Posted by Jane Copland
Name the number one criticism of linkbait: from a technical point of view, the top problem with popular content is that it often doesn’t attract many links. "Linkbait" has become a word we use to mean pretty much anything that gets media attention, no matter whether it is heavily linked to or not. In fact, one could argue that social media sites harm linkbait as much as they help it: if someone can say their piece about a subject within Reddit’s comments or StumbleUpon’s reviews, they may have less need to write about it on an external blog and link to the original source. Some social news sites nofollow their outbound links (StumbleUpon does this 100% of the time and Reddit appears to have a popularity algorithm that dictates when a nofollow is removed). Thus, forty reviews on StumbleUpon and many times more thumbs up mean nothing in terms of traditional, link-based SEO.
If blogging is a site’s primary means of acquiring links, virtually anything it produces could be linkbait of some sort. However, the things you called linkbait when you set out to create them aren’t necessarily what brought a site the most links. This is similar to the difference between viral marketing and content-based link building and I’ve been as guilty as anyone of calling something linkbait when it was actually publicity generation and eyeball-attraction. Ain’t nothing wrong with content like that, either. In addition, if I worked for a search engine (err.. a search engine that indexed content as well as links), I’d be very interested in the net’s collection of viral, linkless information.
StumbleUpon is a good place to go to find content that has passed before a high number of eyeballs but which is often lacking inlinks. Neither individual reviews pages nor popular items lists give away followed links, which I find pretty unacceptable: items receiving over a certain threshold of attention should at least be granted a followed link, especially as that content is what makes StumbleUpon worth using.
After a certain number of up-votes and reviews, the nofollow should be removed.
With the array of metrics available to calculate pages’ strength, trust and success, it’s becoming easier to figure out what sort of content creates visibility and which attracts links and whether it’s worth keeping or scaling back the sort of Digg-topping efforts that result in few inlinks.
So far, Linkscape is confirming a lot of what we all suspected about gimmicky content as opposed to useful posts and projects. We already have very good ways to measure the visibility success of content: 2,500 Diggs, hundreds of Reddit up-mods and many thousands of StumbleUpon views confirm that a lot of people saw something. Social media isn’t SEO, but it’s the fish to SEO’s chips and the value of exposure shouldn’t be underestimated. However, what is the true SEO benefit of different forms of popular content?
Take a comparison of five pieces of SEOmoz content. The five pieces cover the popular-content spectrum rather nicely:
Gimmicky: Alternative Google OneBox Suggestions. This popular post received 2548 diggs and adds virtually nothing to the collective knowledge of the Internet. Not much we do when we’ve been drinking adds much value to anything, after all.
On-topic; Light on content: Gorgeous Website Footers. Matt Inman’s list of beautiful website footers is far more useful than the previous post. It received almost exactly half the number of diggs as the OneBox post.
The "OMG" Ticket: Google Automatically Removes URLs Ending in .0. This post rode the "what the hell" wave into popularity by uncovering a formerly-unknown (and now abanonded) Google practice of throwing URLs ending in .0 out of its index. Webmasters came away from this post with one piece of knowledge, but that one thing was interesting enough to bring in social media traffic and garner inlinks.
Actionable Advice: 17 New Rules for Succcessful E-Commerce Websites. Rand analysed high-performance e-commerce sites and provided easily-implemented, useful advice regarding their effective strategies.
Large-Scale Actionable Research: Google’s Search Ranking Factors. I hardly need to explain why this is the most useful and least noisy type of content. Virtually countless pieces of advice and knowledge, great presentation and 100% on-topic information.
So how do these five different types of content perform when pitted against each other in an epic Linkscape battle?
What is the premier type of linkbait you see on the web? Of course, it’s the gimmicky stuff. It often performs the best in social media, and thus we’re fooled into thinking that it’s the best form of link building. It isn’t even true that this sort of content is the easiest to come up with! It usually takes true inspiration, a bit of wit and often a few martinis to be sufficiently funny to appeal to social media crowds en masse, but you’re often providing them with little more than a cheap thrill. Done often and well enough (Hi, Cracked! RIP Drivl), this tactic develops an enviable following and link profile, but it rarely turns into massive numbers of links if done on a normal scale.
Yes, there are exceptions and we’ve seen them, but the daily offering atop StumbleUpon’s most-popular list won’t necessarily turn into the sort of link love that less exotic pages will eventually pull. Thought you knew this already? Most of us always assumed that this was true, but it’s nice to watch the evidence pile up. Put your site’s or your clients’ linkbait through Linkscape (the above reports are free to run) and see which ones resulted in the highest degree of trust, inlinks and mozRank. A combination of high-quality, high-link items and high-attention pieces achieves both the social media and SEO benefits of link-worthy / viral content.
It probably means less time creating "fun" items, but the link benefit of understanding what produces the best linking results are surely worth it. Although I can’t promise that we won’t drink too much at a conference in the future and write something else that makes a lot of noise on Digg and garners us no links whatsoever 😉
More: continued here