JLH Design Blog

I found an interview with Matt Cutts on Sphinn that I found quite revealing. Normally these types of interviews are very bland and boring where the interviewer gets to build up some sort of street-cred by getting a big name search rock star to talk, and the interviewee gets to spew the normal company spin. This one was a bit different, at least in the fact that I took from it some pearls that can be used in forums to dispel frequently addressed concerns.

Leaving the obligatory social interaction out of it, the highlights of the interview as I see fit are as follows:

1. Syndicating Content - When syndicating content to be published on multiple sources be sure to include a link within the content to the original source of the content. This will help transfer PageRank the syndicated content may get from external links to the original source. When Google is deciding what story to return in the results when there are many copies they apparently, ” a lot of the times it helps to know which one came first; which one has higher PageRank” On the other side of the story, if you are stealing syndicating content and want it to rank higher than the original then don’t include a link to the original source and get more PageRank to yours than theirs.
2. Supplemental Results - There are a couple of undocumented methods for finding your supplemental page count still working. At least one data center is now actually searching the supplemental black hole for 100% of it’s queries, with, “hopefully at more in the future”
3. Link Quality - “a link is a link, is a link; wherever that link’s worth is, that is the worth that we give it” .edu links do not count more than a .com link based on any specific weighting in the ranking algorithm. It just happens that many .edu links are naturally better than your average easy to get .com link. Additionally social bookmarking links follow the same guidelines and are not devalued based on their social network status, if they are weak they are weak on their own, without any help from a calculation.
4. Link Count - The old adage of 100 links per page is a bit outdated and a good example is DHTML throwout menus when many links could be seen on one page. Matt notes that a page with 5000 links would have it’s PageRank so diluted when it came to distribution that the links wouldn’t pass much. The question I think wasn’t answered here was in regards to the DHTML menu structure is that navigation like that tends to be site wide and Google is quite good at determining what part of pages is the template or site wide stuff, vs the actual content. My question would then be is PageRank flow just a simple division of PageRank by link count, or does more weighting go to the actual page content and less to sitewide navigation. Obviously a page with 1000 links (like a sitemap) isn’t user friendly as the designer should have provided a logical tree for the user to find the information, rather than have to read 1000 links and figure it out for them selves. Bottom line is that 100 links per page isn’t a hard and fast number, but keeping it reasonable still applies.
5. NOFOLLOW passing anchor text - In it’s early days there were some rare, and bug-like, instances where the anchor text of a nofollowed link was used in the search results. Those bugs have been killed. Right now, “At least for Google, we have taken a very clear stance that those links are not even used for discovery; they are not used for PageRank; they are not used for anchor text in any way. Anybody can go and do various experiments to verify that.”
6. Predatory Link Buying - Buying links for your competitor in hopes of hurting them is more than likely going to help them as Google is most attacking the link sellers. Kind of goes without saying, but I bring it up only because the original Webmaster Guidelines on the issue only addressed buyers and not sellers.

There’s a lot more in the interview.