site Explorer
Recently SEOMoz decided to 301 redirect old Linkscape reports to Open Site Explorer. Good theory, but resulted in a massive loss of traffic. It’s been fixed with rel=canonical, but the damage has been done.