An SEO's Advice: the importance of fixing outbound links

This is a guest post by Michael Martinez.

As Bing and Google rule out more beloved link building strategies, marketers increasingly turn to supposedly “safe” strategies like broken link replacement (a form of “link reclamation”). I’m not convinced this is as safe a link building strategy as its proponents want to believe, but so far the search engines are not hinting at future changes in their guidelines.

You will always have the right to ask for a link. No search engine can take that away from you. But when you do ask for a link because you believe it will help you build up your search referral traffic then you should assume there is some potential risk involved with that request. The fully realized potential risk is that you will be penalized (delisted) by a search engine for acquiring the link. But you should think of potential risk as a partially-filled balloon that may or may not inflate until it explodes.

Risk potential changes over time, but not all the risks you face concern search engine guidelines, penalties, and algorithms. Let’s just talk about the simple act of placing a link in an article that you publish today.

WIKI LINKS: One of my favorite examples of a high-risk outbound link is a link to any Wiki site that can be changed by its visitors or an active user community. Wiki articles may seem very good to you today but in 2-3 years (or 10 years) they will be very, very different from the content you linked to.

I am a long-time critic of Wikipedia because of the amateurish revert wars that experienced Wiki editors start in order to pervert the content. The way Wikipedia handles these disagreements is to penalize the 2nd person (the one who responds to the reversion) instead of the trouble-maker. Many tens of thousands of people have gone into Wikipedia, made good changes, and then watched in horror as some more experienced user comes along, changes everything back, and watches the article to ensure that the original contributor is blocked by Wikipedia’s reversion rule from keeping the good changes in the article.

If you want to link to a Wiki site that is your choice but you are linking to every idiot, troll, and well-meaning but clueless admin who uses the rules to make good content look bad. There is a lot of risk entailed in linking to any Wiki site, especially if you are expressing an opinion and you feel you are linking to an article that supports your opinion. Someone who disagrees with you can change the Wiki article to contradict what you are saying. Good luck fixing that.

LINKS TO BLOGS: As bloggers we should be linking to other people’s blogs. After all, supporting the community that supports you keeps the community strong. But most bloggers don’t stay with their blogs. If you just link to the home page of the blog in 3 years you may be linking to a dead blog that hasn’t been updated in 2 years.

If you deep-link to an article on a blog your link may survive for a few years but eventually something will change. Blogs are often deleted. They are often moved. The URL structures are changed. And the worst part of this is that you may be the worst offender in your rogues gallery of bloggers who have changed things without notifying you.

I started the SEO Theory blog as a subdomain on Blogspot in December 2006. In early 2007 we moved it to the SEO Theory domain everyone knows today. So that was a double-whammy on changes in URL structures: we went from subdomain.domain.tld to domain.tld.

The article URLs were converted to use the correct root, but at the time we decided to go with just seo-theory.com instead of www.seo-theory.com because we thought the shorter domain URL would be the visitors’ preferred choice (that turned out not to be the case).

When we finally added the www-prefix to the domain and redirected the non-www version I decided that would be good enough. But another decision I made at the time was to host the blog in a subdirectory. I did that because I thought that my employer (who at the time owned all legal rights to the blog) might want to develop some marketing content on the root page. But they already had an “official” Website and, frankly, their offline sales channel was bringing in enough business that they didn’t feel like marketing directly to the Web.

Eventually we dropped the “/wordpress/” folder from all the URLs and moved the content up to the root folder. But I never went back and changed all the links (it would have required far too much time for review because I was writing 5 posts a week at the time AND doing my day job).

And yet as the years rolled by I often found myself linking back to older articles, and the more of those links I generated with the domain.tld/wordpress/ format in the early days the more I unintentionally set up TWO automated redirects. This is one reason why pages on the site sometimes flash when you load them (another being the speed optimizations we have implemented).

Search engines can now handle up to 5 hops in a redirect chain. That’s great for SEO but frankly it creates a bad user experience for me. As I reshare old articles that I feel are relevant I occasionally find to my amused horror that the self-referential links do not reflect the correct structure. I have learned that leaving too many legacy structures in self-referential links eventually leads to trouble so now I review old articles on a random basis to improve the quality of self-referential linking.

REBRANDING KILLS LINKS: I don’t have an estimate of how many sites I have linked to through the years that moved to new domains, but there are a LOT of them. Given the number of Websites for which I write content it is humanly impossible to monitor all the outbound links and keep them updated. Even my close personal friends, who have listened to me rant on and on about how Websites break with rebranded moves, occasionally break links by rebranding their sites.

“Oh, but we always advise people to set up 301 Redirects,” you say. Yes, I tell people to do that, too. In my daydreams people listen to me. In real life they “just don’t have time” or “forgot to do that” or “asked IT to take care of it” and have a thousand other explanations for why it never happened. And there are many of YOU digital marketers whose content I have linked to who have broken my outbound links. Even the most experienced marketers don’t always fix their problems.

Old content may be taken offline simply because it’s “old, outdated, and irrelevant”. And for fear of incurring some sort of imaginary search engine penalty people won’t even redirect the dead URLs to a “that content is gone” page. So there I am, left with dead outbound links on my page and my visitors have no clue as to what I was linking to or why.

Whenever possible I replace rebranded links either with the appropriate URLs or, if the content has changed (or if the page now loads 20 advertisements) I just link to the oldest legible copy I can find on Archive.Org.

But even Archive.Org can fail me because if you set up a “robots.txt” file that disallows ia_archiver it won’t show people the page. I have done this myself simply to fight Website scraping (which, thankfully, is not nearly as bad as it used to be).

My final choice for fixing a rebranded link is to convert the anchor text to an italicized expression, to indicate to me (not so much to you) that there was once a link there to something I felt was useful and the other guy killed it.

IDIOCY KILLS LINKS: Sometimes I will link to an article written by someone I don’t know. They may be saying something I agree with today but eventually it becomes apparent to me that they got lucky with that first article. It’s a bit like being a Skeptic who links to an article about the silliness of Paranormal Research, only to find a year later that the writer is someone who advocates an alternative form of paranormal research (for the record, I try to stay out of Skeptics-vs-Paranormal debates as much as possible).

So there you are, linking to a Website that you now believe is full of nonsense. What should you do? Keep sending your visitors to a lunatic asylum and they will eventually assume you must belong there, too.

Maybe you feel I’m using too strong language here: “idiocy”, “lunatic asylum” are insulting, after all. But think about the way a site you linked to in the past now makes you feel. Would you link to it today? If not, why not? And if you did link to it in the past then you need to realize that you ARE linking to it today as long as your old link is still published and indexable.

Your feelings should play a huge role in how you decide where to direct your links. Trust your feelings, Luke, the Force of your emotions will guide you.

When I see that I once linked to a site that I now feel is substandard I kill the links. If possible I’ll find something else to link to but about half the time I just throw the carcass out into the cold and don’t even italicize the old anchor text. I want to forget that I ever linked to such a site. I want the search engines to stop passing credit, too.

OPTIMIZATION KILLS LINKS: If you have written 10-15 articles on the same topic over the past 3-5 years you’ll eventually come to the realization that you need to clean up that mess. It doesn’t always turn out to be a mess. News sites, for example, need to keep their content differentiated chronologically (and shame on the sites that continually add updates to old content).

But we as digital marketers realize that eventually we start repeating ourselves, and so we either reduce the amount of content we publish on a site or we start consolidating content. I recently did that on SEO Theory and I have done it for other sites. Content consolidation is a great way to reset the clock and give you some breathing space so that you can write about the topic again.

But every now and then when I am reviewing old links I find they now lead to redirected destinations which are terrible attempts to consolidate old content. For example, just before I decided to write this article I reviewed some outbound links on an old SEO Theory article. One of them led to a specific article that has been included in some sort of a category page. I could not find any trace of the article itself on the first page of results in the category listings, so I replaced the link with a link on Archive.Org.

When you redirect your old URLs to a consolidation page you need to show visitors who follow old links that the content they want is still there, easily reached, and important to you. Just following my (and may other SEO bloggers’) advice to implement redirects when you consolidate old content is not good enough (at least not for me).

I want to know what happened to the old content. I want my visitors to know that I am still providing a meaningful linking experience.

I rarely receive any requests from marketers for link reclamation. I would almost never agree to such a request anyway unless I knew the person and thought they were legitimately making a good recommendation for my site. Sorry, digital marketing world, but most of you appear to be hawking really bad content with your guest posting and link reclamation strategies. I have probably agreed to two link reclamation requests in the last five years.

Optimization outreach may lead me to replace old links, but the new links may not be as good as the old links were. At best I am improving a degraded user experience; at worst I am compromising with reality and killing bad links. What I would prefer is for the old article publishers to be consistent in supporting the sites that linked to them in the past.

Sure, it may be hard to show that those links still exist (or still help in any way), but if people are visiting your site through old links you owe it to them and yourself to give them the most relevant experience possible.

WHY IT’S IMPORTANT TO ME

As an advocate of writing timeless content (and I concede that not all my content is timeless), I feel that the links are just as important as the words and images on the page. I want people to know that when they land on an old article (and those old articles get a LOT of traffic) that they can trust what I am telling them.

Sometimes I do update the old articles. It’s necessary to provide some context (such as “this article refers to a service that went offline in 2012”).

Sometimes I take the old articles offline. When I do so I have to decide if I want to redirect the URLs to some other content or leave them “dead”. Yes, I do occasionally orphan inbound links that other people gave me in the past (or that I gave myself).

I know I am creating a bad user experience, but if you have done this then you’ll probably agree that you are compromising with reality and substituting a less bad user experience for a worse one. We may be right or wrong in our judgements.

Eventually I’ll figure out what to do about the content I have taken offline. I don’t want to leave a bad experience in place. But at least now that I can mark posts a PRIVATE on WordPress installations I can quickly see which articles are no longer useful and I’ll be able to think of ways to manage that user traffic.

To me, it says a lot about a marketer’s dedication to the consumer experience when I see them make an effort to resolve dead link problems in a meaningful, user-friendly way. When you just do it for search engines you really imply that you don’t think much about what kind of impression your site makes on visitors. I feel YOUR pain when I take content offline. I want you to feel MY pain when you take content offline.

About Michael Martinez

Michael Martinez has been developing and promoting Websites since 1996 and began practicing search engine optimization in 1998. He is the principal author of the SEO Theory blog.