How Search and Social Platforms are Personalized to you

Updated November, 2014. This presentation covers how Google, Bing, Facebook, Twitter, YouTube and LinkedIn decide which information to present to their users and how as a marketer, an understanding of those algorithms should change your strategy and approach.

Transcript

1. THE LATEST IN SEARCH & SOCIAL December 3, 2013
2. SEARCH ENGINES 101
3. Google and Bing1 create their listings automatically. They use “spiders” or “bots” to "crawl" links (or social shares) to web pages and social profiles and add those web pages & profiles to their index. When a human visitor to the search engine puts in a keyword phrase, the search engine is focused on serving relevant, fresh, quality content that the engines think is matched to the searcher’s intent & personalized for the user. Keyword Phrases LET’S COVER THE BASICS
4. THE GOOGLE ALGORITHM • Algorithm has 200+ factors, some weighted more than others. We know about 40 parts of the 200. • Google changes the “secret sauce” 1.5 times a day. • Internet was built as a thesis reference system – summaries, headers, and “footnotes” matter. Results are influenced by: • Personal, location, and behavioral search • Google Instant • Knowledge Graph • Mobile device • Query Deserves Freshness • Social network and authority
5. RECENT GOOGLE ALGORITHMS
6. GOOGLE’S RULES: CONTENT QUALITY
7. GOOGLE BEST PRACTICE: ONLINE READING
8. BING RULES: ONLINE READING ”Your content should be easy to navigate to, rich enough to engage the visitor and provide them the information they seek and as fresh as possible.”
9. THE MOST IMPORTANT TRANSITION From keyword based results → Entities and connections based results Cartoon on left by Peter Steiner © 2012
10. A QUICK HISTORY OF ONLINE PERSONALIZATION
11. SEARCH ENGINES – 1994-2007 2002 • Google News 2004 • Book Search • Scholar 2005 • Blog Search • Google base 2006 • Google Video 1994 THESIS SHARING. • Search engines needed links to find and crawl your site. • Labeling mattered 1996 Google! • Introduction of PageRank • Some links mattered more than others 2002- 2006 GOOGLE INTRODUCED VERTICALS 2007 GOOGLE COMBINED THEM WITH UNIVERSAL SEARCH
12. SEARCH RESULTS – 2008-TODAY 2008 SEARCH RESULTS BECOME PERSONALIZED • Behavioral intent • Universal search • Geographic location • Previous search history 2009 GOOGLE ANNOUNCES THAT IT MAKES CHANGES DAILY, ADDS REAL TIME SOCIAL RESULTS BING! 2010 GOOGLE ADDS MOBILE SEARCH • Begins to associate the recordings of the words that you ask for with your Google account. 2011 BING-FACEBOOK LIKE INTEGRATION GOOGLE+ - GOOGLE’S IDENTITY NETWORK
13. Paid Ads Knowledge Graph My brother Geo Local results Images from Google+ All personalized to me. Direct answers are now displayed in search. GOOGLE SEARCH – 2012
14. GOOGLE – POWERED BY YOUR GOOGLE DATA
15. Google Search – Nov 2014
16. GOOGLE+ IS THE NEXT GOOGLE (GOOGLE 2.0) Google+ is the official profile with the world’s largest search engine. You are giving Google data: • About you as a company • About who your company is connected to • Identifying your customers (if they follow your page) • Highlighting what other social media networks you have Google is using that information to ensure that they are showing your online content to the people who are looking for you.
17. “Within search results, information tied to verified online profiles will be ranked higher than content without such verification, which will result in most users naturally clicking on the top (verified) results. The true cost of remaining anonymous, then, might be irrelevance.” - from Schmidt's upcoming book, "The New Digital Age."
18. BING’S SOCIALIZED SEARCH
19. BING – POWERED BY YOUR MICROSOFT DATA
20. TWITTER Personalized to you:
21. PERSONALIZATION ON TWITTER 2006 TWITTER LAUNCHED 2012 Rolled out new version of the Discover tab that is even more personalized for you. Includes: • using Cassovary, their graph processing library, to identify your connections and rank them according to how strong and important those connections are to you signals from the accounts you follow and whom they follow • Location • Language preference • Gender Related to searching Twitter: They determine whether you are searching for an “entity in a tweet”, and if so they surface the video or photo tagged with that keyword that has been shared the most.
22. AI & TWITTER Twitter has built a real-time human computation engine that determines what we see: 1. Identifies search queries as soon as they’re trending 2. Send these queries to real humans to be judged 3. Finally, incorporates the human annotations into their back-end models 4. Suggests tweets from people who don’t even follow
23. FACEBOOK Personalized to you:
24. PERSONALIZATION ON FACEBOOK 2004 FACEBOOK IS CREATED 2006 FACEBOOK NEWS FEED IS CREATED 2010 EDGERANK ANNOUNCED • like button released along with instant personalization on other sites • partners may only use your public information experience 2012 EDGERANK ADJUSTS RANKING BASED ON BRAND AFFINITY AND NEGATIVE FEEDBACK RECEIVED 2013 FACEBOOK ADDS GRAPH SEARCH
25. AI AND FACEBOOK Facebook Graph Search • Users can search for people and things as opposed to keywords • Machine learning to spot spam, ensure duplicate posts don’t appear
26. YOUTUBE Personalized to you:
27. PERSONALIZATION ON YOUTUBE Each user’s YouTube’s home feed is personalized: • Comments and subscription activity • Google+ network activity • Popular videos with shared interest audiences. • Previous viewing activity “YouTube will be incorporating machine learning into the algorithm to base recommendations off of the interactions users are making with channels and videos.”
28. LET’S TAKE A LOOK AT: PERSONALIZED SEARCH RANKINGS
29. CHOCOLATE SCREENSHOTS Female, 58, Rockville, MD – anonymous Female, 58, Rockville, MD – anonymous
30. Female, 24, Washington, DC, Ketchum employee CHOCOLATE SCREENSHOTS
31. Male, 34, Bangladesh – logged in Male, 34, Bangladesh – not logged in CHOCOLATE SCREENSHOTS
32. Female, 26, NJ CHOCOLATE SCREENSHOTS
33. LET’S TEST WHAT YOUR AUDIENCE SEES Hypothesis: • Hertz’s On Demand service is not appearing in Google when their target demographic searches for car sharing services Target demographic: • Mobile using • Ages 21-35 • college educated • Doesn’t own a car • lives in NYC and DC (where they have launched their services). Scenario: • You're looking to make a trip to the grocery store, and anticipate needing a car to transport your goods. • We are interested in understanding how you would go about finding and selecting a “car sharing” service to be used in the next few hours?
34. RESULTS OF THOSE TESTS https://www.usertesting.com/highlight_reels/kmHzjVj71UJP4CfyxpTE
35. REMOVING SEARCH PERSONALIZATION Think about your underlying assumption…. When you non-personalized the search results, are you going to see what your target audience sees?
36. CREATING PERSONAS FOR YOUR TARGET AUDIENCE • Demographics (male/female, marital status, age range income) • Job level (if B2B) • Pain points • Level of sophistication with technology • Routine for a typical day • Information sources/social networks • Keyword terms • Mobile use • Objections to attitude change
37. FIGHT BACK BY PLAYING THE GAME
38. TARGET YOUR CUSTOMER 1. Use personas 2. Target quality content to your audience 3. Build social relationships 4. Know the signals and use them 5. Measure it to see if it’s working
39. HOW TO BREAKTHROUGH THE GOOGLE FILTER: • Solid SEO foundation and quality content (webmaster guidelines) • Google+ (verified profiles, posting, Community involvement) • Schema • Social signals and voice data
40. HOW TO BREAK THROUGH THE BING FILTER: • Good SEO Foundation • Social connections to you • Your Klout Score • Schema mark up • Bing verified profile
41. HOW TO BREAK THROUGH THE FACEBOOK FILTER: Facebook Algorithm best practices 1. Timing: • 50% of those who will see your post were on Facebook within the 30 min of when you posted your post. • Facebook posts receive 50% of their likes in the first 1 hour and 20 minutes • Most FB posts expire after 5 hours. 2. Facebook like button on your site 3. Last actor – most recent 50 interactions get more weight
42. HOW TO BREAK THROUGH THE TWITTER FILTER Connections & Timing Matter: For a person to see your tweet, they: 1. Need to follow you. 2. Need to be on Twitter within 18 minutes of you publishing the Tweet. 3. Need to have interacted with your tweets in the past. 4. Your tweet must contextually be something that will drive engagement, or has been interesting to that person in the past.
43. BREAKING THROUGH THE YOUTUBE FILTER 1. YouTube SEO best practices 2. Google+ interactions 3. YouTube community management
44. WHERE ARE WE HEADED? THE FUTURE OF PERSONALIZATION
45. THE FUTURE OF GOOGLE SEARCH: “With your permission, you give us more information about you, about your friends, and we can improve the quality of our searches….we don’t need you to type at all. We know who you are We know where you’ve been We can more or less know what you’re thinking.” - Eric Schmidt, Google
46. THE END OF SEARCH AS WE KNOW IT Matt Bush, Google’s head of performance. The way of doing things yesterday is gone. We need to be looking constantly at data, analytics, and other things to ensure we stay ahead of constantly connected consumers. We now have 570m entities in Knowledge Graph. We are now taking it to the next level – into the area of anticipation.
47. GOOGLE NOW – SERVING YOU INFO WITHOUT SEARCHING
48. GOOGLE ASKS: WHAT ARE YOU THINKING NOW? http://backslidden.net/wp-content/uploads/2011/06/Dyson86237596.jpg Google asks: “What did you want to know recently?” “ It will know at a semantically deep level what you’re interested in, not just the topic... [ but] the specifi c questions and concerns you have. I envision some years from now that the majority of search queries will be answered without you actually speaking. ” – Ray Kurzweil
49. PARTING THOUGHTS…. Know Your Customer • Use their language • Have them test your message • Create search personas and measure against assumptions • Provide online content that is quality and shareable • Build social communities around your brand – especially utilizing Google+ • Store and leverage influencer data • Coordinate all messaging including paid to create a “surround sound” • Leverage social and web data to refine personas and outreach
50. ADDITIONAL LEARNING RESOURCES • How Search Works: http://www.youtube.com/watch?v=BNHR6IQJGZs • Google: How Search Works: http://www.google.com/insidesearch/howsearchworks/thestory/ • SEOMoz’s Beginner guide: http://guides.seomoz.org/beginners-guide-to-search-engine- optimization • Google’s SEO Guide for webmasters: http://www.google.com/webmasters/docs/search-engine-optimization- starter-guide.pdf
51. ADDITIONAL LEARNING RESOURCES • SEO for Bing: http://www.bing.com/blogs/site_blogs/b/webmaster/archive/2009/09/03/ search-engine-optimization-for-bing.aspx • More about Google Human Raters http://searchengineland.com/interview-google-search-quality-rater- 108702 • Staying up to date with online marketing: www.SearchEngineLand.com and www.MarketingLand.com
52. OPTIMIZING CONTENT
53. Put keyword in: • Account name. • Account description. • File title, description, and tags. Tips: • Documents should have an engaging cover page. • Connect account to Facebook. • Subscribe to other people. Comment on their documents. • Link in the document description to your site. OPTIMIZING SCRIBD.COM
54. OPTIMIZING GOOGLE+ 1. Make sure that your first line encourages people to read more. 2. Create urgent attention grabbing headline 3. Make sure you’re following Google’s SEO best practices for quality copy. This also means that Google+ post will (and should) be longer than Facebook posts. Google+ posts can rank in search an independent web pages.
55. OPTIMIZING TWITTER 1. Use keywords or popular hashtags: • Tweets with hashtags receive 2x more engagement than those without hashtags, except or those with more than 2 hashtags – then engagement shows a 17% drop 2. Tweets that contain less than 100 characters receive 17% higher engagement than longer Tweets 3. Tweets that ask followers to retweet receive 2x higher retweet rates than those that do not. 4. Tweets that have links receive 86% higher retweets than tweets without links
56. OPTIMIZING TWITTER 5. You should include in your messaging an ask for the reader to do something. Twitter has created a list of the most successful calls to action. 6. Tweets with images have 2 x higher engagement. 7. Georgia Tech ran an academic study related to which handles were most successful in gaining followers
57. OPTIMIZING YOUTUBE Put keyword in: • The audio of the video (YouTube will translate that into optimized text). • The video page – title of video, description, “tags.” • Use YouTube annotate. It allows you to add links, comments, etc. Tips: • Create playlists by grouping like videos together. • Encourage comments. • Build internal links (links from other video pages on the channel) and external links (links from other YouTube videos or outside websites) to the video.
58. • Jakob Neilson’s www.useit.com/alertbox/writing-numbers.html • Information Foraging: Why Google Makes People Leave your Site Faster. • First two words for links. • How little do users’ read? • http://www.nngroup.com/articles/how-users-read-on-the-web/ • Redish, Janice. 2007. Letting Go of the Words: Writing Web Content that Works. Morgan Kaufman Publishers. • How to write compelling online content that ranks well in search engines. • http://watier.org/katherine/best-practices-social-media-content- based-research/ REFERENCES
59. Katherine Watier Ong Ketchum 202-729-8355 @kwatier www.linkedin.com/in/katherinewatier gplus.to/katherinewatier Katherine.watier@ketchum.com Catherine Douglas Ketchum 646-935-4378 Catherine.Douglas@ketchum.com www.ketchum.com CONTACT