The Science Behind Link Building
While it’s possible to debate the importance of virtually any other ranking factor, what has been found time and again is that links are the most important of all – don’t leave your link profile to chance
Link building became a dirty word in the wake of Google’s Penguin update – not because the practice isn’t necessary, not because link building is in any way negative, but because many practitioners were using tactics that were dishonest and counterproductive. If you can remember as far back as pre-Reddit fora, you will also remember the links that appeared under any post; the same was true (and unbelievably still is) of the comment sections of blogs and articles.
Brands and agencies would either place, or outsource the placing of, enormous volumes of links anywhere where links could be placed – all in an effort to artificially inflate the link profiles of their, or their clients’, websites. The same link-builders also hosted ‘link-farms’ – sites that existed solely as a repository for outbound links.
Once these tactics were rendered obsolete (though not non-existent – it is still possible to earn short term results from such black hat techniques, and brands are still penalised for this kind of activity now, leading to far greater losses in the long term), the term ‘link building’ fell out of favour because of its association with tactics that had been firmly penalised by Google.
However, link building is precisely what outreach is and a host of other euphemisms, yet link building seems a far better description – it is a process of construction. More than this, however, it is a process which has been honed over two decades to become (in the best cases) a science of link building.
The scientific method
We’ve discussed the scientific method before, in the companion piece to this – The Science Behind Content Marketing – but the ideas behind the concept bear repeating.
For a hypothesis to become (in the scientific sense) a theory, it must be rigorously researched, tested, analysed and retested to remove as much doubt as possible about the conclusions drawn.
This is the scientific method. For any strategy to be truly successful, it must be reproducible (see the crisis in psychology to see how seriously the scientific community takes this – the field is in turmoil over reproducibility); and there are few things better equipped than the, itself tried and tested, scientific method to ensure that success can be repeated.
The following sections, as we have done previously, will detail which stages of the overall link building process (many of which you may already be aware of) fit within each level of the process chart and how they can fit together to produce a link building strategy on which you can rely.
Ask a question
There are numerous questions you will need to ask when developing your scientific approach to link building, but amongst the most important are the following:
What do I want the links to achieve?
This is going to heavily impact how you approach your strategy. While, ideally, all links will drive traffic to your site, it is not necessarily the primary motivation – some link building endeavours look to build a site’s authority (by being linked to as an authority, or a producer of a particular piece of research) which, though these links may not drive traffic, they build the target site’s authority through a kind of trust flow from the linking site to the target site. This is why, often, link building campaigns will focus on high Domain Authority (DA) sites. Then there is a third variety which is to build relevance – either by subject or by area.
Who, what or where can I target in an effort to place these links?
Once you have decided what your priorities are, the question you will then have to ask is: how can I do it? You’ll need to consider what sites, publishers or influencers would be the ideal home for links to your site. If you’re looking to build relevance by subject, you’ll need to ensure that the sites are focused on the same subject, look for country code top-level domains (ccTLDs) that relate to your target area, and for well regarded influencers relevant to your field of expertise.
What do I have to offer that is worthy of these links?
While it was not uncommon for agencies to pay for links (another black hat technique which is very much frowned upon), when I say ‘offer’, I’m not referring to payments but to content, research, offers and promotions and possible long term relationships with publishers – things of genuine, if not necessarily monetary, value to your target publishers and publications.
Do background research
Background research for link building should be comprehensive – link profile is still a major (if not the major) ranking factor for search engine results pages (SERPs) and something so vital should not be left to chance. Some research techniques Click Consult employs are as follows:
- Organic Benchmarking
One step can be found in a talk given at Click Consult’s Benchmark Search Conference by Director of Search Alan Reeves. Using tools such as Searchmetrics, it is possible to download all of your competitors referring domains. These can be sorted by DA and the results plotted on a line chart – this will give you an idea of the distribution of links within your industry – allowing you to ensure that your eventual strategy targets link acquisition in a manner that will maintain the organic distribution common to your niche.
- Bell curve targeting
As outlined in the Relationships chapter of Click’s SEO Uncovered series of eBooks, you can ascribe a point system to both your own site and those of potential targets to approach for links, by arranging these scores in a bell curve you can begin to plot key targets relative to your brand’s position on the curve. This can help prioritise targets and can also serve as a measure of progress.
- Content review
There’s no real quick way through this, though there are shortcuts. By content review, I mean that you will need to take stock of what your competition and potential targets are publishing and sharing. Researching the content that has already succeeded will give you additional data points when it comes to approaching target publishers and publications.
Construct a hypothesis
All hypotheses when it comes to a project of this nature will essentially fall into: “if I do x, I will earn a link from y which will have z impact on my site’ – but we have the opportunity here to be a little more bold and define what results we want from an overall campaign. By identifying an initially small group of target publishers and publications, with a common theme/subject and preparing content/proposals that can be reasonably easily adapted across the range, you can then look to decide what you think would represent success for that campaign.
To do this, you’ll have to factor in the cost of producing an asset, or running costs of a project versus the potential gains of improving your DA or increasing your traffic etcetera. Make sure however, that you are setting SMART goals (Specific, Measurable, Achievable, Realistic, Time-Based) – setting the wrong goals, too high a benchmark or too short a time span can lead brands to abandon projects that could be measured reasonably as a success or, at least, as a good springboard.
The chances are, you will be looking to improve your link profile for as long as you have a link profile and – as with any long-term project, even those, such as this, broken into smaller campaigns – you should expect incremental progress as a result. Your first approaches, your first content, may only achieve uptake by 1% of your targets – but as long as next time it is 2%, you are doing something right.
Test with an experiment
There is only one thing to do once the parameters are set for determining success – and that is to try to reach the targets you’ve set. This is a vital step in the progress – as experimentation is in the scientific community. Begin with small subsets of your targets, split into groups that will receive a few styles of communication (contact should be personalised, but there’s no harm in having a wireframe to work from) or content. Measure the results of the various types of contact, the subject lines etcetera. as you would with an email campaign.
In this regard, dependent on aims, it could be the case that a brand chooses to use two or more different varieties of media, content type or tone and trial variations for short term campaigns – so that it can be established which variety gains most traction. It is important at this stage that expectations are calibrated for each of the iterations (based on historical performance, or competitor research) against which success can be measured accurately.
Is the process working?
This step represents an argument for using an email program like Act-On, MailChimp, Hubspot or similar as they offer the ability to monitor open and click through rates, in addition to this, a mix of content offerings for different publishers, different methods of approach etcetera. will allow you to measure and assess various strategies.
In the unlikely event that all of your approaches are successful, the next task is to determine which of the approaches represent the best return on investment ROI – which pieces of content/approaches etcetera were the least expensive to create or took the least time per measure of success versus the set metrics. It may then be possible to mix the best performing aspects of a number of different approaches in order to enhance the next, slightly larger, campaign.
Success is never final and failures never fatal, as the saying goes – so even if few or, at worst, none of your approaches are successful, their failings can sometimes teach you more than any immediate success. Look to recalculate your rating scale for the bell curve, check your data for DA distributions and refine your process for content production and approach. Consider this a step in the process, rather than a hurdle.
Analyse data, draw conclusions
Whether or not your first efforts are rewarded with success, your next step remains the same – analyse data and draw conclusions. There are no guarantees and a style of approach, a variety of content that works one campaign may fall flat at the next attempt. There are few things that cannot be improved through thorough analysis.
As this is the case, it would be recommended that any strategy should loop from here back around to the hypothesis stage several times before progressing to the next step. Through what is commonly called trial and error or trial and improvement, it is possible to funnel the lessons learned in several smaller campaigns into slightly longer, then slightly longer campaigns, all the while tweaking and improving in order to ensure that each successive campaign is better than the last.
For this reason, you should separate each of the aspects of the campaign – the aforementioned open and click through rates, the composition and tone of contact, the styles and varieties of content, the link acquisition rates and analyse them in isolation of one another. While one form of contact may have the best open rate, the content offered to others might have a better rate of engagement and still another publisher may have driven more traffic or delivered more value from their link. Only by studying campaigns at a metric level rather than as an overall campaign, can you ensure that you are taking the best aspects forward.
Communication is no less important in link building than in any other scientific process. Once the process has been undertaken, and you can support your conclusions, it is beneficial to communicate the results to all departments involved in the planning, design and implementation of the strategy. This is not just to prove that you’re worth your money – there will be things that all of the departments can learn from the various experiments and results.
While you may be able to use the results of your investigations and examinations of the results, with the right method of communication between stakeholders and participants, you can actively encourage the positive aspects of the campaigns to become part of standard practice within the appropriate department – allowing all concerned to improve results across the board.
There are a lot of moving parts in a link building strategy, but by careful planning, monitoring and execution, combined with the process of incremental improvement afforded by the scientific method, you can ensure that your results are positive, replicable, and long lasting.
The following resources from can help you build and refine your strategy:
- SEO Uncovered: An Advanced Guide to Organic Search (SEO)
- The Science Behind Content Marketing
- Earned Media Marketing Cheat Sheet
- Organic Benchmarking – SEO Strategy Based on Competitor Analysis
- Content Marketing Boxset – Series 1
- Content Marketing Boxset – Series 2
For more information on our SEO services, use our contact form to get in touch and see how an award winning SEO agency can put you in front of your competition…