Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.

Google partners masterclass Melbourne presentation

Training from Google and Cardinal Path in Melbourne - October 2017

  • Inicia sesión para ver los comentarios

Google partners masterclass Melbourne presentation

  1. 1. Welcome to Masterclass 2017
  2. 2. Every month, in Australia & New Zealand, Google drives tens of millions of direct connections between businesses & their customers
  3. 3. Snowy’s
  4. 4. Many reasons to celebrate our great 2017 together
  5. 5. Today's agenda and logistics 12:00PM Registration, lunch and demos 12:45PM Welcome Rebecca Mason & Jean Magalhaes 12:55PM From a mobile-first, to an AI-first world Aisling Finch 1:20PM Optimisation and Automation Dave Booth 2:30PM Break and demos 3:00PM Measurement and Attribution Dave Booth 4:05PM Agency of the Future Martin Curtis 5:00PM Day Recap and closing remarks Rebecca Mason & Jean Magalhaes 5:10PM Networking, drinks and demos
  6. 6. #Masterclass2017
  7. 7. Marketing Director Google AU/NZ Aisling Finch From a mobile-first, to an AI-first world
  8. 8. Why AI and ML in Masterclass?
  9. 9. Definition AI (artificial intelligence) - science of making things smart, to a point they can perform sophisticated tasks, such as play Go, or recognise the content in a given photo. ML (Machine Learning) - is a technique to develop AI. ML is all about learning from examples, and getting computers to do something without being explicitly programmed to do so 01
  10. 10. 01 Why AI-first? 02 Meet the Assistant
  11. 11. 01 Why AI-first?
  12. 12. Helping us with everyday problems 01
  13. 13. Machine Learning is already improving many of our products Search Search ranking Speech recognition Play App recommendations Game developer experience Gmail Smart reply Spam classification Drive Intelligence in Apps Chrome Search by image Maps Parsing local search Translate Text, graphic and speech translation Cardboard Smart stitching Photos Photos search Assistant Smart connections across products YouTube Video recommendations Better thumbnails Android Keyboard & speech input 01
  14. 14. …and solving old problems 01
  15. 15. Data Centres energy efficiency 01
  16. 16. 01
  17. 17. 01
  18. 18. 01
  19. 19. Breast Cancer diagnosis Biopsy Slide Ground Truth (from pathologist) Model Prediction 01
  20. 20. 01
  21. 21. 01 02 Meet the Assistant Why AI-first?
  22. 22. 02 Meet the Assistant
  23. 23. 02
  24. 24. Assistant: connection between features across our products 02
  25. 25. Google’s speech recognition technology now has a 4.9% word error rate* July 2016 Oct 2016 Dec 2016 Current 8.5% 6.8% 6.1% 4.9% Source: Emil Protalinski, @EPRO, May 17, 2017 *US English only 02
  26. 26. 02
  27. 27. 02
  28. 28. We’re living in a world of moments 02
  29. 29. Thank you
  30. 30. Do you have a question about today’s content? Participate in the live Q&A using Sli.do Step 1: Visit https://www.sli.do/ Step 2: Enter event code #Masterclass17 Step 3: Submit your questions / upvote existing ones Three questions will be answered at the end of each presentation
  31. 31. Founding and Senior Partner Cardinal Path Dave Booth Optimisation & Automation
  32. 32. Aligning expectations
  33. 33. Things that will kill you while swimming/ surfing/ otherwise in the water Things that will kill you P. Sherman 42 Wallaby Way, Sydney, NSW
  34. 34. AI, ML, DL, GDTLAs
  35. 35. IAAI
  36. 36. The heavy-lifting task of optimising for performance Are you making the best use of your time? Are you driving the best results possible? Are you operating in the most profitable way?
  37. 37. The two elements of optimisation we’ll cover in this section BiddingCreatives 0201
  38. 38. Creatives 01
  39. 39. The art of showing generic ads to hyper targeted audiences “ 01
  40. 40. of consumers who own multiple devices switch between screens to complete a task. 90% more likely to be influenced by targeted digital ads. 2x sources consulted before making a purchase. 10+ Consumer journey is more complex than ever Source: https://www.thinkwithgoogle.com/research-studies/mobile-path-to-purchase-5-key-findings.html | https://www.thinkwithgoogle.com/articles/shift-to- constant-connectivity.html | https://www.sociomantic.com/wp-content/uploads/2013/11/What-Consumers-Want-Infographic-Sociomantic-Labs-EN-global.png 01
  41. 41. Consumers want information that is relevant to their moment based on their: Time Device Location Interest Intent Tailoring ads at scale 01
  42. 42. 01 Optimised Ad Rotation Tailored messages
  43. 43. 01 Optimised Ad Rotation Creative success Tailored messages
  44. 44. 01 Optimised Ad Rotation
  45. 45. Different people & contexts ask for different ads 01 3.5B searches happen on Google each day Ad 3 Ad 2 Ad Group 1 Ad 1 Device Time Geo Device Time Geo Device Time Geo
  46. 46. Ad groups with 3 or more high-quality ads may receive up to 15% more clicks or conversions* vs. ad groups with only 1 or 2 ads. 01 Show the right message to the right person Ad 3Ad 1 Ad 2 *Source: Google Internal Data. When you rotate your ads to optimize for clicks or conversions. This will vary by client, business type, among other factors, and Google cannot guarantee this or any uplift.
  47. 47. 01 A practical example *Source: Google Internal Data. When you rotate your ads to optimize for clicks or conversions. This will vary by client, business type, among other factors, and Google cannot guarantee this or any uplift. Ad 3 (It’s probably too late) Ad 1 Forgetful packing incident? Ad 2 Forgetful morning dresser? Ad 3
  48. 48. Optimisation Ad Rotation 01 Optimise ad rotation to show your most relevant ad *Source: Google Internal Data. This will vary by client and business type, among other factors and Google cannot guarantee this or any uplift. +5% more clicks & conversions* Improve quality & traffic Both optimise settings positively impact quality score & CTR Ads using either optimise setting tend to accrue more conversions Increase conversions Simplify campaign management Let AdWords automatically serve the most relevant ad with no additional effort required
  49. 49. 01 Tailored messages
  50. 50. 01 Dynamic Search Ads (old but gold)
  51. 51. 01 of searches are unique 15% How can we optimise and plan for things that haven’t happened yet? Use Dynamic Search Ads to capture relevant queries
  52. 52. 01
  53. 53. 01
  54. 54. Compared with their keyword performance, for the median advertiser, DSA targeting generates... 01 CPCs CPAs 35% lower Clicks CTRs ...and on average 86% of Dynamic Search Ads traffic is incremental 30% lower 30% higher 30% more
  55. 55. Best practices reminder 01 1. Website structure 2. Exclusions 3. Monitor the machine!
  56. 56. 01 Ad Customisers (new feature)
  57. 57. 01 2011
  58. 58. 01 Feed based creatives for everyone
  59. 59. Customise your ads at scale to every context: keyword, time, device & user location. 01 Ad customisers
  60. 60. 01
  61. 61. Tailor your ad to the keyword, ad group or campaign 01 Ad Buy some dinosaur meat. example.com Starting at $49. Won’t spoil for another 325 years! Ad Buy some drop bear meat. example.com Starting at $1.99. Won’t spoil for another 15 days! Eat a 
 dinosaur Drop bears Example Ad Buy some MYSTICAL_CREATURE_NAME meat. example.com Starting at COST. Won’t spoil for another COUNTDOWN!
  62. 62. Feed example for keyword ad customisation 01 Note: 
 You can also use “Target Keyword text” and “Target Keyword Match type”, instead of “Target Keyword”
  63. 63. Tailor your ad to the time of day/ device Ad Got the Monday am commute blues? example.com Eat some unicorn and sparkle. Free shipping on orders above $65. Ad Host a hoop snake BBQ example.com No plans for the weekend? BBQ some hoop snake! 20% off BBQ1. Ad Chupacabra breakfast tacos? example.com You bet! 10% off fresh chupacabra meat. Just add guac & salsa. Friday Laptop Monday Mobile Any day, AM Any 01
  64. 64. Feed example for device / time ad customisation 01
  65. 65. 01 Approach ResultsGoals - Increased overall CTR by 32% - Grew CTR in second-highest volume-producing ad group 
 by 113% - Successfully created and tested ads at the customer level - Used Ad Customisers to create timely and accurate ads - Tested ad copy with specific audience segments - Engage customers with relevant, real-time messages throughout the purchase journey - Efficiently update ad creative in different ad segments
  66. 66. 01
  67. 67. Count down to an event 01 Ad Valentine’s day romantic gifts example.com/valentine Fun activities for couples on sale now! Hurry! Deal expires in COUNTDOWN. Example Ad Valentine’s day romantic gifts example.com/valentine Fun activities for couples on sale now! Hurry! Deal expires in 5 Days. Feb 9 3:00pm Feb 14 9:00am Feb 9 5:40pm Ad Valentine’s day romantic gifts example.com/valentine Fun activities for couples on sale now! Hurry! Deal expires in 9 Hours. Ad Valentine’s day romantic gifts example.com/valentine Fun activities for couples on sale now! Hurry! Deal expires in 20 minutes. Ad Buy some dinosaur meat. example.com Starting at $49. Won’t spoil for another 325 years!
  68. 68. Countdown widget in AdWords 01
  69. 69. 01 Approach ResultsGoals - 8.7% increase in conversion rate at the end of flash sale - 3% increase in CTR on their ad text - 8.3% increase in customer spending per site visit - Implemented the countdown function with Google’s countdown widget - Dynamically showed customers the days, hours, or minutes left until end of sale - Drive quality traffic to website - Create urgency to promote flash sale - Encourage customers to convert quickly
  70. 70. Let’s look at data! 01
  71. 71. Let’s look at data! 01
  72. 72. 01
  73. 73. Customise ads to the user location of interest 01 Location of interest detected Texas, ‘Merica Ad Drink without a worry in DESTINATION example.com PRODUCT for DESTINATION people, starting at PRICE!
  74. 74. Customise ads to the user location of interest 01 Ad Drink without a worry in Texas example.com Drinking holsters for Texas people, starting at $10! “drinking in Texas” Ad Drink without a worry in California example.com Sunscreen flasks for California people, starting at $5! “help me. just please help me.” Location of interest detected Texas, ‘Merica Location of interest detected California, USA Ad Drink without a worry in DESTINATION example.com PRODUCT for DESTINATION people, starting at PRICE!
  75. 75. Feed example for geo ad customisation 01
  76. 76. Schedule your feed uploads 01
  77. 77. Tips / best practices 01 1. You cannot pause customisers But you can pause ads that use them or remove associated data 2. If you remove a data set You can’t “re-start” by uploading a new data set with the same name Better to edit your data to keep customisers running 3. Get creative!
  78. 78. BiddingCreatives 0201
  79. 79. Bidding 02
  80. 80. Automation opens up big possibilities 02Source: *BCG Study ‘Cutting Complexity, Adding Value’, 2013, link Monday Tuesday Wednesday Thursday Friday is strategic 20%80% of time spent is dealing with manual, less strategic tasks
  81. 81. 02 (not that
 many) conversions Millions of variables x = The complexity of the task at hand… Keywords Topics In-market Affinities x All your clients! Time Hour Day Week Month Weekday Location Radius Post code DMA Region Country Device Mobile Desktop Tablet Audience Page visitors Cart abandoners Loyal shoppers
  82. 82. Who’s your audience anyway? 02
  83. 83. Who’s your audience anyway? 02
  84. 84. AdWords Smart Bidding: Auction-powered technology 02Source: *BCG Study ‘Cutting Complexity, Adding Value’, 2013, link Advanced machine learning that drives informed, accurate bidding decisions Unparalleled optimisation frequency and precision that adjusts bids for every auction Rich set of auction-time signals used to capture contextual relevance for every unique search
  85. 85. 010101010101010101010101010101010101010101010101010101010 010101010101010101010101010101010101010 0101010101010101010101010101010101 010101010101010101010101010 101010101010101010101 010101010101010 101010101 010 02Source: *BCG Study ‘Cutting Complexity, Adding Value’, 2013, link Advanced machine learning Keyword Many search queries Across auctions and search contexts Unique insight into every auction, computing data at the search query level W W M Z Q
  86. 86. Rich set of auction-time signals 02 Location Remarketing
 list Every user search is different and bids for each query should reflect the unique contextual signals present at auction-time. Bid Combinations between 
 2 or more 
 signals Search 
 partner Actual 
 query Language AppOS Browser Ad 
 creative Noon 
 EST Smartphone Signals available 
 with bid adjustments Exclusive signals for AdWords Smart Bidding
  87. 87. 02 Example, on the keyword buy new toaster| Example of signals our algorithm will optimise against: Language settings Location Time Browser Operating system CVR 4.0%CVR 2.0% vs. Finding the right keyword bid is hard. Keyword bids are not right for all audiences. Smart Bidding use case
  88. 88. Smart Bidding 02 Flexibility Efficiency Transparency Testing Attribution 01 02 03 04 05
  89. 89. 02 Revenue Conversions/sales Website click Visibility Campaign Goal Recommended Strategy Target outranking share Maximise clicks Target CPA / enhanced CPC (semi-auto) Target ROAS Powered by AdWords Smart Bidding 1. Bidding strategies aligned to marketing goals
  90. 90. 02 Revenue Conversions/sales Website click Visibility Campaign Goal Recommended Strategy Target outranking share Maximise clicks Target CPA / enhanced CPC (semi-auto) Target ROAS Powered by AdWords Smart Bidding 1. Bidding strategies aligned to marketing goals
  91. 91. 02 2. Simulations with great data quality Leverage insights from millions of bids set every second across AdWords Determine the most profitable CPA target and apply the change directly in the simulator tool
  92. 92. 02 3. Unique transparency Monitor a bid strategy performance and understand its current status Understand the strategy status and potential limitations, then take action Monitor the performance of flexible bid strategies key KPIs and watch how it improves over time
  93. 93. 02 4. Seamless high end experiment platform Run well-executed A/B tests to test changes before rolling them out Assess performance Determine the portion of traffic that you want to go through the experiment (360)
  94. 94. 02 5. Go further: bid beyond the last click Pick up a non-last click attribution model Last click First click Linear Time decayPosition-based Rules-based
 models Data-driven
 model great tech gifts highly rated tablets nexus 9 tablet highly rated tablets nexus 9 tablet 2.73% Conversion Rate 2.41% Conversion Rate = 0.32%
  95. 95. 02 IAAI
  96. 96. Thank you
  97. 97. Q&A
  98. 98. Break
  99. 99. #Masterclass2017
  100. 100. Founding and Senior Partner Cardinal Path Dave Booth Measurement & Attribution
  101. 101. Do you have a question about today’s content? Participate in the live Q&A using Sli.do Step 1: Visit https://www.sli.do/ Step 2: Enter event code #Masterclass17 Step 3: Submit your questions / upvote existing ones Three questions will be answered at the end of each presentation
  102. 102. AttributionMeasurement 0201
  103. 103. Measurement 01
  104. 104. 01
  105. 105. 01 On sale! Only 45 shillings Now through Friday’s execution iAnvil 8
  106. 106. Technology has changed everything. People now have a huge selection of devices, channels, and content to choose from, across multiple platforms. This creates a lot of fragmented data which can be hard to track and action for advertisers DisplayOrganic Search BrandDisplay network 01
  107. 107. 01
  108. 108. 01 Legacy, linear paths to purchase were relatively simple to measure. Today’s paths are not. New consumer behaviour broke measurement 01
  109. 109. Rethink measurement 01
  110. 110. Rethink measurement Intra channel Cross channel 01
  111. 111. 90% of people switch between screens to complete tasks 30% of all customer journeys involve more than one channel of interaction Source: Google/Purchased Digital Diary: How Consumers Solve Their Needs in the Moment, May 2016 01
  112. 112. Source: Google/Purchased Digital Diary: How Consumers Solve Their Needs in the Moment, May 2016 45% increase in local shopping queries 01
  113. 113. 01
  114. 114. 01
  115. 115. How much data are you trying to manage? 01
  116. 116. AttributionMeasurement 0201
  117. 117. Attribution 02
  118. 118. 02 Let’s play a game!
  119. 119. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models Identify objectives
  120. 120. Day 1 Day 2 Day 3 02 A customer’s journey
  121. 121. Day 1 Day 2 Day 3 02 A customer’s journey Email Email Desktop Desktop Mobile web In-store visitMobile app
  122. 122. 02 Day 1 Day 2 Day 3 A customer’s journey Email Email Desktop Desktop Mobile web In-store visitMobile app User ID, CD Login cookie Login cookie Login cookie CRM ID Login deviceID Login cookie Login cookieLogin cookie User ID, CD
  123. 123. 02 Day 1 Day 2 Day 3 A customer’s journey Email Email Desktop Desktop Mobile Web In-store visitMobile App User ID, CD Login cookie Login cookie Login cookie CRM ID Login deviceID Login cookie Login cookieLogin cookie User ID, CD
  124. 124. 02
  125. 125. 02 Different people Different needs Different stages Different experiences across at with
  126. 126. 02 Different people Different needs Different stages Different experiences across at with Segments Business goals Customer journey Channels & devices
  127. 127. 02 Metrics & dimensions KPI’s Business goals Different needs Customer journey Different stages Segments Different people Channels&devices Differentexperiences
  128. 128. 02 Business goals Different needs Customer journey Different stages Segments Different people Channels&devices Differentexperiences Segment: Journey: Goal: Channel/device: Existing customer Retention/growth Easily find what they need Mobile website
  129. 129. 02 Business goal: Get visitors to content/features quickly and efficiently Customer journey Channels/devices Websitefrommobiledevice Segment Existingcustomer Attract Engage Convert Retain/grow Advocate Pageviews per visit Pages/Session, segment = mobile Google Analytics Time on site Avg. Session Duration, segment = mobile Google Analytics Bounce rate Bounce rate, segment = mobile Google Analytics Site search usage Sessions with search, segment = mobile Google Analytics Unsuccessful search rate % Search exits, segment = mobile Google Analytics Feature completion (various) (Configured event/goal count/conversion rate) segment = mobile Google Analytics Social activity (Tweets/Posts, negative sentiment (Social monitoring) via BigQuery, ML API Inbound customer service call (Call tracking click to call event)(Inbound calls/sessions) 
 segment = mobile Google Analytics 360, 
 (Call tracking solution) Content engagement score (Custom calculated metric) applied to (key pages), 
 segment = mobile Google Analytics 360, CMS via BigQuery Content velocity (Custom calculated metric) applied to (key pages), 
 segment = mobile Google Analytics 360 KPIs Metric(s) required Data source
  130. 130. 02 Visit 1: Visit 2: Engagement Score / Visit = Custom calculated metrics Scroll 25% Scroll 3 50% Scroll 4 75% Scroll 5 100% Scroll 6 Click Page load / content interaction 1 Tool start 10 Tool complete 10 Login 10 Special offer signup 15 Branch search 5 Branch search exit Download iOS / Android app download 2 PDF content download 1 Whitepaper download 1 Social Share 1 Post 2 Chat 1 Email 1 Comment 1 Video Video start Video 25% Video 50% Video 75% Video completion Category Component Weight 2 + 15 2 2 15 20 56+ + + + = On Page Interaction Scroll 50% On Page Interaction On Page Interaction Scroll 50% Video Start Scroll 50% On Page Interaction Download Event On Page Interaction Scroll 100% Video Start Video Start 50% 15 + 2 20 2 25 20 99+ + + + =15+ 56 99+ 77.5= 2 Visits
  131. 131. 02 Business goals Different needs Customer journey Different stages Segments Different people Channels&devices Differentexperiences
  132. 132. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models Identify objectives
  133. 133. 02 Implementation steps Knowledge transfer Production launch QA / Tag validation ConfigurationImplementation Solution design documentation Business requirements/ KPI framework
  134. 134. 02 Implement against the framework Pay special attention to: 1. Conversion actions 2. Integrations 3. Campaign tracking Business requirements/ KPI framework Solution design documentation Implementation
  135. 135. 02 Don’t forget your integrations!
  136. 136. 02 Don’t forget your integrations! Enable a holistic view: - Single ID across properties - Connected to sales data (incl. offline) - Analysis done with on/offline attributes EDW CRMSingle ID WA Activation
  137. 137. Analysis across the 360 degree view Allows us to: - Create segments from offline / additional data -Analyse against offline / additional conversion events -Leverage additional attributes across data activation platforms 02
  138. 138. Analyse campaign impact across full funnel 02
  139. 139. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models Identify objectives
  140. 140. 02 Default: last-touch attribution
  141. 141. 02
  142. 142. 02 Default: last-touch attribution C1 C2 C3 C4 C5
  143. 143. 02 Other common models C C2 C C4 C5 C C2 C C4 C5 C C2 C C4 C5 C C2 C C4 C5
  144. 144. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models Identify objectives
  145. 145. 02 Attributing value with additional models
  146. 146. 02 Attributing value with additional models
  147. 147. 02 Attributing value with additional models
  148. 148. 02 Attributing value with additional models
  149. 149. 02 Attributing value with additional models
  150. 150. 02 Attributing value with additional models
  151. 151. 02 Attributing value with additional models
  152. 152. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models Identify objectives
  153. 153. 02 Algorithmic/modeled/data-driven attribution
  154. 154. 02 Algorithmic/modeled/data-driven attribution
  155. 155. 02 Algorithmic/modeled/data-driven attribution
  156. 156. 02 Algorithmic/modeled/data-driven attribution
  157. 157. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Identify objectives Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models
  158. 158. 02 Digital multi-touch attribution (MTA) Cookie level Website data Digital media plans Data partnerships 1st party data KPI(s) Customer level
  159. 159. 02 Digital multi-touch attribution (MTA) Measure the lift when a channel is present on the path to purchase Your baseline level of measurement is always a user and their path to purchase 3.0% 2.8% 2.5%
  160. 160. 02 Google Attribution & Attribution 360
  161. 161. 02 One solution, available in two tiers Common cross device & data-driven methodology Google Attribution 360 Full, enterprise-level cross-channel attribution 
 Custom dimensions & metrics for flexibility Optimisation & TV modeling Google Attribution Self-service with simplified options & controls 
 Makes it easy to get started with cross- channel, multi-channel attribution
  162. 162. 02 Foundations of Google’s attribution solutions Insights & actionScience & analytics Data collection & management Considers all tracked touchpoints Google device graph Integrations with buying tools TV attribution Predictive optimisation Reporting 
 & insights Data exports
 and API Programmatic connectors
  163. 163. 02Source: © Google Inc. 2016. All rights reserved. Broad reach across devices The deterministic cross-device graph is powered by Google’s largest web properties, may with >1bn 30-day- active logged in users. Cross-device attribution, powered by Google’s device graph Consumer friendly approach By allowing end-user opt-out our solutions is extremely privacy friendly, and protects the end user. More accurate measurement Our algorithm stitches the path- fragments into complete per-user paths, enabling the model to understand the full customer journey.
  164. 164. 02 Google Attribution & Attribution 360
  165. 165. 02 75% of users switch devices between tasks
  166. 166. DDA has already driven better performance for 
 thousands of advertisers 02 Data-driven attribution (DDA) drives more conversions at the same cost-per-conversion Moving beyond last-click on Search alone delivers 5% more conversions on average, at a similar cost-per- conversion.
  167. 167. 02
  168. 168. 02
  169. 169. 02 Attribution 360: television TV airings data Digital web / conversion data Google Search data Attribution360
  170. 170. 02 Attribution 360: timeline A360 product phase Closed alpha Beta General availability 2018+2017 General availability Through 1H 2018
  171. 171. 02 Prepare for Attribution & Attribution 360 Google Analytics - Enable Advertising Reporting features - Set up Goals and Ecommerce! - Break down channel groupings - Make sure integrations are in place - Clean up your campaign (UTM) tagging - Create separate, dedicated View (no User ID, do NOT strip URL params) Google AdWords - Ensure Auto-tagging is on - Ensure AdWords is linked for cost data import - Review naming conventions / consistency - Link AdWords with Analytics Views DoubleClick (DCM, DS) - Ensure Auto-tagging is on - Add display activity service OR ensure impression & click tracking is in place - If using Floodlight to track conversions, ensure configured & clean - Ensure accurate media cost is there for all campaigns - Review naming conventions / consistency - Link DoubleClick with Analytics Views
  172. 172. 02 Prepare for Attribution & Attribution 360
  173. 173. 02 A path to attribution Digital multi-touch attribution (MTA) solution Algorithmic/data-driven approaches Exploring additional models Standard reporting Implement tracking Identify objectives Conversions integrations campaigns Typically “last click” Econometric (MMM/ESM) Econometric/model based (response curves, channel contribution, halo effects, etc.) Tag, cookie, device- graph based Data driven attribution 
 (DDA) Workbench Marketing channels, campaign stacking Multi-channel funnels (MCF) model explorer, custom models
  174. 174. 02 Econometric approaches - MMM Annual contribution to sales ROI by media channel Campaign direct and halo effects Media response curves CY 2013 CY 2014
  175. 175. 02 Econometric approaches - ESM
  176. 176. 02 Econometric approaches - ESM
  177. 177. Thank you
  178. 178. Q&A
  179. 179. Time to recognise your partnership
  180. 180. Head of Performance Agencies Google AU/NZ Martin Curtis Agency of the Future
  181. 181. Quick exercise Your name What you’re hoping to get out of today’s session One fun fact that no one would know Your role Your company
  182. 182. “The next big step will be for the very concept of the device to fade away. Over time, the computer will be an intelligent assistant helping you through your day. We will move from mobile first to an AI first world. - Sundar Pichai, CEO, Google
  183. 183. Google has a great track record of predicting the future
  184. 184. 01 Total Digital Immersion 02 AI-First World
  185. 185. 01 Total Digital Immersion
  186. 186. Digital = Everything Move away from being the marketing specialist to the marketing expert 01
  187. 187. 01 CMO’s grappling with even more complexity outside 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 VCR TV, radio, OOH, newspaper, magazine Cable AOL The period of unbundling of media agencies from creative agencies DBS satellite Yahoo Google DVR Broadband HD DVD & 
 Blu-Ray Satellite radio HDTV Time-shifted viewing Facebook 7th gen game console Twitter YouTube Netflix streaming Hulu+ Smart TV Tablet Instagram Streaming radio 8th gen game console Chromecast Amazon Prime instant video Roku Smartphone Apple TV VOD Snapchat
  188. 188. People just expect faster, relevant and frictionless experiences, whenever and wherever. 01
  189. 189. 26 Source: A Guide to the Internet of Things, IDC, Intel, United Nations, 2016. 01
  190. 190. We get it, it’s complicated 01
  191. 191. But only 17% of agencies and companies have moved off last click attribution 01
  192. 192. Where we should be as an industry Where we are Advanced attribution strategies Do you want to move off last click? Last click First click Linear Time decayPosition-based Data management platform integrations 01
  193. 193. And it’s only getting more complex 01
  194. 194. The agencies of the future are moving now to measure and manage the entire digital journey 01
  195. 195. 01 Total Digital Immersion 02 AI-First World
  196. 196. 02 AI-First World
  197. 197. A major milestone in Artificial Intelligence 02
  198. 198. Programming Learning 02
  199. 199. 01 We need AI to help us manage this complex journey
  200. 200. of your time will be saved to be reinvested back into your business & yourself 50% 02
  201. 201. This is your opportunity to move away from optimisers back to creative thinkers 02
  202. 202. AI is exponential in nature so you will get left behind Don’t “wait & see” 02
  203. 203. Inflection 02
  204. 204. Day 1 of Adweek NYC We know change is coming 02 - Is the agency of the future still an agency - Leadership and the agency of the future - Agency of the future, becoming the CMOs secret weapon - Creativity to the power of consulting - Advertising needs a rebrand - Yes you should start that new agency you have always dreamed of - Self-disrupt: Creating the next agency model
  205. 205. 02 The largest digital agency is now Source: Ad Age Agency Report 2016 USRevenue $0 $375,000,000 $750,000,000 $1,125,000,000 $1,500,000,000 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
  206. 206. So what can you do? 02
  207. 207. What would it look like? If you were to start your agency today, from scratch If you were to develop your skill sets, from scratch What would they look like? 02
  208. 208. Build capabilities to solve wicked business problems 02 Build skills to solve wicked business problems
  209. 209. Audience data Machine learning Creative thinking & strategy 02 + +
  210. 210. 02 Develop data capabilities Embrace Machine Learning Upskill talent
  211. 211. 02 Develop data capabilities Measurement, attribution, data and feeds
  212. 212. Develop measurement & attribution Develop data & feed management capabilities 02 Where do you start?
  213. 213. 02 Business impact Diversification opportunity Improved client results New revenue stream
  214. 214. 02 Google solutions AdvancedBasic Intermediate Data driven attribution Non-last click Feed & data integration Cross channel attribution Post impression attribution Advanced attribution Data management solution Coming soon
  215. 215. 02 Embrace Machine Learning Automation and Machine Learning
  216. 216. Automate everything 02 Where do you start?
  217. 217. 02 Business impact Saved time Improved margins from efficiencies Time can be invested into creative thinking and strategy to grow revenue
  218. 218. 02 Google solutions AdvancedBasic Intermediate Auto Ad rotation Auto bidding Dynamic Search Ads IF Function Automated Feeds Automated Reporting Auto bidding/creative Audience Integration Automate cross channel
  219. 219. 02 Upskill talent Creative thinking and client service
  220. 220. Who owns creativity & strategy?
  221. 221. It needs to be everyone’s responsibility
  222. 222. WALDO?
  223. 223. TOM MATT MELISSA WALDO 8 14 22 ? = = = =
  224. 224. TOM MATT MELISSA W A L D O 8 14 22 13 = = = =
  225. 225. 02 Change your environment
  226. 226. 02 Keep track of micro creativity
  227. 227. 02 The 10% Rule Dedicate 10% of client’s budget to innovation
  228. 228. 02 Where do you start? Retain the best talent Hire great people Develop your people; creative thinking and strategy skills
  229. 229. 02 What can you do? Mission, transparency, voice Think broad & know what you want Focus on continued learning
  230. 230. 02 Google Partners initiatives ProductDiversity Strategy Lead by WomenWill Masterclass Specialist Challenge Academy Bootcamp (in-person) Academy for Ads (on-demand) Business Coaching Sales Masterclass Partners Connect (co-pitch events)
  231. 231. 02 What the agency of the future looks like Step 1 Step 2 Step 3 Step 4 Develop creative thinking & strategy expertise Automate everything Build measurement, data & feed capabilities Diversify capabilities - video, display, feed management, programmatic Step 5 Consult on wicked business problems
  232. 232. How else can you prepare for these changes? 02
  233. 233. 02 Embrace change Be curious Live life as a student
  234. 234. Thank you
  235. 235. Q&A
  236. 236. Recap
  237. 237. Upcoming initiatives you don’t want to miss Academy Bootcamp: Mobile Sites & Analytics Specialist Challenge 2018
  238. 238. M-Site Sydney
 Tuesday 5 December Analytics Sydney
 Wednesday 6 December Academy Register at g.co/AcademyAUNZ
  239. 239. Do you want to hear from us? bec@example.com
  240. 240. And the #Masterclass2017 best photo winner is… #
  241. 241. Thank you

×