Link Building in Theory - A Look at Unlinked Citations

Does a Citation Have to be a Link?

[martinibuster]

Thought provoking article by Danny Sullivan over at SearchEngineLand entitled, Google: Now Likely Using Online Merchant Reviews As Ranking Signal [searchengineland.com].

In it he cites the recent NYTimes article (Bad Publicity is a Good Thing [nytimes.com]) about a merchant who bragged that his negative reviews helped boost his site to the top of the SERPs, and a Google blog post [googleblog.blogspot.com] in response where Google states they tweaked their algo so that negative reviews no longer help rank a site. Danny calls this tweak, Sentiment Analysis. The article does not say that linkless citations are driving high rankings, only that it is likely that linkless citations may be a factor.

It seems likely that Google is now using these reviews as part of its ranking algorithm... That doesn't mean reviews necessarily override all other ranking signals but rather that they are yet another factor among many to be considered.

Reviews have been a component of the local search ranking algorithm, this was a hot topic of informal conversation at the recent Vegas PubCon. Some suggested hiring link monkeys to visit all sites that are cited in Google local and having them drop reviews and site mentions.

Linkless Citations Beyond Merchant SERPs
Let's expand this discussion to beyond merchant reviews. After all, if this works for merchants, then maybe it would work in the general SERPs. For instance, some people claim that no-followed links help a site rank better. Could it be it's the citation itself and not the no-followed link that could be contributing to ranking a site? It is a citation, right? Citations are the foundation of PageRank.

PageRank is about Ranking Citations
The foundation of PageRank is ranking citations. The 1998 Stanford paper is called, The PageRank Citation Ranking: Bringing Order to the Web [ilpubs.stanford.edu]. Citations have normally been thought of as links. However it's clear that Google uses linkless citations to rank merchants for local search. Google is so hungry for citations they even dig out citations from JavaScript.

• Is it far fetched to assume the possibility that Google has moved or may be moving toward including other forms of linkless citations as part of the ranking algorithm?
• And if so, what other forms of linkless citations could they be using?
• Is it time to at least provisionally redefine what constitutes a citation?

[IanTurner]

Hotlinked images is an area that would be easy to identify and use as citations.

I'm fairly sure they are used by image search - but whether they could affect the main rankings is another area.

[aristotle]

If scrapers copy your articles, wouldn't that be a signal that they contain useful quality content? Even if the scraper doesn't say where the content came from, Google might still be able to determine the original source.

[lfgoal]

I can "get" that a nofollow link may have value because even if a link isn't being endorsed, it's being cited for whatever reason and, therefore, it may be relevant to a topic that's being discussed or searched upon. But linkless citations...wouldn't that depend on the engine's ability to discern the name of a site. Sometimes, the url of the site and the name that's known to the site's common users aren't the same.

Are we basically saying that social media may be playing another role here: mentions of a site in online conversation, whether that be a review or a discussion in a forum, or a tweet, or a facebook post? Are we talking "branding". Didn't Schmidt talk about the importance of brands a while back?

[martinibuster]

But linkless citations...wouldn't that depend on the engine's ability to discern the name of a site. Sometimes, the url of the site and the name that's known to the site's common users aren't the same.

I agree that would be stretching Google's ability to correlate the name of a site with the URL. However it's not a stretch to correlate an unlinked URL with the actual URL. Google is so eager to discover citations that they even find and crawl URLs in JavaScript.

The confusion commonly sets in where a webmaster sees an effect but guesses at the wrong cause. After Matt confirms that the claimed cause is incorrect the community accepts that at face value and moves on, without considering that the statement does not rule out other effects. This happened with the so-called sandbox and with Update Florida. So when a webmaster says, NoFollow Links pass pagerank and Matt says no, this isn't necessarily the end of the story.

We know that Google is discovering unlinked URLs and URLs hidden in JavaScript. Here's an interesting quote by a webmasterworld member [webmasterworld.com] about his experience in how Google treated the anchor text of URLs as actual URLs.

My Webmaster Tools crawl Errors page always contains bad url references that google picked up by reading the abridged version of a url shown in the link text of postings by forum software like vBulletin.

So here is a question that I think is at the heart of this idea of unlinked citations: Does it make sense to count unlinked citations as citations?

Does a citation have to be a link?

[lfgoal]

So when a webmaster says, NoFollow Links pass pagerank and Matt says no, this isn't necessarily the end of the story

Unlinked citations as citations...

It used to be that a link was the highest form of referencing. I guess the idea was that if you actually went to the trouble of linking to a webpage, it was relevant, at least to the linkgiver. Better yet would be a link that appeared on a page that was intended to be somewhat permanent or attempted to be authoritative (such as a list of government resources).

However, the social thing seems to loom larger every month. At times, I've wondered how many new special interest websites and blogs are not being created because people are expressing their interests and targeting information through things like facebook, twitter, forums, yadda yaddy. And how many times do discussions take place that reference a particular website without a link reference being made. If a website is important enough to discuss and this action occurs repeatedly over time, you would think that the discussion itself lends relevance, with or without a link reference. The question would be, how important? It might be the difference between living together and getting a marriage document. Same circumstances, but that extra action of getting the paperwork should confer added significance. So, in that sense, an "honest" link should trump mere conversation.

Having said that, perhaps a LOT of conversation can provide an edge, or even trump a relatively flimsy backlink profile.

If so, I'm not sure I like where this is going. I think the social aspect is not how things should rank, even partly. However, the engines strive to give relevant results. Relevance may not always correlate with correct, or even slightly accurate information. And relevance can be gleaned from low brow jibber jabber.

[CainIV]

Unlinked citations are citations.

Along with the recent 'brand update' in Google, it wouldn't be a stretch to consider that Google has added in better ways to determine brand.

And ways to determine brand might include social mentions local and national link-less citations. One noticeable change during that update was the improvement for brands for their core offering keywords. I monitored at least 4 major brands and found movements in the SERPs that could not be correlated to longer term inbound link increases or other viral / burst links.

I believe this is already happening, and we see it in effect most at the local level.

Let's face it, the biggest nemesis for Google has been the issues caused by detection of, and ability to be manipulated by, paid links.

If I were Google, I would be working hard on developing a system of determining link-less citations so that I had to rely less on link text being such a large factor in rankings

-Reviews + Sentiment analysis
-Business listings, local and national
-FB / Twitter / Social Media
-Google Instant Previews
-Images (and the potential ability in the near future for Google to read images)

[aristotle]

What about when people try to link to your site, but mis-spell the file name or use .htm instead of .html.

For one of my sites, Google Webmaster Tools shows more than 20 of these non-working backlinks in the Crawl Errors section. Most of them are from forums where somebody tried to link to one of my pages in a forum post, but mis-spelled the file name or used the extension .htm instead of .html

I wonder if Google uses these non-working backlinks in some way as a kind of citation.

[buckworks]

Suggestion: when you spot broken links like that, create redirects so that they start working. Use 301's preferably, that send both the user and the link juice to the "real" page.

[aristotle]

Suggestion: when you spot broken links like that, create redirects so that they start working. Use 301's preferably, that send both the user and the link juice to the "real" page.

Thanks for your suggestion. Actually, I've considered doing re-directs for these already. But most of them are low-value backlinks from forums, usually with nofollow tags, and I want to keep all my sites "clean" by using as few re-directs as possible. If I ever see any high-value bad links, then I probably will redirect them.

[g1smd]

There's many types of citations that do not involve links. Mentions of company name, names of staff, business address, phone numbers, product names and codes, and so on, all likely count too.

[OMZen]

"Does a citation have to be a link? "

I do NOT think citation without a link adds any reputation score to the destination; at least not at the moment.

It may add relevance score or involve sentiment analysis as Google claims they do for local reviews, but certainly no reputation equity in the main index IMHO.