28+ Stories of Failed Link Building Experiments
At the heart of link building is attention to detail. If one item is out of place, then it could spell disaster for a campaign.
Therefore, it can be extremely helpful to learn what HASN’T worked. There are so many things that could go wrong, so it’s important to know what those things are, which is why I enlisted the help of some of the most brightest thinkers in the industry.
The tactical takeaways from this post will be what things were overlooked, but ultimately, you’re getting a first-hand look at how great link builders think, which is invaluable in & of itself.
Here was the question:
“What link building experiment did you try that failed miserably? Why do you think it failed?”
And here were their answers.
1. I harvested Google for Australian bank accounts and then scheduled to pay
1 cent into the first hundred accounts with “SEO DEJAN” in the transaction description.
For one dollar I could buy targeted traffic comprised of one hundred bookkeepers, accountants, business owners and managers located in Australia. Basically anyone who has access to either the printed bank statement or online banking. Bonus effect is that I attempted to influence Google suggest so that when people start typing “SEO…” the suggest would come up with “SEO DEJAN”.
I thought my little guerrilla tactic would get some traction and publicity, scoring me links too, but the opposite happened. The whole thing backfired when I got a call from the manager at Commonwealth Bank of Australia investigating a complaint report from one of the recipients of my one cent thinking I was probing bank accounts to see which one was active in order to attempt extracting money from it. Apparently you are not supposed to make unaccountable deposits into trust accounts, which was an account type of the complaining recipient. Apparently they had spent days investigating and consolidating the transaction and after that sent me a pretty nasty letter threatening action if I should do another stunt like that in the future.
I’m so glad we haven’t done this on a large scale. It would have been a total Charlie Foxtrot.
2. When Geocities shut down we developed a crawler to examine all backlinks of now dead pages and have orchestrated a campaign of link repositioning to our own domains containing the content of the disappeared pages retrieved from the Wayback Machine. This got us a massive amounts of link juice, except that we never realised that a lot of webmasters who used to host their sites on Geocities moved to paid hosting with their own domain. We ended up getting tons of complaints and copyright infringement notices The nightmare lasted for over a month, after which we shut down the whole operation. I cringe when I think about it. What a stupid idea.
I had a client in the flooring industry and I asked one of the top linkbaiters of all
time if we could work together on one to see what we could make happen for this
guy. I can’t recall the exact idea but it was something like “Top 10 Best Floors in the Movies”. It did about as well as you would expect it to have done. I can’t really blame the link baiter. He was probably overloaded with offers to do stuff and this was a fairly low-budget affair. But I guess the lesson is don’t take every job that comes along…or something like that.
Wow. In the 15 years I have been doing link building, I have had MANY failures.
But I think this one takes the cake because it got my website hacked – they were so
I’m really into photography. And there was a time when I wanted a photography blog that I had created to do really well. This was around 2005 or so.
At that time, you could do extremely well with “ego-bait.” So, I created a post on my photography blog that broke down the “top 10 photography blogs on the planet.”
Now, I wasn’t actually looking for the best of the best… I only actually cared about who had the strongest link profiles and ranked for the most keyword phrases. Because, of course, I wanted my site to rank better… so I wanted higher authority sites… not necessarily the best ones.
So, I emailed the people that I included in my post and they were all excited and honored to be there. About 3/4 of them linked to the post and a month later, my site skyrocketed with rankings. I wish I still had the analytics charts… because they were awesome.
Here’s where I screwed up. I told the very story that I’m telling you right now on my personal blog…. and… yup… the photographers saw it and were PISSED.
One of them was rather wealthy and had a large following of folks that he shared the story with. One of his followers was a black hat hacker… and he hacked my photography blog.
I logged into my site one day and every single photo on my photography blog was changed to some weird Japanese anime image.
He also told all of the other photographers that were mentioned. They promptly removed the links from their sites… and within a few months, the traffic had gone back to normal.
I lost some serious credibility in the photography community after that… and learned a valuable lesson. Don’t talk about what you do if it works well… which makes me the “Anti Jon Cooper.”
So it’s been some time since I’ve personally done link building work, so here’s a
story from a veteran link builder on our team:
“I once did a massive cold call outreach which was unsuccessful.
I came up with around 30 search strings to find blogs to pitch guest posts for. I input them one by one into SERP Scraper, copied the first 100 results for each string into an excel sheet (so I had around 3000 sites in total). I mass input it to URL Profiler and got the names and contact info. URL Profiler gives you contact info back for around 40% of the sites you input, which means it gave me back around 3000 x 40% = 1200 emails.
I mass pitched them all through a single mail merge. URL Profiler does not give you names for all of the emails it finds, so many of the 1200 approx emails it did find for me were without names. So approx 1200 emails had been sent without checking whether those sites were relevant to the post I was pitching, without checking if the emails were correct and without pulling names for all.
I initially thought atleast half of them would be relevant as they were appearing for very specific search strings I had made and so worst case scenario, I’d get at least a 5% response rate which would lead me to 1200 x 5% = 60 links. For minimum effort that spanned maybe 30 minutes max, 60 links would be a huge win. But in the end, only 8 sites replied back.
What I learned from that was that tools are great but can only do so much. Manual sorting is very necessary and mass cold pitches that are not customized don’t get you far.”
This was maybe 2009, and I was trying to put together an offline event for a
group of influencers in my client’s niche. I’m pretty sure it was the first event I had
ever put on so I was completely going into this thing blind. Plus, I was running this remotely from my client’s location. (In hindsight, I have no idea why I thought this was going to work.)
The whole premise was to target a subset of bloggers in my client’s location, get them to come out and experience this event, and then ultimately blog about it with links back to the client for hosting/setting it up. Outreach was great; I had about 20 influencers committed, and they were super excited to be included. So I got all the logistics set up, and I was feeling pretty good.
Except the day of the event, it rained. Only 6 of the 20 bloggers showed up. They didn’t know where to go. My client was 30 minutes late showing up. And of course, I didn’t even think to give anyone my number (or gasp: be there in person) to help coordinate everything. Total fail. I got maybe 1-2 content pieces published out of it and not exactly glowing reviews.
Last summer we reached out to 20,000 realtors with a pitch for a piece of high-
utility evergreen content.
We got 2 links.
We had earned links previously from 10-20 realtors by pitching local, high-utility content and we believed in them as a prospect group. We wanted to scale up our efforts so we created more high-utility (non-local) evergreen content that we hoped would be applicable.
Here’s why we think our rates were so low:
- As a group realtors are not website/publishing savvy
- We didn’t understand their website audience(s) sufficiently
- Realtors have low perceived value of publishing non-sales-driving information to their website
- Realtors are BUSY ($$$ or gtfo)
- We failed to recognize the importance of geography to realtors as linkers
- Test smaller list segments and see what happens!
- Build out geo-specific lists + content
- Remember that location specificity can increase content linkability
Featured Infographic (AKA: Guestographics)
A core strategy of ours used to be developing infographics on our websites (and clients) which was a gold mine for links from being embedded in blogs with custom descriptions / introductions, referenced as further insights on a topic / expert advice and one of them even being featured within a TV News Report. However, all of a sudden our response rates began to decline dramatically and interest in ‘single image / graphic content’ almost disappeared – this was around the middle of 2013.
Upon evaluation at the time, it seems as though there were a couple of elements that created the demise of ours (and others) strategy:
- Website traffic increasing on Mobile Devices, where infographics are not visually appealing (see Mashables’ 2013 mobile traffic post)
- Popularity of the same infographic strategies (from Brian Dean’s well constructed Guestographic post)
This is definitely an example of how the link building playing field constantly changes based on specific strategies becoming common knowledge (and spamable), technology advancements (eg, mobile usage) and just how your target audience can instantly put up their personal firewall on specific strategies due to others wasting their time.
Nowadays, for us, it is all about ensuring strategies are always unique and that they meet a lot of internal ‘future proofing’ criteria to ensure that we do not get smacked by similar issues again. For example, instead of completely scrapping Featured Infographics, we expanded the strategy to be focused on ‘Content Assets’, creating engaging content, compiling images that compliment the piece, ensuring a mobile friendly (responsive) layout and showing great value to any high quality site.
- Other SEO’s not providing infographics of a high quality, meaning that anything that mentioned ‘graphic’, ‘infographic’ or ‘super awesome information with high quality references – ALL IN JUST ONE IMAGE’ were instantly tagged as ‘spam’ to any website that has a strong audience or boasts quality content.
Buying Expired Domains and Ranking Them….Ok so let me start by saying I’ve
had quite a bit of success with this method, but my story is about my failures, some
really expensive ones.
I was first beginning my hardcore journey into domaining and aggressively buying expired domains to rank as opposed to setting up brand new sites. With a decent bank roll to work with (this was around penguin 1.0 early 2012) my plans was to use my skills to invest in domains and web properties instead of something like the stock market. With the knowledge in hand already, it seemed like the logical place for me to invest.
But here is the problem I ran into, I paid wayyyy toooo much for some of the domains. Two specific examples, one domain was $2600 and the other was $1500.
Prior to that I had swooped a domain for ~$500 and it was a huge win, just crushing Google traffic and the plan I rolled out for the site was working phenomenal. Then I had a couple other domains I spammed links at and they responded very well, at this point I thought I had discovered fire for the first time.
Truth is looking back, I think there was a good bit of luck involved.
So depending on how you look at it, maybe I did alright and bought *enough domains to hit a winner, the losses being part of the game, gambling of sorts.
But it wasn’t all perfect execution on my part, which leads me to the fail miserably part of this question. The $1500 domain had a friggin’ penalty, yup, even my “skilled” analysis missed some sketchy history. Learned my lesson there though, man you really have to go insane mode checking EVERYTHING when prospecting domains. Check every last snapshot in archive.org to make sure some spammer like myself hasn’t already gotten to the domain.
Boom, out $1500 right there. It doesn’t end there though, I spent easily another $1500 on that same project, ~20 hours developing, and ~$300 for a logo. Yeah you can see where this could get pretty bad. Then I sat there watching the rank tracker wondering why the hell every initial ranks were terrible and no movement came after that. Bad times, definitely bad times.
The other fail was a whopper at $2500, now this one didn’t have a penalty jut really didn’t rank that much better than a new domain. So yes it had a crap load of links and looked amazing on paper, but it didn’t do anything spectacular once I threw up some content and targeted some keywords. For that kind of cash you better be crushing the SERPs damn near out of the box.
I learned the hard way how crucial prospecting is if you’re considering investing in a domain that’s anywhere over a couple hundred bucks. Then again, I managed to swoop one really kick ass domain I’m running successfully today. This is still a really powerful tactic, just be patient and don’t make the same mistakes I did.
Mobile Phone Link building idea
We have had some creative ideas in the past which were not passed by legal at large companies I’ve worked in, tho these ideas might be ok for smaller companies tho.
Usually you might suggest a few link building ideas to a client and then the client will pass them off to legal for approvals, well this is how large clients do business.
This one idea we had when I was working at a telco was to make a “non-branded” micro site and basically call it the “Official Australian Mobile Phone Day” and then you use a 3rd party persona to reach out to all your competitors and get them to link back to the mobile phone day website and make some PR around it.
Now once the mobile phone day has “died out” and people have forgotten about the website you then 301 the site back to a money page on your site and instantly you have links from all your competitors. Cheeky idea tho you can adapt this to any business really.
This was one idea we had tho it got knocked back by legal because we would probably end up getting sued.
One experiment I did for a past client that I was really optimistic about was
running a Holiday-based outreach campaign (I also wrote about this tactic last year).
The site I was working on back then was in the online gambling industry (online bingo) – wherein majority of their players are older people.
So we’ve listed all the holidays in the entire year that’s related to their target audience (you’ll actually be surprised to see that there’s one or two in every month of the year). We’ve scheduled and ran different promotions and specials for each holiday.
However, despite the thorough planning and execution, the process I’ve designed for this type of outreach campaign still failed (but I still believe that this is very doable). The problems we’ve unexpectedly encountered – but should really have:
- Time – it’s a very time-sensitive type of campaign, wherein you’d also have to deal with the publishing schedules of the blogs you’re looking to help you promote your site’s holiday-based events. You must do outreach months before the actual event to really ensure that you’ll get ton of exposure to it – and on time.
- Budget – we approached hundreds of mommy blogs (and other blogs we’ve found promoting/writing about the holidays we’ve chosen to build our events/promos around). Since my client was in the online gambling space, most of these blogs charged a bit more than their usual rates to get the promos featured. Weighing which prospects to pursue and invest our budget burned a lot of time.
If ever I get the chance to do this kind of campaign again, I’d definitely do it right next time.
I once used a clients airline miles to buy an all expense paid trip for a family to
go to Disney Land. We targeted mom bloggers and told them to write about the
contest on their site in order to be entered. We got somewhere around 40 LRDs… and nothing happened. Rankings didn’t move at all. My thinking is the lack of relevancy to the client’s vertical (health supplements) didn’t help us in the SERPs. It was a pretty gigantic waste of time.
It’s actually pretty tough to think of one failed attempt that stands out, but there
were certainly plenty of mistakes in the early days. Thinking back to my first ever
attempts at link building, I remember feeling pretty confused about what to do. There was so much information out there, and every single article seemed to contradict another. I’d guess my first mistake was simply not getting started on *something* soon enough.
When I ran my first ever site, I used to get amazing editorial links by buying cheap advertising as a ‘way in’ to the editorial teams. That worked a treat for a little bit, but clearly wasn’t sustainable as the ad guys just got pissed off with me buying £20 ads in the classifieds. There were plenty of other mistakes too – buying too many directory links in one go and getting my wife’s domain penalized probably wasn’t the smartest thing to do.
We have an internal bit of banter here. When someone’s about to do something big, it’s ‘Good luck Geoff – don’t fuck it up!’. Truth is – if you’re too worried about making mistakes, or trying something that doesn’t work out, then you’ll never get started on anything at all. So I’d say the biggest mistake anyone can make is just not cracking on and testing something out.
I was working on a client and decided to totally switch up the outreach email that
I usually used as we were focusing on a new product and I felt like I needed to do a
much more personalized email. It totally bombed. In fact, out of close to 30 emails I sent out, no one responded. I didn’t even get responses to say “heck no” or “take me off your list you cow!”
Here’s why I think it failed: it was too much. I had three paragraphs where I usually just had a few sentences. I wanted to try and make a connection between the product I was marketing and specific content on other websites but in trying to do that, I think I just turned everyone off. I’m wordy anyway so my attempt to get more personal became about as overly wordy as this answer is. Usually I at least get SOMETHING back and out of 30 emails I’d have expected 6 responses with maybe 3 turning into actual links, but nope, I blew it.
The lesson here is that you CAN get too personalized in your outreach. I gave it all away because I thought it might be easier than going back and forth for 5 emails but I think the email was a TLDR and just got deleted within seconds of reading it.
I pitched a discount program meant to target bar associations. The client’s
business was related to B2B financial services and it made a lot of sense for
attorneys and firms to use the client’s service. I had past experience with EDU and other discount programs that worked great. So, I did the outreach, and I soon learned something about professional organizations like bar associations – they only have a handful of revenue streams, and charging sponsors to offer a discount happens to be one of them. They’re called affinity programs, and they’re not cheap if you want to do one with every bar association in the country. We were able to cut the effort off fairly quickly after learning this, and did manage to get a couple links from associations which did not charge to offer a member benefit, but I had a major foot in my mouth after my confident assurance that this program would work wonders for the client.
The second lesson that came out of this is that even a new client will forgive a mistake like that in exchange for your honest assessment that you need to move on to the next opportunity. We’ve had that client over 3 years and have since run a number of successful content and link development efforts.
I’ve had my fair share of link building fails in the past.
All failures come with lessons and that’s incredibly relevant within SEO. It’s better
to test something to see if it works than to not try at all. It’s not always that easy when you’re working on client accounts, but if you let them know of the risks in advance then you’ll be fine. The alternative, and much safer approach, is to test everything on yourself first (I’d recommend this where possible).
One of the failures that comes to mind is from a number of years ago now for a client that I worked with in the UK. They were an AV distributor and we were looking to drive some links through to some of their key product pages.
As part of the strategy, I wanted to run a blogger day as part of a wider event. We would get 10-20 bloggers down to the evening event that we were running and supply them with blogger packs that included product samples, freebies and a USB with loads of imagery, etc. loaded up onto it.
As part of the agreement, we asked the bloggers to do write-ups on their respective tech blogs, including some video blogs at the event where they could get hands-on with the products. This would have been amazing, I’m sure and would have bagged us a ton of great links.
The only thing was that there was a slight admin error and the travel for the bloggers was booked on the day after the event, completely missing it and it was too late to sort it out… oh dear.
One technique I tried recently was finding and reporting Viagra spam pages on
high DA sites. I had a good list of .edu domains and I would Google “site url + viagra”
(or cialis etc). If anything was found it means that the site had been hacked and someone had added spam pages to it. Happens to a lot of .edu domains unfortunately. If I located a spam page I would contact the site and let them know.
More often than not, I received absolutely no reply. Its like they didn’t understand what my email was trying to tell them. Even if I explained it like “Did you know your site has been hacked and you are advertising Viagra?”, they still don’t treat it likes its critical problem.
Not sure if I’m allowed to give examples but here’s one site that was hacked pretty bad:
You would THINK that a site would be alarmed and want to fix the issue. This particular site’s response to me was “it’s not an immediate problem and all the students are in exam season, our Director of IT will be looking into/resolving the issue when he’s done his exams”.
Its hard to then ask for a link when they don’t even see the favor you just did for them.
I think this technique has a low success rate because a) the people who receive the email have no idea what you’re talking about and b) schools are busy places with a lot of red tape and departments.
If I was going to try this technique again, I might just start by emailing the Dean/Principle/President of the school directly as they are likely someone who would take immediate action.
How one word ruined a killer year long strategy
One of our biggest link building lessons was learned the hard way.
Our goal was to chase press mentions. So we implemented a local link building
strategy where we could go and chase press mentions in local news and articles that featured direct quotes and request backlinks to specific local pages from them.
We had long-term success with our client over the span of 14 months, with response rates ranging between 40-50% and between 20-30% conversion rates. Then one month, we saw a drop in mentions, and in an effort to maintain our KPIs, we replaced the word “link to us” with “attribution” in our outreach. It was an innocent misuse of a word that has unique contextual meaning in the PR world.
In short, a personal phone call was made by the journalist to the quoted party of the story, a PR executive for the company we worked with. It resulted in the PR exec shutting down our campaign (as she had no idea it existed). Organizations need to ensure that the PR and SEO teams work closely together and communicate all campaign initiatives. This PR exec didn’t know we had been given a “community” email address from their company, who we were, or why we were reaching out to these link opportunities.
We learned that just keeping the SEO team happy wasn’t enough in this case and should have ensured the PR team knew what we were doing. Now our most low hanging fruit opportunity is gone due to a can of worms that one word brought out.
As hard as it is to admit, I’ve had countless link building projects fall flat on
their face. Most of the time, I can tie it back to the content not being engaging or
interesting enough. For example, a lack luster infographic.
However, my worst experience was with a large scale outreach campaign. We were managing roughly a couple hundred bloggers in the same niche. They were all going to push their reviews of our product at the same time. The idea was that by all of them talking about our product at once we had the potential to get the brand heavily saturated in it’s target niche… That is not what happened at all.
Very few bloggers stuck to their timeline. They were posting their reviews at any moment they pleased despite agreeing to certain time frame; some posted their reviews months later. The bloggers also told several people in their networks that we were giving product away for a review, but we had already reached our cap. This offended several bloggers in the niche, who later negatively commented on posts. Additionally managing the process of getting the product to 200+ bloggers was an absolute nightmare.
We did still get several links, but we never got this sort of megaphone effect we were hoping for. However, the biggest loss, was time. All of us fell so behind on other projects, because this one campaign was consuming all our spare time.
That being said, I do think this strategy could work for the right niche and product. However, I’d definitely be hesitant to try anything like that again.
When I think back to this it stings more than the bucket of dirty dead river fish
water that got flung in my eyes whilst water fighting in Thailand last month.
My most traumatic link building experiment happened two years ago. I’d just published an expert roundup about local SEO and thought I’d try pitching link prospects with a 60 second Snapchat style video. I think the service I used was VSNAP. Can’t say for sure though because I’ve intentionally tried to eradicate every memory of this out of existence!
Before I tell you what happened, I’d just like to add that this went down at a time when my online endeavours weren’t making jack sh*t. I was struggling to make money online and losing hope in ever being able to move out of my parents house.
So as you can imagine…it wasn’t the best time to get a bucket of fish water chucked in the eyes!
Sure, the video actually got A LOT of views. The stats showed that 60 people or so opened the email and watched it. Thing is…only three people replied and their responses were soul destroying.
One simply said “WTF?!”. Another said “Dude…not cool”, and another said “Sorry but we don’t do business like this”.
Long story short, I didn’t upgrade to the VSNAP premium service 😉
That experience actually shook me so hard I almost gave up on blogging.
To sum up, I wouldn’t recommend using video emails to link beg. Instead only use video to connect with someone in a fun way and to simply strike up communication. It is good for that because it does make you stand out and is way WAY more personal than email.
But…never send a complete stranger a video of yourself in a dark room, pretending you know who they are, what they do and with a fake smile and a hint of desperation in your voice, ask them for a link. Just don’t!!!
My biggest failure came early on in my career as an SEO. You see, I failed to
understand that when you are aiming to build links, and there is more than one
person involved (i.e you!) then shit can go wrong. The idea itself was spectacularly simple. The brand I was working for at the time had a major TV commercial going live, which we KNEW would go nuts – it was a single spot buy and during a highly watched UK TV show. It was just that good – and considering the interest it would generate there was no way I was going to miss out an opportunity to garner a few links. So I built a fairly detailed plan.
We built a staging area that would go live the minute the advert aired and would be linked to from the home page.
The area would have a number of assets that would be worth linking to for example:
- We created a “Making Of” video that we would use to further interest in the campaign.
- A back story behind the campaign and comments from people involved.
- A code embed that would allow people to embed assets from the campaign such as images and stills
- A series of out take videos.
- Assets for affiliates to distribute on their own websites
The plan was to publish the campaign as soon as the advert started airing, supported by a press release which would link to the assets page so that media that would be reporting on the campaign would have a resource to link to. Sounds great right? It failed.
It failed because there were too many moving parts and I just didn’t take enough responsibility. Here are the errors that caused the failure:
- I was due to be on holiday 2 days before the advert went out, so I had to plan everything to run while I was away.
- Although we built a staging area, and tested it – the devs messed up and when the campaign went live – all the links on the page were broken because they were pointing to “staging.website.com”. Suffice to say that no one was around at 9pm on a Friday to fix the issue, as the agency was closed. Someone did manage to fix the area, but it was Sunday afternoon.
- The code embed seemed to work – unfortunately the links were to “CDN.staging.website.com” instead of the target URL. It was too late – see next point.
- Affiliate pre-populated email and assets went out within hours of the TV show. Guess what? because the affiliates had access to materials we wanted the public, media and bloggers to have access to on our site, they benefited – because our stuff wasn’t working.
- The press release went out with the HOMEpage URL instead of the landing area we so painfully built – no incentive for media to link to. In fact – a resourceful affiliate sent out his OWN press release with links to his assets page – and gained about 40 links in 24 hours.
In total we guessed we lost 3 weeks work, and about 200 odd links (judging from the links our affiliates got cause of their “coverage” of the campaign).
What I really learnt was – if you want to be an effective link builder, don’t leave ANYthing to chance – you have to almost be a control freak. Link building on auto pilot is only for spammers – if you want to run campaigns – you have to micro manage every part of them if there are too many moving parts. Most of all – don’t go on holiday just before a campaign launches.
The most common link building failure we run into is projecting subject matter
expertise for a client who doesn’t have it built in naturally to their branding,
even if the content is good.
For example we might pitch something on electricity for a non-electrical-focused client – if they sell maybe one or two green-friendly items they’re trying to push – but because the client isn’t a vertical expert, it tends to overall perform worse than subject-matter content, even if specific and really well done.
That’s not to say you shouldn’t ever push content slightly outside your perceived subject matter expertise, but if you do, you’re immediately starting from a position of distrust you have to root yourself out of, and that should be evaluated as a risk factor when entering the arena.
As a programmer, I am always looking to have machines do the work for me.
I’ve long been an advocate of broken link building so I decided to do a personal
experiment with a fully automated process, pushing the technique to its limits. The idea took this form…
- Crawl the web to find links to 404’d PDFs.
- Extract the latest copy if that PDF from archive.org
- Display that PDF in an HTML5 viewer (so the links point to a page not a pdf)
- Download all the backlinks to those 404’d PDFs from the APIs of Moz, Majestic and Ahrefs.
- Extract the webmaster contact info for all of those backlinks
- Email them automatically to fix the broken link with my new HTML version
So, I strung together a handful of APIs and hit go.
It worked. It worked well. It worked too well. Even in a single-threaded version, links were pouring in from everywhere, including EDUs and GOVs. The site was acquiring 200+ root linking domains daily on autodrive and I was receiving countless thank you emails from the webmasters for helping them out.
And then I tweeted about it – not the domain name, just the success. It was less than 24 hours before the site was deindexed and I had an email in my inbox from Matt Cutts asking what was going on. It turns out mass copyright infringement isn’t Google’s favorite thing, especially when used to power a link building scheme. I explained that it was an experiment and pulled the site down. It is still down to this day and will drop in the next couple of months.
In retrospect, I did some stupid stuff and the site got what it deserved. The biggest takeaway, though, was that you should block your experiments with robots.txt. If you are testing new link acquisition methods, you should do so in an environment that does not actually influence the search results. Until you have established the technique as white hat, your best bet is to play it safe behind the protection of robots.txt. Finally, the human element can’t be divorced from good link building. The more automatic, the more risk.
I’ve had tons of failures over the years, recent ones are based on creative ideas I
thought were awesome but unfortunately no one else did, others worked well until
Google completely destroyed them.
I think the biggest error I ever made was based around directories, we all know they worked well historically but I implemented this just before the full force of Panda/Penguin descended on the landscape. I had a client with tons of locations and equally as many individual small websites, so I decided the best idea to acquire links would be to find 4-5 locally relevant directories for each location, and manually submit to each one (yeah I know, manual). I didn’t use dodgy anchor texts, the directories were all real and had genuine quality guidelines so there was no reason why this wouldn’t work. Unfortunately 3 weeks later Google begin to deindex all these directories in mass and the links that were now in place were completely useless.
The project took around a month to complete and was a fair investment for the client, so this was a painful lesson to learn.
Another slightly different mistake was some link removal work I did. My blog had naturally collected links over time, but I had paid for 10 site wide links on some high PR websites (this was 2008). These links worked well, I was driving tons of traffic and even service enquiries, however I was worried that these links would get found and the site would eventually be penalised. I decided to remove the paid links and all my rankings dropped off a cliff, I didn’t expect such an impact so I quickly got the links live again 2 days afterwards but it was too late, the removal of the links obviously flagged something up at Google’s end and I was under penalty for 2 years. I would have been better off getting a manual penalty and using the disavow tool.
Nearly 5 years ago I was just beginning to build out our US based link building
team so there was plenty we’d yet to learn on how to finesse the prospecting and
outreach. We were just getting started with one of the oldest and biggest brands of hearing aids I’ll refer to here as BRAND X. As none of us had any experience or knowledge with hearing aids we did some simplistic brainstorming to develop the personas we would be targeting. Our lengthy process went something like this:
1. Who wears hearing aids? – People with a hearing disability.
Done. Find them all and email them.
Had we spent a little more time developing the personas we would have discovered that while senior citizens who are hard of hearing tended to have a strong affinity to BRAND X, every one else with a hearing disability despised BRAND X with a passion.
Example of an angry response from our link building CRM….
Subject: up yours
Other replies were more verbose and politely explained our faux pax.
That cartoon granny remained taped to the office wall for years as a hilarious reminder to spend a little more time on those personas.
It wasn’t really an experiment, it was more like a small part in a larger link
building campaign that was a complete disaster. We created a small site (around 25
pages) about studying abroad in one country in particular. We hoped to build quite some .edu and .gov links to the domain. This way, we could use the authority of the domain to build links to our clients website, but also to link out to pages on other sites that linked to our client. Getting (indirect) university and government links to both our client and their links sounded like a great plan, but it completely failed. Even though the website had a great, governmental-looking design and solid content, and we sent out quite a few properly written emails, it only attracted a handful of links.
I think it failed for a couple of reasons. First, most of the universities and governmental entities we contacted thought it was their job to inform their students about studying abroad, and not someone else’s. Instead of linking out to our site, some of them simply upped their own game and improved the content on their site. Also, our micro site lacked a proper ’About Us’-page., We couldn’t tell a fluffy story about some kind of non-profit rainbow pony organization that created the site. We didn’t want to lie (it was a commercial company behind the site), so we kept the About Us page fairly basic. We tracked the visitors that came to the site via our outreach emails and analysed their visits. Lots of them aborted their visit on our About Us-page, much more than we expected.
It still frustrates me that this approach didn’t work out (at all!), since I expected quite a lot from it.
My biggest failure is when I started an infographic site (can you tell this was a
few years ago?), and figured I would do broken link building with infographics that
had been moved or taken down.
I found an infographic a lead gen site had made that had been posted elsewhere, and linked to from a Google Sites site from a librarian that was about infographics. I picked up the infographic, put it on my site, and sent her an email, letting her know I’d fixed the link and she could link to my site.
I got in return an extended screed about copyright law and how I couldn’t just pick things up and put them on my site, and she said that she’d never link to my site or anything I do bc I didn’t adequately respect an infographic from a lead gen site’s copyright.
Sometimes you get what you want, and sometimes you get experience.
A few years ago, a colleague of mine was doing some link building outreach and
accidentally used the subject line “link request” after forgetting to change the subject
line from a template I’d reviewed. Weirdly, it actually worked because he followed up straight away and apologized for the subject line – this actually led to him getting a few links!
I tried to replicate this on a freelance client of mine where I was trying to get links to a piece of content they’d built. I emailed maybe 15-20 bloggers and didn’t get a single positive reply! There were only a couple of replies mixed between straight rejections and people asking for money. Suffice to say, I didn’t try this method again!
So I thought I had come up with the most clever play off of broken link building:
creating a fake “broken link” non-profit/startup to use as a single persona for all
broken link outreach.
The organization’s goal was “to try to cure the Web of broken links”, and cited evidence of the problems it’s solving (i.e. in the court of law, a lot of web links to .gov domains are breaking all the time, so past case information is becoming outdated when they cite certain documents).
Once outreach started, however, the idea fell apart. Turned out, these site owners didn’t want my follow up recommendations for link suggestions, as they didn’t perceive me as a subject matter expert. So they may fix the links, but not add any additional ones I suggested, one of which would be the client. Which basically ruined the point of doing the whole thing.
In hindsight – yes, it could work if we only did direct-replacement. But even then, response rates were far from great. Coming off as a tech-savvy persona is usually not ideal for this kind of outreach; instead, go for more of the “casual web surfer” persona to deliver the best results (from what I’ve found).
After a few decent successes with creating linkable content assets for one
of our e-Commerce clients and receiving some great coverage for them we were
quite excited with our next piece which was an interactive map showing where you can charge your mobile device.
When coming up with linkable content for clients we try to make sure that we can cover a number of different use cases especially if we want to create something evergreen. We thought that this map would do that after all how often do you see people complaining about their smartphone battery life. We also added filters to show places that also had free Wi-Fi. If you allow your browser to share your location with the map you can then receive directions to the nearest charger to you.
Why the content didn’t quite capture the attention we had hoped for:
- We had anticipated that businesses would add their own locations to the map and hopefully increase footfall to them
- Many people on social media did not see the map as something useful for them as they lived in these cities already & knew where these places were
- The map became quite cluttered as we added lots of large coffee chains
- The content was very US/UK centric as these were the main cities we covered first
Despite all this the map was quite an easy piece to create and the biggest cost was the time invested in adding the initial locations. Even though the map didn’t quite capture the mainstream attention we had hoped for we did manage to pick up a few dozen good links for the client and there are future opportunities to promote this piece to tourist boards, universities and hotels etc.
One of my most spectacular link-building learning opportunities had to be when
I formulated a plan based around sending outreach emails to about 500 university
admins in the .edu space. Prior to outreach, I had hired a crack writer and graphic design team to create a piece of killer content that was one of the only authoritative guides of its kind. I assumed that, since the content I was promoting was of such high quality and highly useful to university students, this campaign would be a slam dunk, but like most newer link-builders, I quickly learned that even the best campaigns can be quickly laid to waste by even the smallest of details.
The detail in this case happened to be school vacation. I sent out our targeted and personalized email outreach at the end of December and were immediately swamped with a deluge of “Out of the office” messages. It turns out that when the students leave the university, so do the professors and staff. The 8 responses I received out of our 500 hand-crafted emails rapidly taught me the value of background research, and the importance of timing. Now I always make sure to check school schedules before I send out an email blast.
My top experimental/blackhat link building fail was back in 2008. I had a non
paying client that owed me around US$4K and found out he had not paid other freelancers
and was getting a lot of link building for free. I decided it was time for revenge and bought a very similar domain name with better branding and pointed it at his domain so it resolved at his site. The next stage was to outreach to all his top linking sites and tell them ‘we are rebranding and please change our links to the new domain’. The plan was to then 301 all the stolen link juice to one of my affiliate sites.
Yes it worked very well, but one of the webmasters smelt something fishy and emailed him. I then quickly received a letter from a solicitor threatening to sue me and thought it was time to stop, even though they had no hard evidence it was me. Moral of story, don’t do any ‘experimental’ link building that can get you in trouble!
About four years ago, I helped launch a link building campaign that was targeted
at food bloggers. The campaign seemed pretty simple. We were going to host a recipe
contest, where a food blogger would post a recipe creation on their own site (like they normally do) and feature an ingredient that the client sold. The blogger would link one of their ingredients to a product the client carried as well as link to the official recipe contest page as a method to declare their entry in the contest.
We thought this would work well for a few reasons. The barrier to enter seemed very low. These bloggers are posting recipes on their site everyday, they would simply need to incorporate one of the many common ingredients the client sold.
We assumed the campaign would have enough built in virality to spread, reducing the amount of outreach we needed to get it to take off. One blogger sees another bloggers entry, and then decides to enter it themselves.
We launched the contest, with a $200 gift certificate for the winner. We sent outreach to our target list of 20 bloggers we had worked with before. Bloggers we thought would be likely to enter and get the contest rolling.
The end result was not pretty. We got about 8-12 total submissions and links. Almost entirely from websites that were previously linking to us. Our failure to build our own outreach list to target new sites was a major problem. This was before Facebook and Twitter Ads were an option and heavy outreach was needed for it to succeed. By the time we realized this wasn’t going to work out, the contest’s 4 week window was coming up quick.
The concept was good but our poor execution resulted in it being a link building failure.
I had a client that was in the personal gifting space, and this campaign was
around the holiday season. Trying to avoid exchanging product for a link, I thought
it would be creative to send a product to bloggers with no prior correspondence, sort of like a Secret Santa. I obtained addresses via a WhoIs lookup, and was able to find about 25 bloggers within the client’s niche that met our linking metrics. After sending the product out, I anxiously waited for the links to start showing up. Unfortunately, they never did. While I did receive some social media coverage for the client, I received a couple inquisitive emails along the lines of “Thanks for the gift, but ummm how did you get my address???” While the WhoIs lookup is technically public information, it’s not exactly the easiest thing to try and explain to a complete stranger, and definitely not a great look for the client.
Now that the days of exchanging product for a review + link are over, it’s a very thin line to walk when dealing with Google’s Webmaster Guidelines. While I think I may have been on to something with my idea, the approach and execution could certainly use some refinement. At least I didn’t get any responses like what Tony and Richard received!
Many many years ago, in an elaborate attempt to get .EDU links in rapid scale –
I built a slick technology website that claimed to build the best student blog hosting
solution available, at no cost. I then did outreach to IT Directors at universities, and essentially packaged WordPress MU (multi-user) install on their servers, to allow students to setup their own self-hosted blogs on the sites. In return, I would have admin access to create as many (or as little) blogs as I wanted. Note: The school would benefit from providing blog access to students, and I was providing my time to set it up at no cost. It was a win-win for everyone.
Did it work? It actually worked, and I got a university to give me a site! I was given a sub, sub, sub, sub-domain, and it was very difficult to get the content indexed and when I did get enough content and links to the site, I probably could have invested the same amount of time on just getting links to the property I wanted ranked in the first place.
The links never moved the needle, the site never ranked, and I eventually turned the keys back over to the school to run the blog themselves.
Do you have another story of a failed link building experiment? If so, ping me at jcooper at pointblankseo d-o-t com, and if it meets our editorial standards, I’ll add it to the post along with a company name / link!
Thanks for reading! For more link building help, checkout the course.
RQueRoyalDesign.com | Email Marketing | Web Design | Graphic Design | Digital Marketing | Content Marketing | Social Media Marketing | Search Engine Optimization | 571.482.9758
September 6, 2017
December 30, 2016
December 15, 2016