Google engineer Matt Cutts uploaded a couple new videos recently focused on specific link building tactics and what Google thinks of them. I transcribed a couple of the vids and wanted to share one here along with a handful of thoughts and observations. I numbered (in red) behind the sentences I wanted to comment on.
Video title: What is Google’s thinking about links from article marketing, widgets, etc?
Transcription: Video length 2:48
Matt from Mountain View asks, ‘What is Google’s current thinking about getting links from article marketing, widgets, footers, themes, etc?’ (1)
Yeah, I’m really glad to talk about our current thinking which is about the same as our past thinking. But let’s be a little bit more explicit about it. Whenever you get a link from just a WordPress footer or just a random footer or you know when someone installs a widget or they install some theme on their content management system, it’s often the case that they’re not editorially choosing to link with that anchor text. And so you sometimes see a lot of links all with the exact same anchor text because, you know, that’s what the widget happened to have embedded in it or something like that. And even if it’s not the exact same anchor text, it’s like a, you know, it’s relatively inorganic in the sense that the person who made the widget is deciding what the anchor text should be rather than the person who’s actually doing the link by including the widget. (2)
And it’s the same sort of thing with article marketing. If you write a relatively low quality article, just a few hundred words and then at the bottom is two or three links of you know, specifically high density anchor text, (3) then the sort of guy who just wants some content and doesn’t really care about the quality might grab that article from an article bank or something and he’s not really editorially choosing to give that anchor text. So as opposed to something that’s really compelling, when you really like something, when you are linking to it organically, you know, that’s the sort of links that we really want to count more.
So it’s always been the case that these sorts of links that are boiler plate, it’s not really a person’s real choice to really endorse with that particular link or that particular anchor text, those are links that typically we would not want to count as much and so we either use algorithms or we have other ways of saying, okay at a granular level this is the sort of link that we don’t want to trust. There are some people who say I just want links, I’m just gonna take shortcuts, I’m lazy, I’m going to do the bare minimum that I can to get the links that I want to get, and they’ll fall back on some ways where, you know, I’ve mentioned this example before, but if you’re a K-12 school teacher in Pennsylvania and you are trying to get a web counter, you don’t realize that it has mesothelioma hidden in the text of the web counter, that’s not a link that we necessarily want to trust at all let alone give that much weight to.
So the sorts of links where people are really choosing to editorially link to you and not just because they’re paid, but because they think your site is good, those are the links that we are more likely to count more. (4)
(1) I raised an eyebrow (and laughed) when I watched Matt ask and answer this question, he must really want to make a point here.
(2) Matt said: …” And even if it’s not the exact same anchor text, it’s like a, you know, it’s relatively inorganic in the sense that the person who made the widget is deciding what the anchor text should be rather than the person who’s actually doing the link by including the widget.”
My bold to focus on two points.
One draw back to using widgets (and badges) as a way to build links is the issue of footprints. When people want to put a widget on their site, they copy and paste a block of code from the host site to theirs. The code they copy is what makes the widget “go”, it also has a link back to the host site for attribution, navigation and in some cases, link building.
Given the cost and time involved in badge/widget creation, most webmasters make one or two versions which means the code and URL behind the widget are repeated over and over as people upload them. If the widget is popular, this can become a problem since all roads linking back to the host site say the exact same thing or point to the same page. A lot of the same anchor text phrases linking back to a page (especially an internal page with little content) over and over is probably not a natural occurrence so they are ignored or work against you. You’ve spent a lot of time and money creating a tangible asset that is now a liability or at very least, a gilded snitch.
The most interesting part of the video (for me) was this sentence:
even if it’s not the exact same anchor
Now I get why he asked the question and made the video.
First, I think he’s trying to tell us they don’t only look at/find anchors as a way to determine manipulation, they look at the content around it as well. What is around the link in a widget that would be super easy for an engine to detect? The code. Code trails for a code driven algorithm is like Diet Coke to this addict, if it’s out there, I will find it.
Second, he’s also trying to reinforce the point offering different versions of the same anchor won’t escape their notice. You can develop multiple anchor text versions of a badge or widget but chances are the coding will be the same or similar which again, makes it uber easy to detect.
I have heard various Google reps stress the importance of editorially awarded links in the past, since widgets and badges using clean links back to host sites are not considered “editorially awarded” you may want to think hard before investing a lot of time or money into their creation. (Clean links = those not using the nofollow attribute or tracking software)
(3) Matt said: …”at the bottom is two or three links of you know, specifically high density anchor text …”
I can speak to this issue first hand, Google will ding you with a love note if they find your bio has all keyword phrase (high density) anchor text links in it instead of links to your website.
I got a love note recently for my Alliance-Link site, seems Google thought my hyper linking the term “link building training” repeatedly was a problem. I guess it would be if I was actually dropping that anchor but I wasn’t, I have never, ever, EVER (channeling Taylor Swift)…. built links for my Alliance-Link site in the 12 years I’ve had it.
Ever. First off I’m lazy and second, I’ve never had to so why would I, after all these years, suddenly start dropping high density anchor text phrases on a bunch of crap sites?
Two minutes (seriously – 2 minutes) worth of investigating and I found my Search Engine Land (SEL) column was the culprit. Seems a number of my articles had been scraped, including the bio which is featured, using the term “link building training” at the bottom of the page.
The fact my keyword rich bio sits on SEL is no problem, SEL is an authority site, I am known in link building circles, I’m writing on the topic of link building and I’m posting in the link building section of SEL. If my bio stays on SEL it’s fine, but if it gets picked up and circulated on crappy blogs? Not so much. Never mind the fact someone else was stealing my content, I got the love note, the anxiety that comes with it and had to do the work to make things right. Sucketh to be me in this case.
Two take-away’s here:
- Don’t exclusively use keyword rich (high density) anchor text phrases in your author bio and
- Don’t stick all your bio’s at the end (bottom) of blog posts.
(4) Matt said: …”links where people are really choosing to editorially link to you and not just because they’re paid, but because they think your site is good, those are the links that we are more likely to count more.”
If you read the comments under the video you’ll see a lot of people saying things like
you need to look natural in your linking
…be natural; a constant state of SEO leading from well written useful content.
I don’t disagree with either statement but even if you write the most wonderful, thought-provoking, life-changing-Pulitzer-prize-winning content ever, it has to be promoted to see any benefits. This isn’t about being natural, it’s about where you promote. It’s always been about the “where”, not the how.
In the future, links embedded in things like widgets, badges and WP templates won’t count for much since they come with the territory and are not editorially awarded. I also think things like sponsored links (as in buying a sponsorship or donating to charity and getting a link for it) paying for a directory submission where the directory doesn’t have strict editorial guidelines for regular and sponsored links, and even (gasp!!) guest blogging links (see my #3), might not count for much either. I think going as far as saying the link “won’t count” is wrong, just like it’s wrong to say links using the nofollow attribute are not counted. Both, in some respects are counted, as a number if nothing else.
Now more than ever it’s important to incorporate offline tactics, promotional partnerships and use traditional media as a way to build links, drive traffic and promote your brand. If Google doesn’t think the link pointing to your site has been given on merit, there is a strong chance the link won’t carry much weight. Old methods like article marketing don’t work as well now because many of the article directories lack editorial process. If you find a good one, great! Just don’t use the same keyword rich, “high density” anchor text over and over in your bio