Penalty Safe SEO & The Power of Domain Authority
If you are interested in ranking in Google, you have probably noticed how powerfully Google favors brands. In a study I published last year on Moz, I share my findings where I analyzed over 100 keywords, and discovered that a large majority of the top 10 spots were occupied by massive brands with high domain authority.
Before we continue, let’s run through a test together. I’m going to use a few keywords gathered from Wordstream’s list of popular keywords. Let’s type them into Google, and look at the top 10 results:
- Luggage set
- Coffee maker
- Christmas gifts
As you can see from just these 3 examples, the first page results are heavily dominated by large, established, well-known brands. If you choose random keywords yourself and search on Google, you’ll also find a similar trend. Of course the results are determined by the nature of the query – transnational, navigational, or informational – so the key to ranking in today’s Google algorithm is to identify queries that will bring you high ROI, create content relevant to the query, and build authority to the domain so that pages with relevant content will rank. If you build trusted links to the domain to generate domain authority, then when Google finds pages on your site that match a user’s query, the likelihood that your site will rank is much higher than for domains with low trust and domain autority.
Domain Authority: TrustRank combined with PageRank
To understand domain authority, we have to remember the essence of Google’s ranking algorithm. Back in 1996 we had Pagerank, which was the original “popularity” algorithm. In 2009, Google was granted a patent for “TrustRank” which is a ranking modifier based on a measure of trust. Trustrank works based on measuring how closely linked a domain is from a trusted seed set of domains.
To understand how simple it would be for Google to compile this seed set, let’s look at SEMrush’s winners and losers list:
SEMrush spiders the web, parses a site for top keywords, identifies where those sites are ranking for those keywords, and can then anticipate traffic based on keyword data from Google’s Keyword Planner; all of this from a 3rd party tool that doesn’t have even a fraction of Google’s resources.
Let’s assume that Google has a similar list that we can then call a trusted seed set. Google’s trustrank algorithm measures how closely linked your domain is to the set of trusted links, which generates a score for each domain. Google combines this score with popularity to find a holistic ranking, which eliminates much of the “spammy web”.
“Our experimental results show that we can effectively identify a significant number of strongly reputable (non-spam) pages. In a search engine, TrustRank can be used either separately to filter the index, or in combination with PageRank and other metrics to rank search results.”
Rand Fishkin from Moz created a highly informative video about this. In fact Moz’s Domain Authority algorithm closely mimics Google’s algorithm, using many of the same signals. In the video, Rand explains that Trustrank is a combination of trust, registration information, and user data signals using information gathered from Analytics, 3rd party relationships, and search data. He also explains Domain Authority in the context of pagerank, diversity of link sources, temporal data, and link distribution. The 15 minute video will give you a crash course on the basis of Google’s current ranking algorithm. To see the effects, all you have to do is look at the SERP’s for thousands of keywords and you’ll see how trustrank and domain authority impact rankings.
Quality Score Algorithm
In 2011, Google was granted a patent which is considered to be the birth of the Panda Quality Score algorithm. The patent describes an algorithm that associates a document with a search query, derives a value for the document based on the associations, and uses that value to evaluate the quality of a document for subsequent queries.
Let’s consider how Google applies Quality Score to their Adwords systems. Every keyword you add to your account gets a quality score, which is a measurement of relevance (defined by CTR, historical account performance, landing page quality and relevance, and relevance of the keyword to the ad text and ad group). Quality Score is estimated when an account is first launched, but performance of the account will then allow quality score to migrate from the keyword level, to the ad group level, to the campaign level, to the account level. Once the account has a defined quality score, new keywords added to that account will inherit a quality score from the account level quality score (as well as other factors).
In the same way, there’s a quality score in Google Organic whereby a domain is assigned a score based on a value measured by a variety of signals. That quality score then impacts all of the pages of that domain, and affects how those pages rank. In yet another patent, we can see how Google groups pages by domain algorithms and uses that to modify rankings.
Deep Link Ratio and Distribution Analysis
Another footprint used for penalties is Deep Link ratio. When a site has too many inbound links to the home page, this is unnatural. Authoritative, high quality websites that are producing valuable content typically have links pointing to many of their internal pages. Why would a site contain 90% of their links to the home page? Is this natural in this keyword niche? If not, this unnatural pattern sticks out as a footprint and becomes a red flag for penalties.
In Rand’s video (mentioned above), he talks about Distribution Analysis. How valuable is a site that has a high percentage of links going to a couple of internal pages? How many total pages does the site have? And how are these links distributed?
Google’s Webmaster Tools has a report called “Links to Your Site”, which shows how many links each internal page has. Here’s an example for www.contentkarma.com:
This report makes it obvious that Google considers this important enough to track and report to webmasters.
Focusing links on a small amount of money pages is unnatural and reduces quality signals for the domain. To rank well, it’s important to have links to as many pages as possible on your domain.
Block Level Link Analysis
The placement of the links is also important. The block level link analysis patent is proof that Google can parse documents and attribute importance to content or links based on their location on the page.
They can parse the header, footer, and navigation to identify the “main body” of a web page, and then place higher value on those links. Placing links in the footer or sidebar, a practice often used by people buying links, may render them useless by this algorithm. The most valued links are those that are in the main body of the page, particularly towards the top or middle of the article.
User Generated Data, Traffic, and Footprints
Finally, let’s study another determining factor: user generated data. Given the prominence of Google Analytics, we know that Google has substantial data about how users act when they visit a website. We also know they track CTR from the Google SERP’s as a signal of quality, as well as bounce rate – how long does it take for the user to come back to Google?
We know Google’s goal is to provide their users with the best searching experience, as their revenue stream is closely tied to their search product. How often people use Google to search determines how much revenue they generate from Adwords. As such they use all available signals, and user behavior data is incredibly important.
Combining this user data with footprint indicators such as genuine traffic generated from links and total c classes from inbound links helps to both generate a quality score for user experience as well as look for signals of spam. This patent granted in September 2014 to Google mentions many of these footprints and can help us understand how Google uses these elements as rank modifiers.
The Demise of Anchor Text & Money Keywords as a Ranking Signal
Back in the early 2000’s, Google started using keywords in anchor text as a powerful ranking signal. If sites were linking to yours using a particular term, then it was likely that those terms indicated relevance. For 8 years, anchor text was one of THE most important factors in the ranking algorithm, and fortunes were made by people buying links or using link schemes with money keywords in anchor text to rank.
In April of 2012, the first Penguin penalty ended the reign of anchor text. After 10 years of people creating money anchors, it became easy to identify sites building links or manipulating link profiles by simply turning that signal around into a footprint for penalties. The foundation of the Penguin penalty filter is to look for sites that have a high density of keywords in anchor text, and to apply a penalty.
Using data from Adwords to quantify what keywords are transactional, Google can quickly/easily classify keywords in your link profile as “money” terms, and if the ratio of money terms to brand/noise terms is too high, you can automatically receive a penalty. This represents the evolution of the Penguin algorithm. Early Penguin looked at the density of specific keywords, while newer Penguins look at ratios of money vs brand and classify the profile as a whole.
As a result, building links with keywords in money terms will eventually result in a penalty. Having the occasional money link won’t hurt, but the moment the profile starts gaining a critical density – enough for this density to influence the rankings – you’re also waving a red flag to attract a penalty.
Penalty Safe SEO
Now let’s put all of our findings together to identify how to engage in penalty safe SEO.
- Links continue to be a powerful ranking signal
- Sites need links from trusted sources to influence their Domain Authority
- High domain authority is a combination of quantity of links (pagerank) with the trust score of those links (trustrank)
- Domain Authority heavily influences rankings: if your domain is trusted, pages on that domain will rank higher than non-trusted domains
- If your page is relevant for a query based on on-page signals and semantic markers, then the Domain Authority modifier impacts how you rank (your page may be relevant to a query, but if your domain authority is low, less relevant pages will rank above yours).
- Links in to the site should be evenly distributed amongst internal pages
- Links should generate traffic – links without traffic are a signal of manipulation
- Links should be in the main body of the page surrounded by relevant content
- Using money keywords in anchor text is a footprint for penalties
- Time on site, bounce rate, page load speed, SERP CTR and other user behavior signals impact the quality score of your site, which impacts your rankings
Taking these factors into consideration, how should you craft your SEO strategy?
It’s simple. Earn relevant links to internal pages of your site to increase your domain authority. These links should be earned on the foundation of a content marketing strategy. Combine powerful content with outreach and social promotions to attract links.
Links should be editorial, in the body of the article, linking in to a variety of pages on the target domain. Anchor text should be natural and contextual based on the topic of the destination page.
Optimize your site with traditional on-page SEO techniques. The combination of relevancy with domain authority will generate those coveted top rankings.
Most importantly, avoid anything that creates footprints, such as links from the same c class, links from domain with similar page or domain structures, etc. Any unnatural footprint can be detected and eventually result in a penalty.
Those who manage to establish authority will reap massive rewards in the long run. Most of your competitors will get eaten by Google’s scary beasts, while you slowly and steadily build domain authority to help position your at the top of the SERP’s for years to come.
Latest posts by Marcela De Vivo (see all)
- The Evolution of Data: Creating Intent-Led Digital Strategies - 29 January, 2019
- Productive Things To Do When You Are a Freelancer Job-Hunting - 18 July, 2018
- What KPIs Should I be Using to Measure my SEO Campaign - 21 July, 2017