Since its beginnings Google has relied on links as being a key quality signal for grading websites. The basic premise being that the more influential the site, the more links and citations they would attract from 3rd party sites.
Over time, as a ranking factor, this has of course become more nuanced, with other off-page factors, such as the quality of the linking site, the keywords used and contained in the anchor text, and the terms and topic of the surrounding paragraph, all contributing both to the overall impact and topical impact of the link.
Of course, as digital marketers and brands became more aware of the impact of links this quickly turned into a digital arms race, with sites swapping, buying and spamming links with little or no thought to their quality, relevance or marketing impact.
While marketers were acutely aware of the potential up-sides to links, they were also aware that there was no detrimental impact of a poor link, so quality control was non-existent.
This status-quo was seriously harming Google’s core ability to rank sites based on organically gained links as they became buried under a tsunami of gamed and paid links. These links were being created at enormous scale, both manually by offshore teams, or by bots and software.
In early 2012 Google rolled out the Penguin update, originally intended to introduce a negative impact from the kind of lower-quality gamed links that they wanted to discourage.
Almost overnight, adding the risk of penalty killed the demand for paid and spam links.
Estimates are that this affected around 3% of search results, with some high-profile casualties is was as much a PR move as it was an effort to rebalance their index.
By 2016 Penguin had gone through several rollouts and Penguin 4.0 became part of the core algorithm.
Penguin is now more granular. Penguin now devalues spam by adjusting ranking based on spam signals, rather than affecting ranking of the whole site.
This change now meant that instead of a sitewide penalty being applied, which could take months for sites to resolve and recover from, spammy links were now being devalued in real-time with either a limited detrimental impact on a site or page, or simply being devalued completely with no net impact whatsoever.
Google seems to have followed a logical tactical progression over time: no action > penalisation > neutralisation.
Do I think we’re past the days of link penalties?
No, they still exist, but you would have had to have been up to some impressively negligent stuff for you to discover you had been hit by one these days.
While we’re at it, have Google solved link spam?
No, it still works, but to a much lesser extent than is used to, but it still works. Google are only ‘pretty good’ at spotting gamed links “regardless of the site” even when referring to contributor spam on sites like Forbes and The Huffington Post.
But in 2018, with the history of Penguin, manual penalty rollouts and Google becoming increasingly adept as spotting and devaluing gamed links, who is going to take the risk of following a similar strategy with their brand, where real people’s incomes and livelihoods depend on the commercial success of the site?
Who wants to be waking up every day checking to see if that was the day that Google caught up with you?
The discounted link graph & link hubs
I’m sure that following the rollout of real-time Penguin in 2016 that we’re now looking at a situation where an enormous amount of the link graph is now simply being ignored by Google altogether.
Over time I’m sure we’ll see this become even more pronounced.
An interesting patent development was referenced by Bill Slawski last month.
Entitled ‘Pagerank Updated’ this describes how Google may soon be looking to reduce the amount of rankings manipulation by building trust into the algorithm by using ‘seed sites and pages’ to more accurately derive a ‘trust’ score for certain sites.
Assuming this one day becomes part of their core-algorithm, then an absence of citations from seed sites would mean that no amount of linkspam or low-quality links would contribute a positive effect on search rankings.
This would mean that having an effective media relations strategy which incorporates targeting coverage in top-tier media, as well as key industry publications is likely to become an integral part of any successful digital campaign.
So there has to be a mindset change within digital marketing, away from looking at links as a volume-based commodity.
Rethinking campaigns around forming relationships with influencers and developing content that will give brand owners access to the relatively smaller subset of media sites which can have a measurable impact on search visibility.
Why Google can’t get away from links?
In the post-Penguin era marketeers are fully aware that not all links are equal.
Sure, there may no longer be the immediate looming threat of a manual or algorithmic penalty, but when marketing budgets are finite, it is counter-productive to be spending time and money where it isn’t going to have the most effective impact.
The truth of the matter is that Google can’t easily get away from using link metrics at the core of their algorithm.
Content and links are intrinsically interlinked. Great content promoted effectively naturally attracts links. It is this organic process that search engines aim to quantify and measure that is at the heart of their algorithms.
It is also the very foundation of the web – crucial to navigation, user experience, connecting pages, as well as understanding quality and user intent from a search perspective.
There is a logical progression that as more off-page signals are being introduced into Google’s algorithm, so therefore links will start to matter less.
The reality however is that Google isn’t going to move away from having links at the core of their algorithm. Other quality signals can be manipulated, often to an easier and greater extent than links. It’s just doesn’t happen as much, or garner as much attention as they’re not as effective.
Matt Cutts has raised the issues with using social signals and user engagement signals in the past. None are as stable, as premiant, as widespread, give access to a wide dataset of user intention and relevance – none are as good as links.
That brings us back to the 2012 Penguin update – if links are going to fade in importance, then why would Google spend such time and effort protecting the quality of their data?
The bigger picture – behave like a brand
It would be just as negligent to focus solely on link development as your digital marketing strategy just as it would be to remain focused on spam links when considering site level metrics.
Quality factors such as dwell time, social signals, brand search, usability factors, site speed are either proven already, or are logically likely to be being used by Google as ranking factors.
SEO has gone far beyond simply focusing on small elements in isolation and is now about fitting into the marketing mix and bringing together a multitude of skillsets such as design, CRO, UX, digital PR, and copywriting.
Running a successful site is now more than ever about making something that the user wants, that they can access and use easily and provides an excellent user experience.
Neglecting any of these elements is simply leaving money on the table and risking a competitor taking advantage of the opportunity.
Sentiment – The future?
When trying to predict where Google will go next, I ask myself,
- Is it possible?
- Is it logical?
- Does it improve their data, and would it improve their results?
Something else that I am convinced will start to happen soon is that Google will start to bring user sentiment to play a much larger part in the algorithm.
One of the major failings of Google is that link equity doesn’t always correlate with a great product, competitive pricing, or a reactive customer service team.
It has even been the case in the past that companies have used deliberately poor customer service to artificially inflate links to their site.
It’s all part of a sales strategy, he said. Online chatter about DecorMyEyes, even furious online chatter, pushed the site higher in Google search results, which led to greater sales. He closed with a sardonic expression of gratitude: “I never had the amount of traffic I have now since my 1st complaint. I am in heaven.”
Google have already rolled-out Google Seller Ratings in order to collect customer feedback from stores, would the logical progression of this be to start to include some kind of sentiment and quality analysis into their shopping and organic results?
Google themselves has claimed that it is something that they are actively working towards,
As it turns out, Google has a world-class sentiment analysis system (Large-Scale Sentiment Analysis for News and Blogs). But if we demoted web pages that have negative comments against them, you might not be able to find information about many elected officials, not to mention a lot of important but controversial concepts. So far we have not found an effective way to significantly improve search using sentiment analysis. Of course, we will continue trying.
Additionally, link data could be to help to understand deeper sentiment by using the improvements that Google have made in understanding topics and language from linking pages.
It’s nuanced, but link building isn’t ALL about the link anymore, but done well, the link becomes the natural by-product of being creative, newsworthy, interesting, great at what you do –the kind of things great brands have been doing for years.
Create value, communicate that value and you’ll get the links that help you rank. After all, isn’t that what the algorithm is all about – creating value and ranking because of it?
- Links will continue to be at the core of Googles algorithm, they are too central to be extracted completely, and they are less easy to manipulate than other off page factors.
- Google is getting better at identifying and neutralising low quality links. They’re not perfect yet, but how would you futureproof your strategy?
- Move away from viewing links as a volume-based commodity, quality and relevance is where the value lies.
- Google is getting better at understanding natural language and intent. You need to focus your marketing around content that users love and interact with.
- Take a holistic approach to your digital marketing. You need links, but these should be a natural progression from the great marketing activity that your brand is doing.
- Extend this approach not just to your site, but also your product and customer service. A great customer experience extends far beyond your website.
Econsultancy offers SEO training and an SEO Best Practice Guide.
Good round up. And doesn’t “Pagerank Updated” feel a lot a like Majestic’s Trust Flow metric? ie cherry pick a core set of trusted seed sites/pages and base your PageRan/Trust flow score on the resulting link graph? Looks like Google and Majestic using similar types of seed site eg New York Times, etc. I’ve always liked Majestic’s Trust Flow – perhaps Google do too 🙂
Good round up.
Interesting Article, Matt. I agree with you on practically everything here. Good job!
You gave a detailed explanation which is making think , you are right google comes every time with an update which has got to filter he spammy links, thaks for sharing that clear information.
A pretty good roundup 🙂 what’s interesting is that, a lot of people in the industry feel that the beginnings of the link-valuation revolution began much earlier. Certainly, even before the ‘PageRank updated’ patent, it was true that PageRank (1st edition) weighted links from different websites differently. The problem was that the first sites to gain critical mass, did so because of link volumes. As such it makes total sense to have seed sites from which all equity must flow and cascade across the web.
It should be noted that not having links from seed sites wouldn’t necessarily be disastrous for SEO. As long as you had links from sites which had direct or close links from the seed sites, you could still do pretty well. Although it’s not a total defence against link-spam, it looks to be a well-shaped upgrade from the current state of affairs.
Really all of this stuff began in a large way between 2010 and 2011, as Google “May-Day” began to feed into “Farmer”, which though still fairly content-based in nature, paved the way to judging ‘linking’ sites by how ‘spammy’ they felt rather than just blindly following backlink counts. It might be interesting to include some top-line stuff on these earlier updates in this post to really cap it off!
The holistic approach is definitely the way forward. Businesses need to approach SEO as a brand building exercise rather than a money making exercise. If you build a strong brand first, the customers will come.
Thanks for this Matt and I can’t disagree with the points you make. Google’s aim is to provide the very best user experience it can, whatever that looks like and if that means changing the way they view links then so be it.
You asked “…but how would you futureproof your strategy” and my answer is simple. Don’t buy links. Don’t try and game the system.
If you approach the problem as a Marketing challenge the aim is to get more people to buy more product more often (thanks Sergio Zyman…) and links won’t directly do this.
Create something that people want to buy and then any links, no matter their quality, become earned and by definition relevant.