The root of this discussion stems from this comment from Matt Cutts in an interview with Eric Enge back in July. Straight away there were people saying the sky was falling on infographics and SEOs should stop doing them. Even I found this an over-reaction and I’m not the world’s biggest fan of infographics. I’ve seen some truly awful ones that are not only badly designed, but are just plain wrong in their fact checking and presentation of data.
There is an argument here though – if these “bad” infographics still get links, which was the goal of it, it doesn’t matter. If the goal of an infographic is to get links and it does, one could argue that the design, data and presentation doesn’t matter. But that is me talking as a pure link builder. I feel this is the type of thing that Matt is getting at with his comments.
But let’s take a closer look at his comments and try to not just interpret them, but say what this could mean in reality. I’d like to open a discussion about what Matt actually said and if they really are looking to discount these links, how could they do it?
“In principle, there’s nothing wrong with the concept of an infographic. What concerns me is the types of things that people are doing with them. They get far off topic, or the fact checking is really poor. The infographic may be neat, but if the information it’s based on is simply wrong, then it’s misleading people.”
Imagine for a moment, Matt was making these comments in relation to written content. The same principles would apply right? Google do not want to reward an author who publishes content that is plain wrong and misleading. This is nothing new from Google, they do not want to reward bad quality – we all know this. It should not come as a surprise to anyone that Matt is saying this.
The key here is how Google can algorithmically determine the accuracy of an infographic? How can they look at a graph inside an image, extract the data points and see if the story being told is true? I can see how this would be possible with written content, but doing this for infographics on scale feels like a step too far for Google right now.
The conclusion here for me is that Google can’t verify the facts or data contained within an infographic (for now anyway) but this clearly doesn’t mean you shouldn’t care about them. If you churn out any kind of content for a client that is inaccurate, you are not making your client look good and whilst you may still get links, it will catchup with you eventually and the client will not be happy.
Can Google devalue infographic links based on this? Algorithmically, very doubtful.
“The other thing that happens is that people don’t always realize what they are linking to when they reprint these infographics. Often the link goes to a completely unrelated site, and one that they don’t mean to endorse. Conceptually, what happens is they really buy into publishing the infographic, and agree to include the link, but they don’t actually care about what it links to. From our perspective this is not what a link is meant to be.”
How can people not realise what they are linking to? By copying and pasting your embed code that contains a link at the bottom – when they publish it on their website, the link is often at the bottom and kind of hidden away. To me, this is what Matt is getting at here. I’m not quite sure how else someone could not realise what they are linking to – unless the link is embedded within the image itself which is of course perfectly possible.
In terms of Google detecting this, it is pretty easy for them to determine the position of a link on a page. They can already do this, so tweaking this (if they need to) to find the types of links Matt is referring to will not be that hard. In fact, if you imagine that most infographics are pretty long, have a few lines above it as an intro, you can imagine how easy it is for Google to determine the distance from the top of the page to the bottom and change the value of the link accordingly.
Can Google devalue infographic links based on this? Almost certainly yes.
The link going to a totally unrelated site is very common and I can see what Matt is getting at here. This is often born out of necessity from the SEO and can happen when they are stuck for ideas that are closely related to their niche. Commonly SEOs will complain that they work in boring niches where the scope for creating relevant content is limited, so they will create something unrelated but still link it back to their client’s commercial website.
I don’t feel like Google would need to run analysis on the infographic itself in order to work out of the link is relevant to the target website or not. I’d imagine that the title of the infographic (usually a text heading) will describe the content of the infographic and the website itself will be very easy for Google to look at.
However this question has been asked for years – does it matter if you get links from websites that are not relevant to yours? What if it’s a very high authority website but not relevant at all? What if it is a certain category of the website that is relevant?
Exactly the same questions apply here.
It doesn’t feel like this is what Matt is getting at though, he knows that this is the nature of the web. People will link to apparent irrelevant websites sometimes but to me, what really matters is the overall diversity of a website’s link profile. If the majority of it is irrelevant, then that is more likely to cause a problem.
Can Google devalue infographic links based on this? Yes, but no more than they already do for “irrelevant link”, if they do at all.
“The big key is that the person publishing the infographic has to know, and agree with, including an endorsement to your site as attribution with the infographic.”
I know what many of you may be thinking, how can Google possibly “know, and agree with, including an endorsement to your site”. They can’t get inside the publishers mind can they? Well, not yet at least, but soon maybe. 🙂
But it is possible for them to figure this out in my opinion. To work it out, let’s think about how someone may not know and agree with an endorsement to your site. There are a few ways this can happen:
- They just copy and paste the standard embed code which everyone else uses and includes a link
- They receive an outreach email with a description of the infographic and just copy and paste it – including a link. This is the same description everyone else who got that email uses
- The link is “hidden” inside the infographic itself when the embed code is pasted
Let’s be clear, these are not editorially given links. In fact, if a blogger is simply copying and pasting a description from an email, how much do they care about the quality of their blog? If a blogger finds a piece of content genuinely interesting, they will take the time to comment personally with a unique piece of text. They will also link to the source in the way they see fit – which probably won’t be identical to how everyone else has linked to the content.
Now, let’s look at the question again. Can Google detect that a blogger can “know, and agree with, including an endorsement to your site”. Yes they can if we imagine they use signals such as the bullet points above. What they look for is an active and clear endorsement, if they can’t find it, it is pretty easy for them to devalue that link.
Can Google devalue infographic links based on this? In my opinion, yes they can.
“This is similar to what people do with widgets as you and I have talked about in the past. I would not be surprised if at some point in the future we did not start to discount these infographic-type links to a degree. The link is often embedded in the infographic in a way that people don’t realize, vs. a true endorsement of your site.”
To me, this is the clearest indicator of how Google may choose to discount infographic links. It also backs up many of my points above that relate to the use of embed code. I feel like it is this embed code that not only would be easiest for Google to detect, but also what Matt Cutts may be most concerned with.
In the same way that link building with widgets can get out of control, the same principles of detection and devaluing can be applied.
Can Google devalue infographic links based on this? Yes, they have done it before with widgets, they can do it again.
What else could Google look at?
Let’s go outside the scope of the interview and look at what else Google may use to detect good quality infographics.
The types of links they generate
There is one thing that Google can very easily discount right now, if they haven’t done so already – websites that look like they are made just for linking to infographics. Adam over at SEER did a great write up on a list of sites which I maintain that accept infographic submissions. There are some websites in this list that do look like they have been made just to publish infographics. This is fine to an extent but if I were Google, I’d be looking at other signals around that domain to see if it is a genuine resource or if it accepts anything that people send it.
What Google could do: look at the ratio of editorially given links vs. links generated by sites that are clearly just made for infographics. If you have more of the former, they may trust that content and the associated signals a bit more.
Adam covered this pretty well in the post above but it is worth mentioning again. Google can very easily look at the social signals surrounding the links you get to an infographic, as well as the social signals around the infographic itself. This is super easy and they do this already when it comes to other forms of content.
Unique content surrounding the infographic
This is key for me. Imagine that an infographic gets embedded on twenty websites, all of which link back to a single source. So it is pretty easy for Google to detect who created the content and who should get credit for it the most. Now imagine they gathered all of the written content around this infographic on each of the twenty websites that published it.
If they find twenty copies of the same description, with the same link structure, would they trust that these were editorially given? No.
If they find twenty unique descriptions, each semantically relevant to each other and with their own ways of linking, would they trust that these were editorially given? Yes, I believe they would.
The second scenario shows that the infographic has reached an audience who have felt compelled enough to engage with it and have taken the time to comment on it. This shows that the infographic is probably accurate, is well designed and on-topic – exactly the types of things that Google want to reward.
There is the possibility that these twenty people wrote about the graphic because it is terrible – but some semantic analysis would probably be able to detect this.
Right, that’s me done for this post. Please let me know your thoughts in the comments and if you agree or disagree!