How to Know Your Customers by Amazon Senior Product Manager

Editor's Notes

1. Senior Product Manager on the Customer Reviews team at Amazon. Primarily focused on Trust. Build systems that ensure the integrity and trustworthiness of the customer contributions system at Amazon. Before that worked on augmented reality and holograms, gaming, enterprise software and genetics research. Mixed background.
2. Brief preview of what we’ll talk about today.
3. Jeff Bezos, 2017 Letter to Shareholders First impression: wow, this makes my job really hard… how will I ever keep up? But this is also an opportunity: harness your customers’ voracious appetite and discontent to build a better product, the one they really want.
4. What is “working backwards?” We’ll get into that next. Ideally, you get into the customers -> incrementally better product loop forever.
5. TRANSITION INTO WORKING BACKWARDS Great products and services come from deeply understanding your customers. One way to achieve this is by ““Working backwards.” The idea is that you start with the customer in mind first and work backwards toward the ideal product, rather than the start from the product and try to make a connection to customers. Audience engagement: -do you own a smartphone? -does your smartphone have GPS in it? -do you have a credit card? -when do you typically use a taxi?
6. “UbercCab” original pitch deck. Can you see how some of the differentiators there clearly work backwards to the Passenger customer segment? Don’t forget about the Driver segment as well.
7. Uber actually has more than one customer segment: it also has drivers. And if this customer isn’t happy, they’ll go elsewhere -- like Lyft. Uber had to spend more time on other customer segments to keep their overall customer base happy. “Last year, in the midst of the many scandals and lawsuits that roiled the company, Uber launched its “180 days of change” effort aimed at improving relations with drivers. Highlights included a new in-app tipping option for drivers, allowing drivers to message riders, additional layers of feedback for bad ratings, and more money for out-of-the-way pickups.”
8. All successful products roughly have these two components. Right now we’re primarily focusing on the first one. Having done your research, what else might you need to deeply understand the customer need?
9. TRANSITION INTO HANDLING CUSTOMER FEEDBACK Empathy: Are you the same as all of your customers? Unlikely, but if so your job is usually a little easier. Most of the time, that won’t be the case. Avoid: dismissal, skepticism, bias, defensiveness, ignorance, blindness which are all natural reactions to feedback. It’s not about you, it’s about the customer. Empathy is a muscle which must be exercised. -- The New Yorker: “Silicon Valley’s biggest failing is not poor marketing of its products, or follow-through on promises, but, rather, the distinct lack of empathy for those whose lives are disturbed by its technological wizardry.”
10. Jack Ma, co-founder of Alibaba (was an English teacher) “Principle of treating people well helped launch Alibaba in 1999.”
11. Any time you talk to a customer, give them an opportunity to provide feedback. You need a minimum amount of feedback to start to understand themes. Bullets: Listening vs. talking: Listen 90%, talk 10% You’re there to get feedback, not lecture the customer Neutral vs. directed: Ask neutral, not directed questions How short was Napoleon? vs. How would you describe Napoleon’s height? or “"Wouldn't you agree that...?" Prompting and silence-avoidance: Don’t prompt answers, silence is okay Uncover hidden motivations: Find true motivation with tools like the “Five Whys” Show Empathy: reiterate that the customer’s voice matters and be an active listener. It’s not about you. Five Whys: Why? - The battery is dead. (First why) Why? - The alternator is not functioning. (Second why) Why? - The alternator belt has broken. (Third why) Why? - The alternator belt was well beyond its useful service life and not replaced. (Fourth why) Why? - The vehicle was not maintained according to the recommended service schedule. (Fifth why, a root cause)
12. Taking critical feedback is hard… natural instinct to get defensive. Instead, see this as an opportunity. Leave your ego at home. Every piece of negative feedback is gold. The hardest customer feedback to encounter is the harshly negative feedback that you know deep down is right.
13. Product managers give the strongest argument when they can align both data and anecdotes. Data gives credence to specific anecdotes. Anecdotes do not always match the data. The data cannot always be clearly explained without complementary anecdotes. Anecdotes can give you clues for where to look in the data.
14. Data helps you identify: -themes, patterns, trends -aggregate behavior -behavior over time -any of the above that a customer wouldn’t tell you, i.e. hidden insights Data will let you validate the prevalence of any given anecdote among the population.
15. If your company doesn’t use SQL, that’s okay too… it’s just a way of getting the data you might want. Replace that with some other way of getting the data you want and you’re good.
16. Aliza Rozen Twitter noticed the phenomen of “cramming.”
17. Fig. 1: Character distribution of tweets across languages. Fig 2: Worse cramming problem means fewer people tweet Increased character limit (to 280 chars) less friction to tweet more users tweeting, more tweets greater engagement Twitter saw more people tweeting with the character count increased to 280 characters. This change was controversial! Don’t forget to match customer data and anecdotes + consider all customers.
18. TRANSITION INTO AVOIDING COMMON MISTAKES Humility: According to Jeff Bezos, people who were right a lot of the time were people who often changed their minds. Voice your opinions strongly and passionately, but be prepared to change your mind when you encounter better information or a new perspective. Your job isn’t to be right the first time, it’s to be right just as you launch the product.
19. In psychology, the false-consensus effect or false-consensus bias is an attributional type of cognitive bias whereby people tend to overestimate the extent to which their opinions, beliefs, preferences, values, and habits are normal and typical of those of others (i.e., that others also think the same way that they do).[1] This cognitive bias tends to lead to the perception of a consensus that does not exist, a "false consensus".
20. If you know exactly what the customer needs you just do exactly that and you’re set -- right? No. This is very hard and I could have an entire separate talk on this, but will cover the must-knows here.
21. Assumptions are the kryptonite of product management.
22. Joe Gebbia, co-founder of Airbnb, went to Rhode Island School of Design (RISD)
23. Multivariate testing introduces more variables and can additionally measure how they interact
24. A/B testing is a little bit harder in physical products. Lola tampons (mylola.com). Chemical-free cotton. Can get 4 or 8 week subscriptions. How do we know if customers are loving the box customization feature? Measure these metrics by cohort. Always test with the minimum possible blast radius required to get the signal you require. Potential outcomes: customers loved it and all metrics were favorable