Rand Fishkin — Search Ranking Factors in 2015 What Data, Opinions, and Testing Reveal

Transcript

1. Rand Fishkin, Wizard of Moz | @randfish | rand@moz.com Search Ranking Factors 2015 What data, opinions, and testing have revealed about how Google’s rankings operate.
2. Slides Will Be Publicly Available October 20th, 2015
3. A look at Google’s algo in 2015 according to 150 professional SEOs
4. We used to show graphics like this to illustrate the relative importance of different areas of optimization to Google’s 2013
5. But a pie chart suggests that you can only get so much value from any given set of features. In reality, factors like higher link authority on your domain have as almost unlimited ability to positively influence rankings
6. Thus, we’re switching up how we illustrate opinions about the factors:
8. Most interesting to me is what’s happened to SEO professionals’ opinions over time…
9. 2015 (in blasphemous pie chart form to illustrate comparative change)
10. 2009 2011 2013 2015
11. A few of the opinions about factors in particular stand out:
12. Page-Level Link Features Domain-Level Link Features Page-Level Keyword Features 2009 2011 2013 2015 43% 22% 19.15% 14.54% 2009 2011 2013 2015 24% 21% 20.94 % 14.60% 2009 2011 2013 2015 15% 14% 14.94% 13.97%
13. 1) Professional SEOs feel that, on average, the algo is flattening, and the days of a single factor having an overwhelming impact are fading. Takeaways:
14. 2) After years of dominating the algo, links, while still powerful, don’t feel like an overwhelming ranking force to SEOs. Takeaways:
15. Takeaways: 3) Engagement data is on the rise. If growth rate continues, by our next survey, it may be in the top two features.
16. How do various web metrics correlate with higher Google rankings in 2015?
17. An important reminder about correlation: Correlation DOESN’T tell us why one page ranks higher than another. It DOES tell us what features higher-ranking pages tend to have over their lower ranking peers.
18. Do correlation coefficients in the 0.1 – 0.4 range (typical for single factors in search engine studies) mean anything? Debunk statements about what’s NOT causal in rankings 3 Useful Applications: Show relative potential influence ID factors for more testing / investigation
19. Debunking myths with correlation data is easy: Google are losers! The more ads you show, the higher they rank you.
20. A negative correlation of -0.03 disproves the idea that more ad slots = higher rankings.
21. Coefficients can also be used to show relative correlation: The best SEOs use multiple repetitions of keywords in their titles. I guarantee it works better than some fancy LDA model.
22. On average, content that better fits an LDA topic model dramatically outperforms KW repetition in the title
23. Correlation numbers can lead us to interesting theories that we can then validate through other means: Could it be that partial match anchor text now has equal or greater ranking influence than exact match?
24. Let’s go run some experiments to see if this is true y’all!
25. NOTE: In an algorithm with 100s – 1000s of ranking inputs, we shouldn’t expect any single element to have the kinds of high correlations seen in less complex input scenarios. Single factors correlate with higher Google rankings in this range.
26. In May 2015, Moz collected 16,521 unique SERPs from Google.com (US). Full methodology here
27. Look Familiar? Link metrics’ correlations w/ rankings have remained similar for ~6 years
28. Moz & Ahrefs For the first time, we compared Mozscape’s link correlations against Ahrefs… And found nearly identical results for both.
29. Social Shares Correlations are down ~10-15% from their high in 2013.
30. Traffic & Engagement For the first time, we measured usage data. While traffic looks strongly correlated, engagement metrics have weaker numbers. Traffic and engagement metrics via
31. Keyword Use & On-Page Optimization As we get more sophisticated in our text-modeling abilities, we’re seeing higher correlations (though still low relative to links & social shares)
32. For the first time, we also broke correlations down by category of keywords/SERPs
33. Health websites that link out more tend to rank higher. Dining sites see almost no correlation between linking out & ranking.
34. It tended be more present in higher ranking sites for these verticals Anchor text had a smaller relationship w/ high rankings in these verticals
35. Those meager restaurant websites? Looks like Google doesn’t mind much. Buzzfeed & Upworthy are always showing how lengthier articles perform better for them.
36. Twitter & Facebook have very similar relative correlations, which fits w/ Google’s statements that they don’t directly use either. In some verticals, social sharing is much less connected to ranking positions than others
37. 1) Correlations with links have remained relatively similar, suggesting that perhaps links haven’t faded in influence as much as some in our industry have suggested. Takeaways:
38. 2) We need more sophisticated on-page analysis tools. With the right algorithms/ software, we may find real opportunities to improve rankings through content. Takeaways:
39. Takeaways: 3) Correlation is even more useful (and interesting) on subsets of SERPs than on an entire corpus. In the future, calculating correlations for the SERPs you/your company care about may become standard.
40. A Look at Links & Social Shares in Google’s Rankings
41. We know that links can still overwhelm other ranking signals. Via Rishi Lakhani on Refugeeks Pointing a few anchor-text links at this blocked-by-robots page on Matt’s blog made it rank (even in 2015).
42. 20 Elements of a Link’s Ability to Influence Ranking: 1) Anchor Text 2) PageRank 3) Relevance 4) Domain Authority 5) Location on the Page 6) Internal vs. External 7) Quality of Other Links on Page/Site 8) Editorial Weight 9) Engagement w/ Linking & Linked Pages 10) Follow vs. Nofollow
43. 20 Elements of a Link’s Ability to Influence Ranking: 11) Source Depth 12) Text vs. Img 13) Link Age 14) Topical Authority of Source 16) Spam Signals 17) Speed/Acceleration of New Link Sources 18) Author Authority 19) 1st Link to Target on Page vs Duplicate Links 20) Prior Links to Target from Source Domain 15) Javascript vs. HTML
44. This stuff mattered a lot when we did manual link building to move rankings But today, many of us just let content build links for us, right?
45. Moz & Buzzfeed joined forces for a report looking at 1 million pieces of content. Data via Buzzsumo & Moz’s Joint Study
46. Content + Social Sharing = Links? Median # of links across a million pieces of content in Buzzsumo’s database?.... 1 linking root domain. Data via Buzzsumo & Moz’s Joint Study
47. This is a power law distribution – the top content gets the overwhelming majority of links and shares.
48. The reality of social amplification and earning links is… 0.028? That’s too close to 0 to infer any consistent, direct influence.
49. For the most heavily shared content, there’s a little bit more of a correlation, but it’s small enough that relying on social shares to earn your links is probably folly. We tried segmenting the samples:
50. This data shows why I can’t endorse either of these common maxims in SEO and content marketing: Create good, unique content and Google will figure out the rest. The best way to earn links is to create great content.
51. In the past, I presented a concept that, based on this data, now appears to be fundamentally flawed:
52. Publish Amplify Grow network Rank for slightly more competitive terms & phrases Get links Grow authority Earn search traffic
53. 1) Social shares by themselves almost never lead directly to the quantities of links necessary to rank well. Takeaways:
54. 2) Content that performs extraordinarily well on social networks and ranks well in search engines may not be benefitting solely from links. Takeaways:
55. Validating Some of Google’s Statements On Secure Sites
56. Via Google Webmaster Central Blog
57. Via Rand’s Google+
58. HTTPS URLs have a 0.04 correlation w/ higher rankings… much lower than many features Google says don’t impact rankings.
62. Investigating SEOs’ Longstanding Theories re: Raw URL Mentions
63. Using data from Fresh Web Explorer, we can see how many mentions a URL receives in a given day/week/month
64. The correlations w/ URL mentions are pretty high – in the range of social shares and links (0.19 for full domain, 0.17 for root domain)
65. Via Stone Temple Blog (and IMEC Labs) So, the crew at IMEC Labs ran a test!
66. Via Stone Temple Blog (and IMEC Labs) Not the easiest graph to read, but the results suggest that raw URL mentions had no impact on rankings, certainly nothing like the impact that links do.
67. bit.ly/rankingfactors2015 All the data from the ranking factors report can be found at:
68. Thank You! Rand Fishkin, Wizard of Moz | @randfish | rand@moz.com