The article emphasizes a need for search to get more social and better understand intent. In my view, understanding content is much more important.
Example: What do I want when I search for a product review?
I want the top few hits to be:
* reviews of the exact product I'm interested in
* well organized, well written, and thorough
* mention pros and cons
* if appropriate, compare to other competing products.
What do I get?
Spammy sites that have mastered SEO (primarily links)
Very popular sites that have shallow reviews
Very popular sites that just have a few sentences and a link to another review
Product pages for Amazon or other online sites.
The actual best reviews are hard to find. They are occasionally on large well established sites like ars technica but more often than not are on small niche specialist sits or small blogs. One is lucky to find even 1 such link in the top 10 search results.
Social can help with this type of query (i.e. the Amazon review ranking system) but not every kind of query is amenable to ranking in this way.
What a human can do that search algorithms currently can't do is determine which reviews are of high quality. In other words, understand and evaluate the quality of the content. Perhaps this is currently too hard to do, but I'd love to see search move in this direction.
It's easy for SEO specialists to fake link strength and perhaps even social popularity. Much harder to fake quality, I would think.
Example: What do I want when I search for a product review?
I want the top few hits to be:
* reviews of the exact product I'm interested in * well organized, well written, and thorough * mention pros and cons * if appropriate, compare to other competing products.
What do I get?
Spammy sites that have mastered SEO (primarily links) Very popular sites that have shallow reviews Very popular sites that just have a few sentences and a link to another review Product pages for Amazon or other online sites.
The actual best reviews are hard to find. They are occasionally on large well established sites like ars technica but more often than not are on small niche specialist sits or small blogs. One is lucky to find even 1 such link in the top 10 search results.
Social can help with this type of query (i.e. the Amazon review ranking system) but not every kind of query is amenable to ranking in this way.
What a human can do that search algorithms currently can't do is determine which reviews are of high quality. In other words, understand and evaluate the quality of the content. Perhaps this is currently too hard to do, but I'd love to see search move in this direction.
It's easy for SEO specialists to fake link strength and perhaps even social popularity. Much harder to fake quality, I would think.