Hacker News Reader

Comment by jrockway | Hacker News Reader

SEO tips seem a lot like black magic to me. Nobody ever explains why, and nobody shows any data... they just say things you are supposed to believe. I find this annoying.

(I have never focused on SEO, and my website always comes up for terms that I expect it to come up for.)

View on HN

The people selling black magic tend to use techniques that end up getting them busted in the long term. But there are some things that are just good practices from an SEO perspective and have nice usability and accessibility benefits.

There are companies that have the data to back up the statements in this and similar articles. They just don't give it away for free. :-)

There are companies that have the data to back up the statements in this and similar articles. They just don't give it away for free. :-)

Not just directed at rgrieselhuber so much as the whole community, what would you say are the predominance of such companies relative the whole SEO market?

My experience so far has suggested a vast number are just floating on borrowed time until someone busts them, but I'll be the first to admit my sampling might be biased.

From what I've seen, the industry as a whole tends to get a bad rap from the people doing stupid stuff. My experience comes from working with global brands so it's probably skewed, but there is real business value to be derived from doing a few things right.

I end up having a lot of people ask me "SEO" questions which basically just amount to understanding PageRank and the basics of Google's ranking-fu. What I've seen is that the desire for SEO comes from and ok place -- people just seeing page ranking as black magic and wanting someone to demystify it. Nerds tend to take for granted understanding the basics of web search. A surprising number of people in the web world have no basic grasp of the concepts, but still know that showing up in search results is important for them.

The "why" inevitably boils down to "we have found this to be effective while black-box testing a constantly changing algorithm developed by an adversarial company with many, many more engineers on staff than we have".

I show data, occasionally. Most people think it is competitively sensitive.

I'd be happy to say more (why, data, etc.) on anything in particular. The post was already getting a bit long, which is why I left it at that. There is some why in there already though.

I was mostly interested in this one:

"Use directories instead of subdomains."

I am pretty sure that this is because Google sees different subdomains under the same domain (ie. a.example.com, b.example.com) as different websites, and so if you got a link to a.example.com, maybe a.example.com's pagerank would improve, but example.com's pagerank will not improve.

This is why choosing one main domain for you website (either www.example.com or example.com) and 301 redirecting all requests from the one not chosen to the chosen one is recommended.

This is true. This is why we decided to set it as http://blog.dawdle.com instead of http://www.dawdle.com/blog . We have some freelancers that want to write about gaming in general (so they might get a callup to get paid), so it's editorially independent, and the blog can pass independent PageRank to the main http://www.dawdle.com/ domain.

The point is: don't automatically just shove everything in subdirectories; think about if you want your subdomains to have independent PR or not - it might make sense in certain situations.

On subdirectories vs. subfolders, see http://www.seomoz.org/blog/understanding-root-domains-subdom... - essentially, it's about benefiting from the authority/trust/link metrics assigned to a root/sub domain rather than starting a new one.

Like I said, I'm interested in data, not speculation. Reading this is like reading a review of a power cable that makes your speakers sound better. People can convince themselves of anything they want to.

Reply on news.ycombinator.com