How easy is it to determine a good or a bad link?

Is that a Good Link or a Bad Link?

I played with a new tool this morning. It was some kind of link evaluation tool.

It purported to tell you whether a link from a URL was good or bad or somewhere in the middle.

Cool, I thought.

So I gave it a go and popped in 6 URL’s. All came up with wildly wacky results, all were deemed to be spam, all suggested I should do something funny with them and run away screaming.

Haha.

I’m not going to link to this tool, but I applaud the effort and love the little bit of it that will generate lots of discussion both for what it’s trying to do and some of the more wackier outcomes.

It might be fun to actually think some of it through and explore the whole notion of good and bad links and what might be a good signal and what might not. The bottom line is of course, that none of it matters really as ultimately it’s what Google or (if you like the traffic) Bing think of it.

It’s in that vein that I write this.

Link Tactics, History and Interpretation

If we track back over the years, we might find all manner of references to how Google determines what is a good link and what is a bad. We’ll find that their position has shifted over the years as they’ve reclassified their determination of what the web should be and how useful or useless a resource might be.

There’s been reams of discussion around quality rater documents and how they’ve classified URL’s as offensive, offensive being not becoming to the standards that Google wants to index or rank very highly.

Pagerank, specifically; linking structures that been determined as schemes designed to have manipulated Pagerank or link juice, have been called out as ‘dangerous’ and against the Google guidelines. Pagerank sculpting as it became to be known was thrown in under the bus as a tactic that might get you in to hot poodo.

There’s also been the general ongoing war on paid links and the various high profile dings given to companies who’ve been caught out.

Usually, these have been fairly high profile brands. At other times, they’ve been targeted at bloggers who’ve been perceived as having a fairly loud voice and audience. The tactics identified as bad, seem to change yearly. What was ok in 2006, isn’t necessarily ok in 2013.

The general rule seems to be that the moment an effective tactic is discussed within the SEO community to the point of ridicule, then Google comes out and makes a pronouncement, usually by video or if especially egregious, at some high profile marketing or search type conference.

Here’s a list of a few tactics that spring to mind that have drawn commentary at one point or other.

Directory Submissions – Generally accepted as spammy and a low quality signal.

This discussion from some time back gives a little insight in to how the ground had shifted from a position where a directory link from a place that had editorially evaluated a listing and wasn’t a free for all was a good thing (Yahoo, ODP) to a position where due to everyone and their cat building a directory to cash in and effectively sell links for page rank and anchor text purposes, wasn’t.

Debra: In the past Google/you have stated directories with strong editorial policies were OK to submit even if they required a submission fee to be reviewed. Is this still the case?

Matt: That’s still the case, but bear in mind that Google will ultimately decide which domains or directories to trust. Just because a directory claims to have strong editorial oversight doesn’t mean that it will meet Google’s criteria or that Google will trust the domain.

Debra: If an editorially run directory offers a sponsored listing option, do you consider them (the sponsored links) paid links and against your TOS?

Matt: Adding a nofollow attribute to sponsored links remains the best practice for any website.

In other words, Google decides, and if you ask folks to pay to play then it might be a good idea to wrap the link in a condom, else risk the wrath of a Googler in a bad mood some time down the line. 

Sponsored Blog Posts – Not very bright if done wearing size 15 shoes.

This seems cut and dry right? If you go to one of these websites that pays mummy bloggers to write about stuff on those blog networks, then you’re asking for trouble. Easy to identify, be it from a manual or algorithmic perspective.

if(($post_count_with_money_kwanchors > $defined_percentage) or ($links_per_post_to_singular_domain > $defined_number OR ( $links_to_money_kws > 1 AND $web_graph_shows_more_of_the_same))  might be one easy way of folding in an algorithmic stance which could be overlain  against whatever other set of metrics they might like to check $spends_money_on_adwords $is_a_brand etc etc.

Press releases – A rapidly diminishing tactic of efficacy, abused by users and platform holders

For a time, and even today these can be and are a useful tactic to employ both for general PR and SEO link benefits.The idea is that you get your great news item out there and all the cool news folks interested in your niche will see your stuff and write about it too. Hey, even Google will use your stuff and output it in their news results so you can even get on the first page for car insurance.

You can still get on the first page for car insurance, but in the UK at least you’ll find yourself at the bottom of the page and using places like PR Web for example just won’t cut it anymore.It doesn’t take a rocket scientist to work out that someone in the Googleplex might just have read about things like this http://service.prweb.com/pricing/package/advanced/ and decided that generally, weightings from such things should be adjusted in some way.

Barry posted a link and a quote from a Googler that said:

I wouldn’t expect links from press release web sites to benefit your rankings, however.

Quite unambiguous, you might think and hey, they can’t catch them all but its pretty clear in its intent which is “hey you guys, press release tactics are on our radar too, so be careful if you use these too as we might just ding yo ass if we catch you”.

Advertorials – A recent Google post by Darth Cutts set the record straight on these

Please be wary if someone approaches you and wants to pay you for links or “advertorial” pages on your site that pass PageRank. Selling links (or entire advertorial pages with embedded links) that pass PageRank violates our quality guidelines, and Google does take action on such violations.

The message is simple – do this and you’re dinged. In this particular example they dinged Interflora and stripped out some of the Pagerank from newspapers who were selling these too.

Widgets, Themes etc – A great idea, oft over done for search ranky benefits

Here’s the thing. You create a cool theme or widget for bloggers or site owners which is cool as it does stuff and adds value. However, from a Google perspective it’s not cool at all, if you also get a little link love from your actions as it’s not really earned in the way that they’d like it to be.

As far as widgets go, the code it uses might be a little javascript or an iframe that folks can just pop on in.

Of course, once people like us get a hold of them we tend to get link erections and offer up funky ideas to boost your link pop too. We might have said stuff like “Hey mr browser, just in case you are really old and can’t parse this html stuff, we’ve include a special bit of code called noscript which will show people what to do. Yeah, it has a link to our clients too but don’t worry about that, it’ll help your rankings…”

I am being funny of course, not everyone did that kind of stuff but some did. Some didn’t even bother hiding it and some just left a little brand link to the originator.

For themes it’s not half as egregious, but it can be over cooked too and generally, 1000’s of links from a footer in 2013 just isn’t the best thing to have in a link profile, especially if you come under scrutiny either manually or through Penguin.

Here’s Google’s head of webshhpam Matt Cutts again, talking about the value of links from widgets and themes

… links from article marketing, widgets and other pre-curated content types are unlikely to drive search rankings or visibility

Competitions – Link to me using this link text and win a prize/cash payment etc

Some folks took a view that one of the other possible reasons behind the Interflora penalty that lasted mere days because they are a brand and spend lots with Google  was that they were getting bloggers to write about them via a mechanism called ‘Competition’ the rules set out that all you had to do was do something simple like write about flowers and you’d win some roses or some other silly gift that wasn’t worth very much. However, the link value to Interflora was another story altogether which might well have helped them to rank for cool search terms like flowers or roses or valentines day.

Back in 2007 a very nice chap named David got dinged for something similar. He ran an awesome competition with some fabulous prizes. Someone named Jenni tried to warn him at the time, but he decided that it was good to go. Word spread and everyone was buzzing and bloggers talked about him and entered and linked and…you can read about how he recovered here.

So these tactics aren’t new, but they require a little thought and consideration around how they might be interpreted first.

Affiliate Schemes – Pay people referal fees and get them to boost your link profile whilst at it

There’s 1000’s of affiliate schemes out there. Outside of tradedoubler or CJ for instance, some folks roll their own. For a time these aff links were great for site owners and all very innocent. Some people realised that with a little server side jiggery pokery folks could have their cake and eat it. People using your affiliate add code would often use on topic anchor text as they wanted people to click through and hopefully make a purchase. this way, they’d get paid and the merchant would be a happy bunny.

Some of us realised that there was a bit of a missed link opportunity here and that with a little thought, such schemes could be tweaked to add a secondary benefit. Users could be redirected with session Id’s or Cookies and link love could flow through to a decent organic landing page too. header status 301 mr Google and boom, watch those rankings rise.

Of course, like anything a little bit crafty that’s done to the extreme and bragged about, Google eventually gets to hear and makes a pronouncement on it, usually expressing the view that you use nofollow. Subtext of which is that if you don’t, then it might be interpreted as an overt manipulation of your backlink profile.

 

Guest Blogging/ Hosted Content –  Let us provide you lovely content for your readers, all we ask is that you link back to our client with a few silly words in the anchor text

Dear publisher, we love your site so much and was wondering whether you’d care to run a cool piece on your website. We’ll even pay towards your hosting costs as we think that content like yours is just the best.

I’ve had emails like this. I’m sure you have too. On the face of it, there’s nothing wrong with this at all. But in Google’s world there’s everything wrong with it. You’re using money to get links to manipulate your link profile and if they just let things like this slide without nofollow attributes being applied then, they’d never make any money cos people might say, “Sod adwords, this is more profitable for me”

The tricky thing is that it is a good tactic for boosting your rankings, provided that you do it in a semi intelligent way. If that footprint is all too present and in the opinion of Google your content is of low quality or otherwise spammy, then you’re really asking for either a demotion of the ability for your blog to rank or pass link juice. If you are the guest blogger and you’re not thinking your tactic through, then you risk poisoning your link profile and seeing your site tank.

You might want to ask yourself whether it is really natural to get #n new links per week from blogs for money keyword that are the same.

Here’s that Matt chap again, talking about guest blogging

 

Forum Sig Links

Most half decent forum software these days has cottoned on to the ways of the Google rumpers of this world and have given site admins the ability to either nofollow signature links or hide them to the likes of bots and what not. That said, there’s a ton of unpatched or old software out there that gives people an easy link and an opportunity to inject some juicy link text.

These can be pretty powerful (amazed that they still work) and some firms actually hire people to go out and acquire. Some clever bastards write software too that enables you to do a mass blast, or stepped attack and build up x number of links in y days. xrumer being but one of many.

Comments on Blogs

Hey parse_poster_name, I loved your post on string from title tag  before reading this I knew very little, I’m now a little wiser. Thanks Money keyword in name field

Similar to the above, folks are still running around and doing this type of stuff manually and programmatically. If you’re a blogger then you know all about these wankers and sigh if they make it past your comment filters.

Fact is, they still work and people are still massively ignorant about the reasons why. Innocent/not very bright bloggers often think that these commenters are just being nice so happily leave their comments in place, pleased that someone has taken the time to comment on their stuff.

Google’s position is clear and defaults to the standard nofollow, link to them at your own risk perspective.

So you get the general picture. People do a lot of things to get links and get people talking about them on the web, that’s a given. Yet what isn’t so clear or easy to divine is how these tactics are interpreted. Do we really have to nofollow EVERYTHING to be safe? What about people who link to us naturally? Don’t they sometimes use methods that might be considered dodgy? Do we really have to run around webmaster tools or majestic seo constantly evaluating our link profiles, contacting webmasters, forum owners, blog commenters etc?

If every link on the web was nofollowed then it would be a level playing field. Nothing would count and we could all get on with working out what it was that did 😉 But of course that isn’t going to happen so we are left in a bit of a pickle.

We can hope that we never get misinterpreted by a human or an algo and just carry on building, writing and serving our clients, readers etc and hope that the various signals we’re creating everywhere are enough to keep us safe from the vagueries of the black box that is Google

Or

We can be proactive and fuss over every little link in our link profile and chase webmasters and site owners to alter a link to comply with Google.

Or

We can hope that someone builds a very cool tool that takes away and a lot of the legwork and alerts us when things are going tits up and helps us to evaluate them before we go all disavow crazy.

But building that tool isn’t easy either. Sure, we have all manner of data dumps we can inject in to such a tool and draw all sorts of conclusions around link placement, link duration, singular, sitewide, link makeup, link numbers, site authority, social mentions, links in, link ratios, pages indexed, pagerank, markup to name but a few and even with these, it’s not exactly easy.

So lets look at what might be good and what might be a bad link (I’m getting there)

How do we determine a bad link?

So, if a page has been around for say, 2 years and it had no Pagerank and that site had no pages indexed in Google then we might determine that from a quality perspective, it would score low. We might use one of those metrics from SEOmoz or A hrefs or Majestic to bear this out.

How about if that same site was just born? No pagerank, do follow links, poor quality scores from the link scrapers cited above? Would that make it a bad link? Should we chase site owner?

What if there were 100 of these? And if there are, are they on the same IP address? Are they using the same code base? Do we really know when they were created? How do we know that the site owners hadn’t blocked the link scrapers?

We might look at the number of links on the page. Are there 100’s of links to other domains?

We might look at who they are linking to. Are the sites they link to reputable for instance? Are they porn or pills or poker type domains that are totally outside of out niche?

Does our tool even allow us to stipulate a niche? Is it niche specific? If not then why not.

Are the links on the page hidden either through css or some other silly method like noscript or noembed or insert what other method you can think of?

If the linking site is quality and is ranking in Google, then does the link stand the test of scrutiny? Is it a natural earned link? Or has someone paid them to put it there.

If it’s a dofollow comment or forum link then is it a genuine one?

Is it a sponsored blog post? Would it be apparent to Google?

I’m not even going to ask about what is a good link as I think we can all agree that it’s all pretty messy and ultimately, Google decides.

If you care about ranking in Google then you need to keep a regular check on your link profile and manually review those that look suspect and even then you could be going OTT and actually diminishing your ability to rank.

A decent well thought out tool that asked lots of questions, that dug deep, that iterated out a few times, that queried multiple sources, that cross correlated with the Google Api might be a good way of taking the donkey work out of it, but I don’t think it exists today.

 

Rob Watts
Kickstart your business today - Get an SEO Consultation or just talk to Rob about your online aspirations. With over 20 years experience in building traffic he's pretty much encountered most markets and scenarios
Posted in SEO and tagged .

2 Comments

  1. So when are you building this tool then Rob? 😉

    The problem is most of us are ultimately trying to game the system through desire to succeed, client pressure, a feeling that ‘everyone else is doing it’ or for purely financial reasons.

    Rather than over scrutinise and worry about what Google says or does time would be better spent worrying about your content strategy and how you generate links from a multitude of sources and don’t become over reliant on common practises.

    In most cases it over reliance on single sources that causes issues the like of Interflora experienced. Whether you put it down to advertorials or anchor text is irrelevant, it’s the extent to which these were done which was the trigger and the signal. No one will ever get penalised for ‘some’ of anything, you get penalised for lots of it.

    With a content strategy incorporating as many sources of links as possible you will spend less time looking over your shoulder at the competition and the big G and more adding value to your audience.

    • Haha – Rob, if only I had the time or desire. There are some good tools out there, they certainly help, but are no substitute for the old grey matter!

      I think clearly, it’s about gaming the system in an intelligent way that stands any test of scrutiny. Competitive needs, KPI’s contracts etc are all well and good, but if they go too far or are packaged in unimaginative ways then as you point out, shit begins to hit fans and people get torched (albeit momentarily for Interflora )

      The takeaway for most should be that there IS a tactic, there IS a way of winning and these don’t include boiler plate unimaginative methods that can be knocked up at the press of a button.

      I don’t have a problem with people doing spammy things, provided they apprise folks of the relative risk and reward. Lots of people have churn and burn strategies that are great for rev gen which they’re happy to rinse and repeat.

      It’s great for Google too of course as it makes the game that little bit more harder and increases the costs of doing organic promotion. If it comes with a degree of risk, then bigger brands (and aspirant ones too) might think twice before committing budget to SEO and a little more for PPC.

Leave a Reply

Your email address will not be published. Required fields are marked *