Trust UX - Balancing Privacy & Personalisation (Strata+Hadoop World, 2015)

This connection we have with our smart-everythings, it is no longer merely about exchanging data back and forth. Our connection with computers now does revolve around values which we normally find in human-to-human relationships: understanding and trust.

We humans expect that these machines do the computing effort to understand us; smart personalisation. We also expect that we can trust these machines – and the companies behind them - to keep what they learn about us to themselves. We expect them to respect our privacy. Our security.

And as designers, we need to deliver great, personal experiences. We also need to deliver trustworthy products. We owe it to both our users and the people who hire us to actively think about privacy, and to implement privacy in the flows and designs we deliver.

Presented at Strata + Hadoop World, London 2015.

Transcript

1. TRUST UX ann wuyts @vintfalken sentiance customer experience & UX BALANCING PERSONALISATION AND PRIVACY TO CREATE UNDERSTANDING AND TRUST Strata + Hadoop World, 2015
2. HAPPY TO INTRODUCE YOU TO MY NEW BEST FRIENDS THESE THINGS KNOW ME
3. THE FOUR PHASES OF COGNIZANT COMPUTING 1. SYNC ME Store copies of my digital assets and keep it in sync across all end points and contexts 2. SEE ME Know where I am (and have been) on the internet and in the real world. Understand my mood and context to better align services 3. KNOW ME Understand what I want and need and proactively present it to me 4. BE ME Act on my behalf based on learned and explicit rules 94% COMPLETED 82% WEARABLES UPDATE 31% … LOADING DATA INITIALIZING.. 12%
4. Meanwhile, at the Apple Watch announcement, 19× Personal. relevant are you engaging at the right moment? glanceable can you deliver value in milliseconds? personal do you approach people in the right manner? EXPERIENCES FOR THE MOST PERSONAL DEVICE EVER ARE
5. No other Apple device has ever been so connected to the wearer. It is important to be mindful of this connection . Apple Watch Human Interface Design Guidelines, 2015
6. UNDERSTANDING PERSONALISATION TRUST PRIVACY H2H MACHINE-TO-HUMAN RELATIONSHIPS ARE NOW ABOUT HUMAN-TO-HUMAN VALUES M2H
7. 2nd Annual Poll on How Personal Technology is Changing our Lives - January 2015, Microsoft   Concern about privacy jumped 5 points between 2014 and 2015. THE PRIVACY CHALLENGE
8. 91% of adults ‘agree’ or ‘strongly agree’ that consumers have lost control over how personal information is collected and used by companies. Pew Research Privacy Panel Survey, January 2014 LOSS OF CONTROL
9. MISCOMMUNICATION People are fearful of sharing their data largely because companies and government have not been good at clearly explaining how they use it. The Data Dialog, Demos 2012
10. TERRIBLE HEADLINES In the future your insurance company will know when you are having sex. Kashmir Hill, Fusion.net - 2015
11. 38% of those who say not concerned about privacy online say they do mind companies using information about them 71% of those who are happy to share personal information with companies and brands that they like are concerned about how information collected about them is being used by companies Personalisations vs Privacy, Ipsos MORI, 2014 USER INCONSISTENCY
12. THE FACEBOOK PARADOX THE FACEBOOK PARADOX 91% of adults feel consumers have lost control over how personal information is collected and used by companies. 58% of the entire adult population (and 71% of internet users) is on Facebook.
13. Having knowledge of the customer is the only durable competitive advantage for companies. Bruce Kasanoff, the author of Smart Customers, Stupid Companies
14. Increased knowledge of the consumer and the fine-tuning of offers that are perceived as personal and highly relevant should lead to an increase in spend. Gartner on the monetization of contextualization
15. DATA VALUE EXCHANGE insights in own behaviour customer insights peer comparison optimising processes & products personalisation (right message) relevance (right person, right moment) accountability (on ADDD) data driven decisions CONSUMER BUSINESS magic moon standard moon no moon value on the consumer side should be equal or more
16. privacy THE UNTRUTH UX VS The truth is that collecting information about people allows you to make significantly better products and the more information you collect, the better products you can build . Dustin Curtis, “Privacy VS. User Experience” (2014) #FALSE (or at least a very one-sighted truth)
17. Whether or not better products can be made by collecting more user data is a matter of… •  Context •  Opinion •  Causation/correlation •  Data quality (accurate? relevant? complete?) •  Define: better product MORE DATA, BETTER PRODUCTS?
18. BUSINESS VALUE CONSUMER VALUE PRIVACY GREAT UX personalisation PRIVACY AS A FUNDAMENTAL COMPONENT OF THE PRODUCT EXPERIENCE
19. PRIVACY @ SENTIANCE UX DESIGN Businesses  can  deliver  a  grand  user   experience  and  tremendous  value  to   both  customers  and  the  company,     only  if  they  safeguard  their  users’   privacy  and  security.   →
20. Niveau 1 Niveau 2 Niveau 3 aggressive driver … … waking up … city worker sleeping walking running car subway train bus boat zigzagging standing airplane* biking arriving at home, work, the gym, ... shopping lunch inactive noisy environment* in a meeting horse* @home working in company* couch potato workaholic sportive N2 moments N3 profiles N1 events watching tv motorcycle agitated* suddenly stopped tram climbing* SENTIANCE FROM MOBILE DATA TO SMART LIFE
21. why great UX requires privacy what UX & privacy have in common 10 examples to do both Privacy & UX right
22. DEFINE: PRIVACY Personal data is any information relating to an individual, whether it relates to his or her private, professional or public life.
23. (it is however, an essential requirement) SECURITY = PRIVACY “Security is a very important topic, but it’s primarily a technical topic, and to a large extent it’s a very well- understood one. If you pay attention to security, it is possible to get it right, whereas privacy is something that’s much more fluid and is much more about social norms, expectations, implicit contracts between consumers and providers.” Pilgrim Baert – co-founder of AlertMe
24. THE INTERNET OF EVERYTHING: DESIGNER ROLES ARE UNDERGOING CHANGE UX design has been extended to address all aspects of a product or service as perceived by its users – that includes the control they have of their personal dta, their privacy.
25. BEING CREDIBLE useful usable desirable credible valuable findable accessible User Experience Honeycomb (Peter Morville) credible 2004 the information you present to users 2015 taking responsibility to keep personal data safe
26. PRIVACY-BY-DESIGNER: DELIVER BOTH PERSONAL & TRUST We owe it to both our users and the people who hire us to actively think about privacy, and to implement privacy in the flows and designs we deliver. B. We need to deliver trustworthy products. A. We need to deliver great, personal experiences.
27. 1. You need to fully understand the end goal (by asking the right questions) WHY Why are we doing this? What do you want to achieve? What is required to achieve this? What is the best way? WHAT WHO Who is impacted by this? What do they expect? GOAL DATA PIA
28. 2. There are rules, guidelines, toolkits. (which continuously evolve) UX PRIVACY •  Apple, Android, .. design guidelines •  Interaction patterns •  Best (and worst) practice examples •  Models & frameworks •  User research methods •  Emerging trends •  … TOOLKITS: omnigraffle, illustrator, fireworks, pen & paper, … •  Existing & upcoming EU Law (GDPR) •  Local privacy act & royal decrees •  Local telecommunications law •  Privacy watchdog recommendations •  ToS of the platform (iOS, Android) •  Internal policies •  … TOOLKITS: information classification, risk assessments, privacy policies, PIA framework, …
29. 2. There are rules, guidelines, toolkits. (which continuously evolve) General Data Protection Regulation EXPANDED SCOPE any organization processing personal data of EU residents PRIVACY-BY-DESIGN & DEFAULT EXPLICIT INFORMED CONSENT DATA BREACH NOTIFICATION DPA & possibly consumer DPO REQUIRED Users/month threshold or location data DATA PORTABILITY
30. 3. Less is More. (value & proportionality) collection processing storing purpose PROPORTIONALITY
31. fewer data fewer less detailed shorter period in time de-idenfity asap Less is more, not just in quantity but also in identifiability & time
32. value of knowing cost of identifying IDENTIFIABILITY Why an IP an sich will no longer be personal data… (unless you are an ISP)
33. avoid when possible geohashes coarse location beware location GDPR: this is ‘special data’ → extra safety measures
34. 4. You can not do it alone. (it is multi-disciplinary and cross-departmental) Privacy requires a clear mandate to get things done. Everybody accepts it is important – but not a single department has it as a priority. Have privacy as part as the project plan and estimates as soon as possible. A continuous need to explain the significance of privacy in the overall product & company picture Have privacy as a deliverable, avoids the delays & soring costs of adding it after the facts.
35. Privacy is not only a fundamental right, it can also be a competitive advantage . Neelie Kroes Conform to EU legislation? Ready for the world market, then. People can trust you with their digital identities → sets you apart from competition
36. 5. The devil is the details. (and the cost of mistakes is high) •  up to 1,000,000 EUR fine or up to 2% of the annual worldwide turnover in case of an enterprise, whichever is greater (Draft GDPR, art 79) •  customers leaving •  •  customer complaints •  customers leaving UX Privacy
37. Most of these were hacked (security), but as soon as personal data leaks: Privacy & trust disaster too.
38. 6. Practice Honest Communication. (from the start) Consider a breach likely – and prepare accordingly. VISA’s ‘Responding to a Data Breach – Communications Guidelines for Merchants‘ guidelines. do not play the victim express regret take ownership be accountable
39. 1. What happened? (tell what you know at that time) crisis communications (works for downtime communication too) 2. What is being done *NOW*? (investigate, take systems offline, ..) 3. How does this affect your customers? (both short- and long term) 4. What are you doing to minimize risk? What can your customers do? 5. How do people get more information or updates? (folluw up) 6. What are you doing prevent this from happening again?
40. Privacy does not benefit from a “do first, ask forgiveness later” strategy. (avoid: “Hey, we just lost all this data of yours you did not we had in the first place.”) which data you gather & what for set correct expectations
41. FROM THE START: informed explicit consent (avoid: “Hey, we just lost all this data of yours you did not we had in the first place.”) clear affirmative action use plain language
42. 1.  You need to fully understand the end goal – Ask the right questions 2.  There are rules, guidelines and toolkits – Rules & tools evolve. Fast. 3.  Less is more – Value & proportionality 4.  You can’t do it alone. – Multi-disciplinary and cross-departmental 5.  The devil is in the details – and the cost of mistakes high 6.  Honest communications – from the start As designers, then what can we easily do that improves both UX and privacy?
43. GOOD UX AND PRIVACY 10 examples to get it right
44. 1. Design for Explicit: Opt-In By signing this contract, you agree we have the right to collect and pass on all your information. In case you do not want your bank to pass on your credit information to third partners and other divisions, please write ‘I do not agree’ on the contract and hand it over to the person behind the till. EXPLICIT EXPLICIT NOT EXPLICIT (hidden opt-out) NO YES IF YOU AGREE, PLEASE CHECK THIS BOX:
45. 2. Design for Informed: No Surprises
46. 3. Design for Choice: Consent In your designs and flows, take into account both having and not having the data. Design personalized experiences for when you have data. Design good alternatives for not having the data. Today will be sunny Weather for Olen, Belgium where we know you live. Check out the weather! Antwerpen
47. Privacy as a trading function Accelerometer Gyroscope Microphone Camera In-App Usage GPS Expected value User acceptance Wi-Fi Browsing History The more permissions are required, more added value is expected from the mobile app. Calendar SMS Light GSR
48. Clear & consistent, so people can trust you to point out privacy related features & settings. 4. Design for Trust
50. 5. Design for Because.. Explain your magic. When users know of the existence of a certain algorithm, their satisfaction with the product increases over time , probably as they start to understand its workings better. Yet when they discovered an algorithm they were previously unaware of, users felt betrayed.
51. Worst case scenario “In the extreme case, it may be that whenever a software developer in Menlo Park adjusts a parameter, someone somewhere wrongly starts to believe themselves to be unloved. ” – Eslami et all.
52. Because... allows people to correct you when wrong. Something we best figure out before algorithms get to act on our behalf.
53. 6. Design for Transparency Show people their data selfs. If we are going to allow algorithms and expert rules to steer our behaviour, we must know they understand that correctly. Allow for: -  Correction -  Reset
54. 7. Design for forming secure habits “Burner accounts” Kinja introduced these for anonymous commenting. They made private keys understandable through metaphor. “…if you lose the burner key initially issued we will not be able to retrieve this information for you or reset the account. Save your key.”
55. Reward secure behaviour Users that enable two-step security on their accounts will now receive a 10% discount off their monthly bill Mailchimp bill.
56. 8. Design to encourage privacy Access duration People forget to ‘revoke’ things. Supply limited time access options: -  WeChat: discoverable for 10 minutes (default) -  LinkedIn: access duration settings (weeks → months → years)
57. 9. Design for an Exit Offer Alternatives Make it easy to leave, but think about WHY people are leaving, and offer alternatives. Eg. -  “snooze” services -  less-email-option -  reset profile/account -  ..
58. 10. Design with P2P privacy in mind Do you want to know if your friends are (action/mood/..) ? Do you want your friends to know if you are (action/mood/..) ? Don’t allow OK
59. We influence what is acceptable.
60. Snowden Challenge at SXSW Combine exceptional ux with privacy at INCEPTION, not afterwards “ Combine exceptional UX with privacy at inception, not afterwards.” Edward Snowden’s Challenge to Startups at SXSW
61. As the need for permanent access to data increases, so does the need for ethics & morality. Weak AI (expert systems) Strong AI (singularity) Machine Learning Deep Learning Recommender Systems Autonomous Systems Transition period (Danger Zone)
62. GOOD UX AND PRIVACY Privacy is about more than data. Privacy does not benefit from a “do first, ask forgiveness later” strategy. It’s their data. Not yours. (Safeguarding it is a joint effort, though). It is their choice. Design the best possible experience, regardless the choice.
63. HELP BUILD THE FUTURE BOLDLY. DO SO RESPONSIBLY ann wuyts @vintfalken sentiance customer experience & UX Strata + Hadoop World, 2015 Thank you.