Analytics Academy 2015 Presentation Slides
Download as PPTX, PDF1 like767 views
This document discusses how analytics can inform digital strategies and operations at NPR. It begins by outlining some key goals of NPR's digital and social media presence, such as informing the public and tracking changing audience behaviors. It then discusses how NPR analyzes various metrics like website traffic, mobile usage, social media engagement, and email effectiveness to understand what content and platforms resonate best with audiences. Insights from analytics are used to optimize strategies, like prioritizing mobile-friendly content or creating local Facebook stories that drive higher engagement. The document emphasizes using data to continuously learn and improve offerings, rather than just focusing on short-term optimizations. Context is also important to make data meaningful.
Analytics Academy 2015 Presentation Slides
- 1. WHY ARE WE TALKING ABOUT ANALYTICS TODAY?
- 2. Photo: Tobias Seiderer, Flickr
- 3. Photo: Scott Akerman, Flickr
- 7. INSIGHTS ACTIONS YOU GET WHAT YOU MEASURE
- 8. “I’m gonna write me a new minivan this afternoon!”
- 9. WE DO ALL THIS AGAINST A BACKDROP OF A WORLD OF DATA Max Hancock for the Noun Project
- 14. Thank you. Enjoy the day!
- 15. We Are What We Measure Steve Mulder
- 16. 16
- 17. 17
- 18. 18 What we measure becomes what we reward
- 19. 19 What we reward becomes what we do and what we are
- 20. 20 What are you measuring now?
- 21. 21 In the world of NPR… Broadcast radio listening Digital analytics Donations and membership Underwriting revenue Important things to measure, but do they tell the whole story of what you are trying to do and be? Are there things you’re doing less of simply because they’re not as easy to measure? We tend to measure what is familiar and easy to count
- 22. 22 Community impact More informed public, influence on culture, policy change, etc. Audience insights Listener attitudes, unmet needs, changing behaviors Digital Less familiar to many people in our organizations Constantly changing and hard to keep up with Overwhelming amount of data at our fingertips It’s harder to measure what is less quantifiable, newer, changing, or difficult to track
- 23. 23 Using data effectively is a lot about mindset
- 24. 24
- 25. 25
- 26. 26
- 27. 27 Trap #1 We measure the wrong thing
- 28. 28 Trap #2 We make decisions by blindly following data
- 29. 29
- 30. 30 Trap #3 We become optimizers instead of free thinkers
- 31. 31 Optimization via A/B or usability testing The danger of optimizing to a local maximum Local maximum
- 32. 32
- 33. 33 1. Data is the platform from which we leap
- 34. 34 For NPR, ratings and analytics data tell us about listening
- 35. 35 Third-party research validates trends in media fragmentation and digital adoption
- 36. 36 Channels proliferate, presenting both opportunity and challenge “Radio isn’t going away, it’s going everywhere.”
- 37. 37 Our own research and personas reveal changing consumer behaviors & opportunities
- 38. 38 It’s not about building things that users know to ask for
- 39. 39 New digital interfaces in familiar locations (connected car)
- 40. 40 Experiments with new listening experiences: NPR One Exploring experiences that meet needs users can’t yet articulate
- 41. 41 2. Data helps us know how high we’ve jumped and whether to keep going
- 42. 42 The Facebook experiment: Local station stories geo-targeted on NPR’s Facebook stream
- 43. 43 Local stories saw consistently higher engagement and grew local audience
- 44. 44 Pivot and double-down: We created a workflow tool so we can scale this offering
- 45. 45 and more about continuous learning 3. It’s less about achieving goals
- 46. 46 Yes, this sounds like Lean Startup thinking Assumption/hypoth esis Minimum Viable Product Analytics, research, testing
- 47. 47 Designing choice in the new NPR One app Original design: Give users more control/choices Hypothesis to test: Fewer immediate choices + simplicity = longer listening Results: Listening time up 8%
- 48. 48 Walking the narrow path
- 49. 49 Ingredients of an analytics culture
- 50. 50 Bring data to life, make it memorable and actionable
- 51. 51 Bring data to life to make it memorable and actionable
- 52. 52 Emphasize what matters most
- 53. 53 With great power comes great complexity. Don’t let analytics feel like this.
- 54. 54
- 55. 55 Thanks for the metric. Is that good or bad? Context is the engine of meaning
- 56. 56 Context can turn data into insight and action
- 57. 57 Context over time, against averages, or against custom cohorts
- 58. 58 Embed data in regular processes and habits
- 59. 59 Embed data in regular processes and habits
- 60. 60 Data without analysis and storytelling is like food without taste
- 62. 62
- 63. Putting Data to Work How Digital Analytics Informs Operations Analytics Academy March 19, 2015
- 64. Overview What is your business goal? How does digital communications support that goal? What are the goals of digital strategy? What can you learn about user behavior to determine if you are achieving that goal? How can you use that knowledge to improve your strategy?
- 67. WEB & SOCIAL Strategies for inbound traffic
- 68. Goals of the Website • Help prospective applicants discover the Harvard experience • Hear authentic current student voices • Understand the application process and successfully complete the application • Access information about application process and admission decision
- 69. Analyzing Web Traffic Google Analytics Monthly Report Metrics Report Interpret
- 70. Interpreting the Metrics • Seasonality - The Admissions cycle affects traffic to the site. By understanding that traffic we can plan for server capacity and uptime during our most critical time. Regular Decision Early Decision
- 71. Understand Traffic Sources • Traffic Acquisition - How does it change over time? January 1, 2014 January 1, 2015
- 72. Analytics-driven Content • Landing pages: identify where users are entering your site, other than the homepage • Highlight key messages • Make content easily accessible across the site
- 73. Understand Traffic Sources • Mobile Traffic - Mobile traffic has grown from 15% in 2014 to 20.5% of all traffic in one year on smartphones alone. On EA decision day, mobile traffic was nearly one quarter of all traffic.
- 74. Mobile Strategy • Optimize for mobile. Assume users will view on mobile first –Responsive design. All content must work on mobile– web, email, tour –Test early and often on many devices, browsers
- 75. Social Media & Google Analytics Acquisition > Social > Users Flow
- 76. Social Media Many sources, constant change • What can we measure? We can’t we measure? • Aggregating social data from various platforms • Measuring engagement at the post level to inform content strategy
- 78. Informing Social Media Content • Our audience are 16-18 years old. • We are monitoring usage to go where they are, rather than asking them to come to us.
- 79. EMAIL & OUR SEARCH LIST Strategies for outbound communication
- 80. Goals of Outreach • Search list are highest priority users– actively recruiting these prospective applicants • Targeted outreach focused on students from diverse socioeconomic and cultural backgrounds • Drive traffic to our website, engage with our students, start an application
- 81. Measuring Email Effectiveness • Engagement is measured in open rates, clickthroughs • Deliverability, bounces, quality of list • Use UTM codes to link back to website • Send through Slate, integrated with admissions database. More on that later!
- 82. Google URL Builder ● utm_source: Identify the advertiser, site, publication, etc. that is sending traffic to your property, for example: google, citysearch, newsletter4, billboard. ● utm_medium: The advertising or marketing medium, for example: cpc, banner, email newsletter. ● utm_campaign: The individual campaign name, slogan, promo code, etc. for a product. ● utm_term: Identify paid search keywords.
- 83. Screenshot of Email Metrics
- 84. Email Report
- 85. Email Content Optimize based on learning • 70% of emails opened on mobile. Design for mobile – Responsive design – Length of message – Placement of links • Know when to send based on audience – 3-5pm for our audience • Test subject lines – A/B testing
- 86. SLATE (DATABASE OR CRM) converting prospects to applicants
- 87. CRM & Admissions Database • Content strategy is based in goals for recruitment & yield. CRM tracks progress towards that goal. • Also makes online reading of applications possible • Implemented this year- opportunity to reevaluate strategies and better analyze our data!
- 88. Student’s Path in App Cycle • Applicant data is centralized throughout all stages of the process– as a prospect, inquiry, and applicant. Prospect Inquiry Applicant
- 89. Getting Specific with our Message Recruitment communications for targeted prospective students Low socio-economic status seniors Hispanic, African American, Asian American, Native American Low Application schools Low SES Juniors HFAI UMRP HCC First in the family to graduate from a four-year college First Gen
- 90. Measuring Digital Recruitment Track offline interactions and aggregate data in Slate • Interactions with current students • History of engagement is kept in the applicant record • Query and pull reports based on interactions
- 91. Measuring Digital Recruitment • Widgets, queries and reports demonstrate how recruitment leads to applications.
- 92. 15797 17173 18498 20132 2018 2019 Searched +8.71% Not Searched +8.83% Number of Applicants to the Class of 2018 and 2019
- 94. Thank you Amy Lavoie | aalavoie@fas.harvard.edu Victoria Marzilli | vmarzilli@fas.harvard.edu
- 95. Hit or miss? How to tell, and what to do next
- 97. What does ‘success’ mean? ● Metrics paired to business goals o Pageviews (reach) o Time on site (attention) o Email signups (loyalty, audience building) ● Anecdotal feedback o internal (local leadership, faculty) o external (media, influencers, awards)
- 98. What does ‘failure’ mean? ● Failed to meet expectations ● Elicited negative responses ● Opportunity cost
- 99. From the Harvard channels
- 100. From the Harvard channels
- 101. Campaign checklist Using reporting from start to finish 1. Decide what’s worth sharing 2. Build consistent tracking methodology 3. Share prepared content 4. Implement analytics collection process 5. Put analytics through feedback loop 6. Evaluate effectiveness, determine takeaways
- 102. Deciding what to share — and how to share it Editorial decisions *Editorial calendar *Trending topics (Twitter, FB, Iconosquare) Google trends Topsy Encore Alert and similar services Format Platform updates Mobile proliferation *Current research (Sharing Fast and Slow)
- 104. Choose content: Format Sharing Fast and Slow 104 POST PRIMED FOR SHARING/REACH POST PRIMED FOR INTEREST/CLICKS 200 clicks to story176 retweets, 13 new followers
- 105. Build and execute consistent reporting framework
- 106. Put analytics through feedback loop
- 107. Evaluate effectiveness, determine takeaways
- 108. Thank you! Any questions? Let’s talk! rebecca_wickel@harvard.edu
- 109. The art and science of effective social media in campaign messaging. Painting By Numbers
- 110. Usable Knowledge #ThroughEducation Launch/learning/measurement Launch/learning/measurement
- 111. Usable Knowledge #ThroughEducation Launch/learning/measurement Launch/learning/measurement CAMPAIGN PRIORITY (DEAN) CAMPAIGN ENGAGEMENT
- 112. Usable Knowledge: A History
- 113. Strategy
- 114. Strategy Validate and experiment for hypothesis: Does social deliver?
- 116. Strategy
- 117. Strategy
- 124. Social is like Moneyball
- 128. #ThroughEducation
- 130. National trending
- 131. #ThroughEducation
- 137. @HGSE Impressions over the past month – focus on 9/19
- 138. Momentum…
- 139. Brad Paint By Numbers
- 140. Et cetera
- 148. thanks.
- 149. Copyright © President & Fellows of Harvard College Behind the Retweet: Learning from Social Analytics Harvard Analytics Academy, March 19, 2015 Robert Bochnak, Assistant Director, Alumni Marketing and Communications, Harvard Business School, Harvard Analytics Academy, March 19, 2015
- 150. -Assistant Director of Alumni Marketing and Communications at the Harvard Business School where I manage the office’s social media channels -Have helped generate over 18,000 unique alumni interactions via Twitter since January 2013 -Graduate of UMass-Amherst and the Tufts University Graduate School of Arts and Sciences -Author, Social Media Matters blog, http://robertbochnak.wordpress.com/ -Superhero sidekick of Aniceta 15 0 A little bit about me….
- 152. Our Approach: Identify HBS alumni who are on Twitter and add them to the appropriate list. 15 2
- 153. Our Approach: Engage alumni through retweets, original tweets promoting their work, tweets sharing events, tweets connecting alumni with each other, tweets to alumni based on their interests, etc. 15 3
- 154. Our Approach: Leverage the relationships we have developed with alumni so they participate in discussions. 15 4
- 155. But before all this engagement could happen we needed to collect data on our alumni…. So we decided that each alumni interaction we had on Twitter would be tracked on an excel spreadsheet. This sheet would include information drawn from the Twitter profile and alumni directory file of each alumnus/a and would include information such as city and state of residence, interest, year of graduation, and section. 15 5
- 156. Data Collection 15 6
- 157. Data Use With this data—as earlier slides conveyed—we’re able to send targeted tweets to alumni based on their interests. 15 7
- 158. Reporting and disseminating the data -Each interaction with an alumnus/a is “scored” on the tracking spreadsheet: 1-2 “touches” a month equals “minimally engaged”; 2-4 “touches” a month equals “moderately engaged”; and 5 or more “touches” equals “highly engaged.” -Interactions are retweets, favorited tweets, or original tweets by individual alumni. 15 8
- 159. Reporting and disseminating the data cont. 15 9
- 160. Reporting and disseminating the data cont. -We also measure the connections we help make 16 0
- 161. Reporting and disseminating the data cont. -Twitter-based interactions are included in monthly reports. 16 1
- 162. Reporting and disseminating the data cont. 16 2
- 163. Reporting and disseminating the data cont. 16 3
- 165. YOUR GUIDE TO EMAIL METRICS DETOX
- 167. EMAIL METRICS Which camp do you fall into?
- 169. WHAT DASHBOARD?
- 170. BEWARE OF METRICS BLOAT ○ Numbers replace your business goals ○ Confusion among team members ○ Lost alignment within organization ○ Spread thin aggregating metrics ○ Metrics no longer used to inform business decisions
- 171. HOW DO YOU TREAT METRICS BLOAT? a. Invest in more fancy tools b. Wither and die c. Other
- 172. METRICS ARE PEOPLE, TOO Show some empathy, would you?
- 174. “How are you going to build that relationship?
- 175. PROVIDE VALUE Events ○ Workshops ○ Lectures ○ Screenings Interesting content ○ Articles ○ Online videos ○ Digests Service ○ Online application ○ Customer support Stewardship ○ Funds ○ Volunteer opportunities
- 176. MEASURE VALUE Your business goals should align with how you provide value.
- 178. THINK ABOUT METRICS IN TERMS OF PEOPLE AND VALUE
- 179. EXAMPLE
- 180. EVENTS Place your screenshot here
- 182. Place your screenshot here PUBLISHING CONTENT
- 184. STEWARDSHIP Place your screenshot here
- 186. Place your screenshot here SERVICE
- 188. ONCE YOU FIND YOUR BEACON...
- 189. START SIMPLE Email Attendees December invitation 25 January invitation 30 February invitation 5 March invitation 25
- 190. BUILD TRENDLINES
- 191. ASK QUESTIONS ○ What happened in February? ○ Was the creative different? ○ Was the email sent at a different time? ○ Did the email go to the wrong list?
- 192. CHOOSE YOUR ADVENTURE ○ Wait for more ‘beacon’ data ○ Tweak your strategy (one variable at a time) ○ Follow the trail with more analytics
- 193. CATHERINE’S GUIDE To Email Metrics Detox
- 194. 1. IDENTIFY BUSINESS GOALS ○ What value do you offer? ○ How are you building a relationship with your audience?
- 195. 2. CONNECT YOUR COMMS STRATEGY TO YOUR GOALS What is the path to your end goal?
- 196. 3. TRACK THE END GOAL Use the trendline as a proxy for success.
- 197. 4. WORK BACKWARDS ○ Ask questions about anomalies. ○ Dig in deeper, if you must.
- 198. 4. TWEAK One variable at a time!
- 199. CATHERINE’S EMAIL METRICS DETOX PLAN 1. Identify your business goals. 2. Connect your communications strategy to your goals. 3. Track the end goal. 4. Work backwards. 5. Tweak
- 201. Tracking the Life of an Event Using Analytics
- 202. Event Life Cycle
- 203. Define Success and Plan Accordingly
- 204. Measures of Success • Registration • Revenue • Attendee satisfaction • Target demographic representation • Smooth event flow • Numbers of first-time attendees • Social amplification • Smooth registration process • Email engagement • Social connections during/after event
- 205. Communications Strategy • Who is your audience? • What channels? • How frequently? • How long is your campaign? • What is your content strategy?
- 207. Stay Nimble
- 209. Save the Date 1st Invite 2nd Invite Save the Date 1st Invite 2nd Invite Your Harvard: Mexico Your Harvard: China
- 210. Tell the Whole Story
- 213. Dive Into Attendee Data
- 218. • Registration • Revenue • Attendee satisfaction • Target demographics • Effectively applying learnings from past events • Smooth event flow • Numbers of first-time attendees • Social amplification • Smooth registration process • Email engagement • Social connections during/after event Measure Success
- 220. Escape Excel Hell: Make Your Reports Memorable and Actionable
- 221. Hi! I’m Mike Petroff • Digital Content Strategist • Loves GIFs (but not emojis) • Twitter: @mikepetroff • New Dad (excuse any puns)
- 222. IS THIS YOU?
- 223. IS THIS YOU?
- 224. IS THIS YOU?
- 226. I’ve been there. I’ve tried.
- 227. I’ve been there. I’ve tried. I’ve failed.
- 228. If you build it, they will come?
- 229. If you build it, they will come?
- 230. Data alone is not inherently helpful. A brief history lesson.
- 231. Data and maps.
- 232. London Underground Tube Map ▣1908 Source: datavizblog.com/
- 233. London Underground Tube Map ▣1908 Underground lines were laid out geographically, often superimposed over the roadway of a city map. Source: datavizblog.com/
- 234. London Underground Tube Map ▣Meet Henry Beck an English engineering draftsman.
- 235. London Underground Tube Map ▣1933 Source: datavizblog.com/
- 236. London Underground Tube Map ▣1933 Beck believed that passengers riding the trains were not too bothered about the geographical accuracy, but were more interested in how to get from one station to another, and where to change. Source: datavizblog.com/
- 237. Format matters. What is the intent of the reader?
- 238. Bring focus.
- 239. Bring focus.
- 240. But… We’re buried in raw data. How do we bring focus to what’s important?
- 241. ‘’ From 2013 to 2020, the digital universe will grow by a factor of 10 – from 4.4 trillion gigabytes to 44 trillion. It more than doubles every two years. Source: emc.com
- 242. ‘’ Most companies estimate that they are analyzing a mere 12% of the data they have, according to a recent study. Source: Forrester Research
- 243. Fix the reporting process. Find what matters.
- 244. The Failed Reporting Process collect report
- 245. The Failed Reporting Process
- 246. The Successful Reporting Process collect analyze present
- 247. Collect ▣ Good data in, good data out ▣ Track consistently ▣ Share the collection process, and make it transparent
- 248. Analyze ▣ Provide insights ▣ Highlight opportunities ▣ Identify things that can be changed
- 249. Present ▣ Bring focus to the the most important thing ▣ Find the best delivery method ▣ Be consistent, but be memorable
- 250. Simplify Show the goal’s performance.
- 251. Simplify
- 252. Simplify
- 254. Kickstarter
- 255. Gates Foundation
- 256. Insights Show the reason it matters.
- 259. Daily Gazette report What: Quick view of daily Gazette email performance Who: Senior leaders, Gazette editors, Media Relations, Digital Strategy Why: Identify what resonates with readers, engagement trends Place your screenshot here
- 260. Daily Gazette report Not very helpful.
- 261. Daily Gazette report Better, but room to grow.
- 262. Weekly Gazette web traffic What: Google Analytics dashboard Who: Senior leaders, Gazette, Media Relations, Digital Strategy Why: Identify overall traffic, explain sources of traffic, anomalies, review owned traffic performance, social impact
- 263. Weekly Gazette web traffic Identifies not only top overall stories, but highlights traffic generated from owned sources like Harvard.edu, Daily Gazette email, and Harvard’s social channels.
- 264. Weekly Gazette web traffic ▣ A written summary of select referral traffic, including influencers on social media ▣ Identifies media pickup and impact on traffic
- 265. Monthly report What: Combine metrics across various digital platforms Who: Senior leaders, HPAC, AA&D Why: Analyze trends, large spikes and dips, measure against previous month/year
- 266. Monthly report ▣ Compare against previous month and previous year, use conditional formatting
- 267. Monthly report ▣ Provide most important findings and with each report; explain changes and future opportunities
- 268. How to get started ▣ Define the goal and the audience ▣ Identify metrics that measure the goal ▣ Collect metrics consistently ▣ Simplify to show performance against the goal ▣ Present insights that are memorable and lead to action YOU control the context.