Keeping Up With SEO in 2017 & Beyond
Download as PPTX, PDF115 likes138,820 views
This document summarizes Rand Fishkin's presentation on how marketers can keep up with Google in 2017 and beyond. It discusses 5 key things marketers must do: 1) diversify traffic sources to reduce reliance on Google, 2) create a scalable link building strategy with decreasing friction, 3) use searcher intent and changing SERP features to break through Google's results, 4) evolve keyword targeting to match Google's sophistication, and 5) focus on searcher engagement which may be a new key ranking factor. The presentation explores how Google is increasingly using machine learning and user data in its algorithms.
Keeping Up With SEO in 2017 & Beyond
- 1. Rand Fishkin, Wizard of Moz | @randfish | rand@moz.com How Marketer’s Can Keep Up With Google in 2017 and Beyond
- 3. In the last 2 Years, Google’s Made a Lot of Big Changes
- 4. Via Google Algorithm Change History on Moz 17 Major Algorithm Updates (that we know about) from Oct 2014-Oct 2016
- 5. Via Backchannel A Move to Machine-Learning First
- 6. It’s like spotting a snow leopard in the wild. Via Mozcast The End of 10 Blue Links
- 7. Via Nick Eubanks AdWords Redacting Data Like It’s the CIA
- 8. Via Merkle PPC Growth Continues to Level Off
- 9. Google Fights Back by Making Ads Subtler But… Where Can They Go From Here?
- 10. Thanks to Clickstream Data, We Know More About How Searchers Interact
- 11. Keyword Length of Search Queries 1-Word Queries 21.71 % 2-Word Queries 23.98 % 3-Word Queries 19.60 % 4-Word Queries 13.89 % 5-Word Queries 8.70% 6+ Word Queries 12.12 %
- 13. The average searcher performs ~3 queries per day on their desktop/laptop Search Activity
- 14. 1.19% of Google.com US Searches result in an ad click Distribution of Clicks on Paid Results
- 15. ~51% of all the clicks that happen on Google.com US search results go to organic, non-Google results Distribution of Clicks on Organic Results
- 16. But 49% of clicks go to Google properties of one kind or another (Maps, YouTube, Ads, etc) Distribution of Clicks on Organic Results
- 17. ~40% of searches result in no clicks at all For weather queries like this, it’s ~75% Lots of “No-Click” Searches Happen, Too:
- 18. ~25% of desktop queries come through Chrome Instant, and many of those that load result in no clicks (as searchers are still typing/refining/choosing) Search Suggest May Be Biasing Behavior
- 19. 1) youtube.com 2) wikipedia.org 3) facebook.com 4) amazon.com 5) stackoverflow.com 6) pinterest.com 7) reddit.com 8) ebay.com 9) zillow.com 10) imdb.com 11) linkedin.com 12) yelp.com 13) xvideos.com 14) quora.com 15) tripadvisor.com 16) walmart.com 17) answers.yahoo.com 18) pornhub.com 19) nih.nlm.ncbi.gov 20) netflix.com Together, these 20 sites account for 22.8% of all Google search traffic referrals Top Sites Receiving Google US Traffic
- 20. Top 20 Websites 22.8% Top 100 Websites 30.59 % Top 1,000 Websites 45.58 % Top 10,000 Websites 62.66 % Distribution of Google’s Search Clicks
- 21. Massive Change… But What Does It Mean for Us?
- 22. The Big 5: What Marketers Must Do Differently in the Years Ahead
- 23. Diversify Traffic to Keep Earning Visits from Google #5
- 24. Google Knows Where Traffic Flows on the Web (mobile and desktop) Together we see all… We know all… There is no escape…
- 25. Google knows where visit paths begin… Where they go next… And where they end up.
- 26. If Google Knows Your Site Gets 80%+ of Your Traffic from Them, They May Take a Closer Look at Why Via Vinny La Barbera
- 27. Via SimilarWeb
- 28. via Reuters Institute US Daily Video Views (Note: Facebook counts views @ 3s, YouTube @ 30 s)
- 29. Via Cohlab YouTube is the World’s 2nd Largest Search Engine
- 30. Via DuckDuckGo And DuckDuckGo Continues its Rapid Growth
- 31. Via Search Engine Land Amazon vs. Web Search
- 32. Don’t Ignore Search Channels Just Because They’re Not Google
- 33. See Where Your Competition Gets Their Traffic: Via SimilarWeb Pro These are the sites that sent traffic to WarbyParker.com in March, 2016
- 34. Apply different tactics to reach audiences on different sites Amazon’s ranking factors can be found here. Conversion rate plays a big role.
- 35. Consider Content for Multiple Engines Moz first puts videos up on our site, then later uploads to YouTube to benefit from both classic SEO and video
- 36. Evolve Our Keyword Targeting to Match Google’s Sophistication#4
- 38. Google’s minimized the desktop search experience to make it more closely align with the mobile experience.
- 39. And they’re putting more and more “answers” atop the results to remove the need to click a 3rd-party site for information
- 40. We expected answers like these would siphon away traffic from our site… In reality, we got more!
- 41. But some answers really do remove traffic (estimates of 50%+ traffic loss to web results after SERP
- 42. Consider Click-Through-Rate When Choosing Keywords The organic results on this page probably get only ~60% of the clicks from searchers
- 43. Consider Click-Through-Rate When Choosing Keywords But, the organic results here are likely getting 100% of the clicks
- 44. You can make these estimates yourself when evaluating keywords for SEO effort:
- 45. Or you can use a tool like Keyword Explorer to get the CTR Opportunity scores 100% CTR Opportunity 60% CTR Opportunity
- 46. It’s Possible to Use SEO to Get Into Featured Snippets (and earn big CTR boosts) They beat out Wikipedia by phrasing the content to match the *answer* Google wanted for this search query
- 47. More on How to Become theAnswer: Via Dr. Pete (and more Dr. Pete)
- 48. The Classic Source of Volume Data is GoogleAdWords, and there’s a Lot of Weirdness There
- 49. These numbers actually represent ranges, even though they show as estimated averages
- 50. Russ Jones showed how Google’s volume ranges work in his post about Keyword Planner’s Dirty Secrets.
- 51. Google Trends (which we’ve found to be generally more accurate) tells a different story?!
- 52. AdWords is Now Conflating Volume of Related Keywords in Frustrating Ways Via Bill Slawski on Go Fish Digital
- 53. Keyword Suggestions inAdWords Hide a Lot of Useful, High-Volume Queries Really Keyword Planner? Really?!
- 54. It sure looks like Google knows about some closely related keywords that have search volume!
- 55. Don’t Use Google’s Volume Numbers asAbsolutes, Only as Relative Comparisons There are probably more people searching for “champagne flutes” than “toasting flutes”, but 12,100 and 1,900 almost certainly aren’t the real quantities.
- 56. Google Trends is Good for Volume Comparison “Champagne flutes” is likely ~5-10X the volume of “toasting flutes”
- 57. KW Explorer Uses Clickstream Data and Volume Ranges that May Give GreaterAccuracy In our tests, these ranges contained the true search volume 95% of the time.
- 58. Want the Best DataAbout Volume, Conversions, and Trends? You’ll need to buy the keywords and measure directly in AdWords.
- 59. Don’t Rely Exclusively on KW Planner for Keyword Suggestions/Ideas Search Suggest and Related Searches are smart additions (and totally free)
- 60. 7 Kinds of KW Research Expansions to Try: 1) Search Suggest 2) People Also Search For… 3) Similar Pages Rank For… 4) Semantically Connected Terms 5) Topically-Related Searches 6) Questions Containing these KWs 7) AdWords (Commercial)
- 61. Tools that Have These Discovery Options: 1 ) 2 ) 3 ) 4 ) 5 ) 6 ) (Free & Paid) SEMRush (Free & Paid) KeywordTool.io 7 ) (Paid) SimilarWeb (Free & Paid) Moz KW Explorer (Free) Ubersuggest (Free) Answer The Public
- 62. Use Searcher Intent & SERP Features to Break Through Google’s Changing Results#3
- 63. 18 Unique Types of SERPs that Show Up in 0.5%+ of Google’s Results One of my favorites: the “disaster type”
- 64. In our data, only ~3% of results are the “classic ten blue links” kind
- 65. SERPs like these are far more common
- 66. These types of results can dramatically reduce organic CTR (in this case, Moz estimates only ~24% organic CTR)
- 67. Infuriatingly, Google’s restricted who can appear in certain types of listings (e.g. video is now YouTube or Vimeo
- 68. And on mobile, even more kinds of searches are limited to particular networks (Google Play & iPhone App Store).
- 69. Analyze Which Types of SERPsAppear Most in the Keywords You CareAbout These show me how many of each SERP type appears in the results for this keyword list (via
- 70. Then Determine What Verticals & SERPTypes You Need to Optimize For It might pay to generate some visual charts in addition to a text-only version…
- 71. If certain searches are impossible to target…
- 72. Aim to Influence Search Suggest More on Whiteboard Friday
- 73. Be On the Right Platform For Your SERPs: Videos: E-Commerce: Podcasts: Local Businesses: News: Apps: YouTube Vimeo Facebook G Shopping Amazon eBay Etsy G Play iTunes G Maps SoundcloudLibsyn Apple Maps Bing Maps G News iTunes G
- 74. Google’s Become Masterful at Understanding a Searcher’s Intent
- 77. Keyword Matching is No Longer a Competitive Advantage Via Backlinko
- 78. On-Page SEO is No Longer Satisfied by Raw Keyword Use Either These keywords are nearly as important to use on a page targeting the query as the query keywords themselves
- 79. It’s Still Wise to Use Keywords
- 80. Matching Your Content to What SearchersAre Seeking is Critical, Too Trulia knows searchers want home prices, by neighborhood, with trend data, zipcode filters, and a zoomable map.
- 81. Use of related topics to indicate the content’s relevance Serving KWs w/ matching intent together on one page Thorough answers/solutions to the searcher’s query Unique value over what other sites in the SERPs Intelligent keyword use in page title, meta description, URL, and top few paragraphs (just as before) On-Page SEO in 2016 Requires
- 82. More in this Whiteboard Friday Video
- 83. Create a Link Strategy That Scales with Decreasing Friction #2
- 84. Every Link Builder Needs a Strategic Roadmap: Link Goals Strategic Approach Tactical Initiatives KPIs/ Metrics
- 85. An Example: Bookserf, An Istanbul Based Book-Sharing Community
- 86. Bookserf’s Long-Term Link Strategy Rank in search for book names (especially in Turkey & other operating regions) Encourage members to link to their profiles & favorite books; partner w/ retailers & authors to earn links from their Link nudges during account setup, and upon completion of activities on the site. Direct outreach to authors/retailers. 1. Links from members 2. Links & mentions from authors/retailer s 3. Rankings for key terms 4. Search visits Link Goals Strategic Approach Tactical Initiatives KPI Metrics
- 87. We Also Need… Buy-In on Experimentation Via PointBlankSEO and Allie Brown Link tactics like an event photo gallery might work, but it also might fall flat. Link building needs room to try and fail.
- 88. The Right Expectations Regarding Time-to-ROI Via WB
- 89. A Balance Between Long-Term Investments & Short-Term Hacks High upfront costs Pay (in time/$$) as you go Long-Term Investments Slow to show ROI Earn links while you sleep Non-existent Spam Risk Can Show Fast ROI Effort In = Links Out Can Have Spam Risk Short-Term Hacks
- 90. Great Link Builders Focus on Their Flywheel
- 91. Moz’s Flywheel KW Research + Industry Intuition Publish Content Promote via Social Channels Push to email + RSS subscribers Earn Links + Amplification Grow social, email, RSS, & WoM channels Grow Domain Authority Earn Search & Referral Traffic Rank for More Competitive KWs
- 92. KW Research + Industry Intuition Publish Content Promote via Social Channels Push to email + RSS subscribers Earn Links + Amplification Grow social, email, RSS, & WoM channels Grow Domain Authority Earn Search & Referral Traffic Rank for More Competitive KWs In 2006, I was lucky if new content led to any links or rankings at all
- 93. KW Research + Industry Intuition Publish Content Promote via Social Channels Push to email + RSS subscribers Earn Links + Amplification Grow social, email, RSS, & WoM channels Grow Domain Authority Earn Search & Referral Traffic Rank for More Competitive KWs But, by 2016, nearly every post earns a handful of links, & some earn a lot!
- 94. Moz is a Business Built on Content
- 95. Dollar Shave Club was Built on a Press & Media Flywheel Via NYTimes
- 96. Dribbble was Built on a UGC/Community Flywheel
- 97. But… Almost Every Flywheel Finds a Point (or Points) of Friction Publish Content Promote via Social Channels Push to email + RSS subscribers Crap… We’re not reaching anyone who will link/amplify us
- 98. Link Building Growth Hacks Can Accelerate a Flywheel
- 99. Hacks aren’t necessarily evil, spammy, or without value. They can be highly useful when applied to a sound link strategy.
- 100. The Formula for Link Building that Will Work in Google for Years to Come: Strategic Roadmap Scalable Flywheel Long-Term Investments Hacks to Remove Friction + + +
- 101. Searcher Engagement May Be the New Silver Bullet in Rankings#1
- 102. RankBrain is Only the Most Obvious of Google’s ML- Based Ranking Elements Via Gianluca Fiorelli on Moz
- 103. Rankbrain helps Google know that results like these are relevant to searches like this
- 104. And that all of these queries probably share intent and should bring back similar kinds of
- 105. Google Uses User, Usage, & Engagement Data to Determine How to Rank Content I hate bullet points, but this slide from Google Search Engineer Paul Haahr, shows how clicks are used to grade performance Via How Google Works on Slideshare
- 106. This is a good SERP – searchers rarely bounce, rarely short-click, and rarely need to enter other queries or go to page 2.
- 107. This is a bad SERP – searchers bounce often, click other results, rarely long-click, and try other queries. They’re definitely not happy.
- 108. Via Steve Hammer
- 109. Someday,Algorithms Built by Machines Will Outperform those Hand-Selected by Engineers Potential Ranking Factors (e.g. PageRank, TF*IDF, Topic Modeling, QDF, Clicks, Entity Association, etc.) Training Data (i.e. good search results) Learning Process Best Fit Algo
- 110. Google Leverages Machine Learning Despite Not Knowing for Sure What It Uses: Via SERoundtable
- 111. We may soon hear that algorithmic elements are no longer applied universally. Rather than one algorithm, we get thousands or millions of them.
- 112. Google’s PublicAbout Their Commitment to ML Techniques… Won’t Be Long Now. Via BackChannel
- 113. Focus on Signal:Noise Ratio – Don’t Let Bad Pages Drag Down a Good Site Quantity of Pages Earning SERP Visits Relative Time on Site, Bounce Rate, Pgs/Visit, Searcher Satisfaction Site’s Search Engagement Reputation =
- 114. These don’t look so good; probably worth investigating.
- 115. Find Ways to Beat Your Competition’s & YourAverage Ranking Position’s CTR If you rank #3, but have a higher- than-average CTR for that position, you might get moved up. Via Philip Petrescu on Moz
- 116. Serve Multiple Searcher Intents, Not Just Your Own Interests You can’t just try to sell sous vide cooking devices to the <1% of searchers who are ready to buy after performing this query. To compete long term, you’ll need to empathize with and serve ALL of these intents with your ranking
- 117. Note that NONE of these SERPs are trying to sell equipment directly. They’re serving the *intent* of the widest range of searchers.
- 118. Speed, speed, and more speed Delivers an easy, enjoyable experience on every device Compels visitors to engage, share, & return Avoids features that dissuade or annoy visitors Authoritative, comprehensive content that’s uniquely valuable vs. what anyone else in your space provides Make UX a Cornerstone of Your SEO:
- 119. Deliver Dramatically Better Engagement Than Your SERP Competitors & They Will Be Hard Pressed to Catch Up See this list of 10X Content for examples and resources on how to create it
- 120. Massive Change… Bring It On.
- 121. Rand Fishkin, Wizard of Moz | @randfish | rand@moz.com Bit.ly/keepup2017