Human-centered data: using data science and human-centered design to grow your product
0 likes380 views
The document outlines five principles for leveraging data science to enhance customer experiences: starting with human needs, emphasizing context, personalizing experiences, maximizing customer value, and recognizing that analysis alone cannot predict the future. It highlights the importance of blending qualitative insights with quantitative data to understand customer motivations and behaviors. These principles aim to help companies innovate and effectively engage with their customers to drive growth.
Human-centered data: using data science and human-centered design to grow your product
- 1. 5 Principles: using data science to create great (human) experiences
- 2. A hearty welcome from your friendly Sprocketeers. Tres April Justin
- 3. We’re a data-loving growth strategy firm. Using analytics and imagination, we help our clients: 1. Better understand their customers 2. Capture more value from existing customers 3. Innovate and grow in new ways
- 4. What the hell is data science, anyway? The art of using data to drive decisions, inform and power processes, and enable businesses to grow and move forward intelligently.
- 5. Data Science + CX. The time is now. 1. It is no longer acceptable to not be a ‘data person.’ 2. Companies who don’t fully exploit the value of data will be at a severe competitive disadvantage. 3. Most importantly, data is digital exhaust from human behavior and a human-centered lens unlocks its potential.
- 6. LET’S GET TO WORK
- 7. Jargon check. Persona Journey map Machine learning Natural Language Processing "How might we" question Beer
- 8. Introducing the 5 Principles. 1. Begin with humans and their needs 2. Context is critical 3. Make the experience more personal, easy, and enjoyable 4. Don't leave customer value the table 5. You can’t analyze your way into the future
- 9. Let’s learn about the 5 principles by talking about BEER. LAKEBOARD BREWING CO.
- 10. Principle 1. Begin with humans and their needs. • All good product experiences begin with human needs • Blend qualitative understanding with available data and unsupervised machine learning to paint a picture of your customers/users • Example persona template in your packet
- 11. 1. GOALS 2. NEEDS 3. BEHAVIORS 4. DESCRIPTORS
- 12. Qualitative insight, descriptive analysis, and machine learning help you understand groups of customers/users. Qualitative insight Discover goals, needs, and motivations. Gain deep understanding of why people do what they do. Quantitative insight Define “to what degree” customers have certain goals, needs, and motivations. Machine learning! Let the machines help us find patterns and like groups of customers. Personas Research-based descriptions of different customer types with like properties. + + =
- 13. 13 deal making dana. beer makes for good business. 60% Male 35-54 y/o 6+ annual occasions “Sometimes business is done in the boardroom, no doubt, but more often deals get inked when everyone is relaxed. Beer takes the edge off when the SOWs come out.” Job-to-be-done: • Build relationships and advance business deals during the course of a work week Defining behaviors: • Has 2-3 top breweries and bars he conducts business in • Sometimes rents a room if hosting larger groups • May go to breweries on the weekend, but primary day parts are after work during the work week Food options Beer taste EnvironmentKid friendly Trendy
- 14. Natural language processing (NLP) + Machine Learning = Qualitative insight at SCALE
- 15. Let’s look at Facebook Portal.
- 16. We (politely) asked the machines to find us 10 clusters of positive and negative reviews.
- 17. 17 Here are three of the clusters about particular features. Easy integration with Messenger 12% “But this whole thing works so well for my family because it's built on top of Facebook Messenger.” “Since practically everyone has Facebook it's easy to get in touch with people.” Alexa is integrated. 11% “Alexa works well on this device until all three of my kids begin yelling contradictory orders or random questions at once, which makes Alexa shut down (and I don't blame her).” “It also makes it easy to have Alexa already connected.” High quality video calls. 25% “The camera and video quality was smooth and sharp.” “Great camera follows you real well.” “It delivers great video and sound quality.”
- 18. 18 wanted to do our own installation in our current home wanted to change to a cellphone type system wanted to feel a little safer wanted to expand my system wanted one that's updated wanted a system that we could take with us wanted to move and Frontpoint is wireless wanted something that had mobile capabilities wanted a temperature sensor and a smoke detector wanted something a little more modern wanted to add elements to the system wanted to make sure that we got a good experience wanted to have a home security because I was moving wanted to go cellular wanted to have a more up-to-date system wanted the police to be notified immediately Natural language processing and JTBD are meant for each other.
- 19. Now back to Dana (and beer). Check out what people are saying about LakeBoard.
- 20. Principle 2. Context is critical.
- 21. Principle 2. Context is critical. Plus up journey maps: • Natural language processing of online reviews • Quantification of surveys and existing analytics
- 22. Principle 3. Make the experience more personal, easy, and enjoyable. (exercise) 1.Grab Dana’s persona, journey map, and natural language processing (NLP) results 2.Using post its, ideate 3 ways to make the experience more personal, easy, and enjoyable
- 23. Principle 4. Don’t leave customer value on the table. Think from BOTH a customer and a commercial standpoint to identify opportunities to make the cash register ring. • Connect personas with transactional behavior • Specify and quantify incremental behaviors • Avoid accidental beneficiaries
- 24. Principle 5. You can’t analyze your way to the future. Start creating the future! • Innovation involves risk • Mesh data with creativity • Use our Experiment & Learn Sheet to define an experiment now!
- 25. Start with a hypothesis…what do you believe? Define your experiment…what will you do to test? What metrics will you measure? What does success look like?
- 27. 5 Principles in Review. 1. Begin with humans and their needs 2. Context is critical 3. Make the experience more personal, easy, and enjoyable 4. Don't leave customer value on the table 5. You can’t analyze your way into the future
- 29. That was fun. Let’s do it again soon. sprocketcx.com