Over time, the analysis from Moz has become more complex with a move from trusted experts’ opinion to additional correlation based analysis. In many ways, the original breakdown of on and off page factors is always preferred for its simplicity, yet there are many more factors to consider now of course. In today’s post, we will highlight the main ranking factors, with a brief analysis of the implications for marketers who aren’t full-time SEO.
The Experts’ opinions on ranking factors
The best quick summary on ranking factors for non-specialists is, for me, still the Experts’ analysis. Here’s the latest from 2015, where respondents rated the relative levels of influence exerted by broad areas of ranking factors on a scale of 1 (not influential) to 10 (highly influential):
The chart shows, according to Moz, that:
- Links (to the domain overall and individual pages) are still believed to be the most important part of the algorithm. ‘Link-building is not dead’ as some content marketers have proclaimed, but ‘link earning’ is where we should be focusing.
- Keyword usage on the page is still fundamental, and other than links is thought to be the most important type of factor.
- SEOs think social sharing has an influence but it is relatively low
Studies according to correlation studies
For the study, Dr. Peters examined the top 50 Google results of 16,521 search queries, resulting in over 700,000 unique URLs (methodology). Moz cautions that correlation does not equal causation. However, correlation studies give a quantitative method of showing the potential relative influence of different factors. The full report has many correlations of which this analysis, at a page level shows that at page level overall page authority is most important, particularly unique linking domains, but with internal links having some influence.
This confirms that simple measures of the number of keywords like the number of keyword matches in the title or main heading doesn’t have an influence as it would have in the early days of SEO. Instead, semantic similarity, as shown by techniques like Latent Dirichlet Allocation or TF*IDF are significant today. Essentially, this means that’s it’s important to not only target specific keywords in titles, headings or body text, but to clearly show the theme of a piece of content by using synonyms and related concepts for a theme. Further correlations based on site engagement (dwell time) show that these also have an influence as many have speculated since Google can detect time between queries.