SlideShare a Scribd company logo

Links As SEO Ranking Factor: Still Very Powerful!

•Download as PPTX, PDF•
•
Eric Enge
Eric Enge

Our in-depth study used advanced statistical analysis to show that links to your site are even more powerfully correlated with ranking than most people think. See the complete study at stonet.co/link-study.

1 of 24
A Study by Eric Enge
Complete Study at stonet.co/link-study
You’ve seen the claims...
“So many new ranking factors, and they’re all important!”
“Content quality is all that matters!”
“RankBrain has taken over all Google algorithms!”
Um...no. In fact, links remain a very important ranking factor. Let’s dig into the data...
Google’s Andrey Lipattsev youtu.be/l8VnZCcl9J4 Content and links going into your site are the two most important ranking factors.
Link correlations in Moz & SearchMetrics studies bit.ly/moz-ranking & bit.ly/searchmetrics-rankfactors
Problem: Those are significant correlations but, about the same as other highly- correlated factors in those studies. So why would Google say links are still so high up in their ranking factors?
Two Key Factors: Both studies evaluated each SERP individually, then took the mean of all results. Both studies also focused solely on commercial terms.
We took a different approach that statistics experts told us was more valid (details at stonet.co/link-study)
We also tested more varied query types, including: Commercial head terms Commercial long tail terms Informational terms (In fact, 2/3 of our queries were informational)
We found a correlation of total links to a site as higher than correlations of Moz DA or PA.
And, in fact, the correlation of total links to as site was much higher by our methodology than in the other two studies
We also normalized the links across SERP positions in aggregation to counteract false negative correlations.
This “smoothing” process yields even higher correlations for total links to a site.
And when we took the aggregate correlations across blocks of 10 search results, the Spearman Correlation jumped to perfect!
I believe that these aggregated calculations show a far truer picture of the true power of links than the mean-based methods of the other studies.
By the way, a manual analysis of several hundred results revealed that the following types of results are likely NOT as influenced by links: • Local results (not maps results, but results that are locally influenced) • Query deserves diversity • In Depth Articles These accounted for about 6% of the results.
But all this does not mean your content quality is irrelevant. Far from it! If your content is not relevant or competitive, links won't help ranking. If it is, links will make the difference.
Our own experience working with many Fortune 500 clients bears out the value of good backlinks to quality, relevant content.
IMPORTANT: We don’t find that links can rescue poor quality content, or cause low relevance content to rank. Also, all of our efforts focus on getting recognition from, or content published on, very high authority sites.
Complete Study at stonet.co/link-study

More Related Content

Links As SEO Ranking Factor: Still Very Powerful!

  • 1. A Study by Eric Enge
  • 3. You’ve seen the claims...
  • 4. “So many new ranking factors, and they’re all important!”
  • 5. “Content quality is all that matters!”
  • 6. “RankBrain has taken over all Google algorithms!”
  • 7. Um...no. In fact, links remain a very important ranking factor. Let’s dig into the data...
  • 8. Google’s Andrey Lipattsev youtu.be/l8VnZCcl9J4 Content and links going into your site are the two most important ranking factors.
  • 9. Link correlations in Moz & SearchMetrics studies bit.ly/moz-ranking & bit.ly/searchmetrics-rankfactors
  • 10. Problem: Those are significant correlations but, about the same as other highly- correlated factors in those studies. So why would Google say links are still so high up in their ranking factors?
  • 11. Two Key Factors: Both studies evaluated each SERP individually, then took the mean of all results. Both studies also focused solely on commercial terms.
  • 12. We took a different approach that statistics experts told us was more valid (details at stonet.co/link-study)
  • 13. We also tested more varied query types, including: Commercial head terms Commercial long tail terms Informational terms (In fact, 2/3 of our queries were informational)
  • 14. We found a correlation of total links to a site as higher than correlations of Moz DA or PA.
  • 15. And, in fact, the correlation of total links to as site was much higher by our methodology than in the other two studies
  • 16. We also normalized the links across SERP positions in aggregation to counteract false negative correlations.
  • 17. This “smoothing” process yields even higher correlations for total links to a site.
  • 18. And when we took the aggregate correlations across blocks of 10 search results, the Spearman Correlation jumped to perfect!
  • 19. I believe that these aggregated calculations show a far truer picture of the true power of links than the mean-based methods of the other studies.
  • 20. By the way, a manual analysis of several hundred results revealed that the following types of results are likely NOT as influenced by links: • Local results (not maps results, but results that are locally influenced) • Query deserves diversity • In Depth Articles These accounted for about 6% of the results.
  • 21. But all this does not mean your content quality is irrelevant. Far from it! If your content is not relevant or competitive, links won't help ranking. If it is, links will make the difference.
  • 22. Our own experience working with many Fortune 500 clients bears out the value of good backlinks to quality, relevant content.
  • 23. IMPORTANT: We don’t find that links can rescue poor quality content, or cause low relevance content to rank. Also, all of our efforts focus on getting recognition from, or content published on, very high authority sites.