Back links are of serious importance for getting good Google rankings. They establish site authority and boost your site’s perceived importance to particular search engine queries. I’ve seen sites that weren’t at all targeting competitive keyphrases in their page copy, but were able to rank for them by virtue of keyword rich text links from authoritative sites. So these things are important, but I’ve never seen any studies, or even much discussion, about how people naturally link to sites. Here are my common sense observations about it, and how you can expect search engine listings to be affected by it.

Types of Rankings

For a site, I see a couple ways top rankings get acquired:

1. The site gets a lot of links to its home page from authoritative sources, with lots of keywords, or maybe no keywords, in the linking text. The site links to internal pages with keyword text, and by virtue of its home page links (authority), starts ranking for those terms it’s targeting on the internal pages.

2. Keyword rich links point directly to a page causing that page to rank for the keyword phrases in those links, especially if they are the same keywords or similar to the keyword phrases the page is targeting.

There are lots of details and variations left out of these two observations. How 1 or 2 are weighted … if they are weighted … where “unnatural” linking comes in – these are questions I have, but don’t expect to get answered. You can also expect Google might take the time to identify themes of pages or sites that link to another page or site to help determine rankings, but I suspect this would be of much lower consideration than points 1 and 2 above. With all the documents on the web, it’d probably be difficult to take the time to identify and make calculations based on themes. Search engines move slow - they process a lot of data – so if this was something they were doing, you might expect it’s only done for certain sites or certain queries.

Commercial Vs. Non-Commercial

For a company site, it seems to me, excluding blog content or something that seems very purely informational, your links will tend not to include keywords, as far as natural linking goes. I do a lot of unnatural linking for companies, and even if I explicitly request links that contain specific keyword phrases, it’ll be a lot more common to get someone linking to the site with generic text or the site’s brand. Company sites are typically identified by their brand. If I want Pure Visibility to rank for a phrase like “internet marketing firm,” I’d try to acquire links with this phrase in it, but it probably looks nicer and makes more sense for other sites to link to it with “Pure Visibility” as the anchor text. So websites that have an original brand name, will typically rely on building authority and then using keyword targeted internal linking to rank well. That’s more like how it’d occur naturally, but SEO is a lot of battling against what’s natural. Some sites out there still use the tactic of including keywords or exact phrases in their brand to rank well. This was more effective in the past, but probably still works to some extent now for exact phrases. In this case, the brand and the keyword phrase being targeted are the same, so the division between points 1 and 2 is blurred.

Informational resources are probably a lot more likely to get an exact keyword phrase as a text link than a company is. These links would be more likely to look like a citation. For example, if I’m discussing internet marketing, and I’m using niche terms I don’t feel like defining, I might just link that term to a Wikipedia entry. If I’m discussing an article from a generally known source, I might link to that article with its title. These informational resources have a bit of a boost for getting keywords in anchor text, because of how people behave (or how I think people behave). These sources are typically called non-commercial. At least, that’s how I refer to them. In no way are they actually non-commercial. A lot the information is ad based, or supported by a company – that hardly makes it non-commercial, but I’m abbreviating I suppose. They are non-commercial in the sense that they are informative, opinionated, and not directly focused on a product or service being sold.

There’s a practical reason to believe that Google favors non-commercial sources for a lot of queries (in the wacky sense I defined above). It’s done algorithmically based on the linking behavior of ordinary folks. Any search engine that uses link text (and is resistant to artificial “bombing” of the results) should favor “non-commercial” sources. That’s the theory I’m sticking to for now, at least.