Episode 090
In this episode of Product Voices, we talk to Sheetal Rajpal, an amazing product leader with experience at Amazon, LendingTree, and PepsiCo. Sheetal shares insights on how machine learning, AI, data, and more are impacting B2C subscription products, and overall advice for product professionals. Topics include augmented reality, customer-centric product management, and the future of ML and AI. Valuable insights are shared on balancing customer needs, delivering value, and using data to drive decision-making.
CONNECT WITH SHEETAL:
RESOURCES MENTIONED
Thinking in Bets Annie Duke
Thinking Fast & Slow Daniel Kahneman
Never Split the Difference Chris Voss
TRANSCRIPT
Intro 00:03
Welcome to Product voices, a podcast where we share valuable insights and useful resources. To help us all be great in product management. Visit the show's website to access the resources discussed on the show, find more information on our fabulous guests or to submit your product management question to be answered in our special q&a episodes. That's all at product voices.com. And be sure to subscribe to the podcast on your favorite platform. Now, here's our host, JJ Rorie, CEO of great product management.
JJ 00:36
Hello, and welcome to Product Voices. On today's episode, I've got a great guest, Sheetal Rajpal, who is a product leader with a ton of great experience at Amazon and others. I'll give you four bio in just a moment. But we're gonna have a great conversation on machine learning how to build those, you know those technologies and into your products when it's appropriate. We're going to talk about how to focus on b2c subscription products, which is not always an easy thing to build and succeed in. And then the usual just you know, advice for product folks out there so couldn't have a better guest today. Very excited. Again Sheetal Rajpal is a Senior Product Management leader with 15 years of experience building software products, most recently an Amazon leading kids Plus subscription product, where she owned the full customer lifecycle acquisition, engagement and retention. What a cool job that is. She has expertise in driving growth for customer facing ml driven products. And she holds an MBA from Wharton. Sheetal, thank you so much for joining me.
Sheetal 01:42
Thank you for having me, JJ.
JJ 01:45
I'm excited to have this conversation with you. I as people who've listened and know me, in the real world, they know that I am a product nerd. I love it. I've done it for many, many, many years. But I'm not a technologist. I don't have an engineering degree. I don't have a computer science degree. And so I've always been really close to technology, obviously in my role, but I, I don't, I don't know it all that well, to be honest. And it's kind of an odd thing in these days in such a technology driven world. So I always love to have these conversations, I love to talk to folks like you who've who've really driven these products, whatever your background, right in terms of, of degrees and education, you've you've really led the world and building these kinds of products. So I'm really, really excited about this conversation. So let's get started with a little bit more color on your background. So obviously, I gave a really quick by bio, which is great on its own. But tell me a little bit more about your background and how your journey into product management has looked.
Sheetal 02:51
Yeah, for sure. So I tell you about my journey, it's really to tell you about my background. So my story is all about adventure. My I grew up in seven countries. And what that meant was literally picking up and starting new in a new culture, new new country, new language many times. So as I was growing up, I learned the secrets of like connecting the dots and in how to make friends and all of that. So that's kind of how I view creating technology products. I started off as an engineer, but I realized, I because of my upbringing, I was so curious about like, how these products were being used and learning all all the different things that go into product development, like market trends, what tech is doing, what the user behaviors are, and all of that. So. So I transitioned into product management pretty early and went to went to Wharton so and from there, like did a stint outside of tech came back into tech. And I just I just love building products. I feel like it's a great way to connect with people.
JJ 04:19
You know, I love your story. It's amazing. First of all grew up in seven countries. That's crazy. By the way, which countries? I'm just curious. Yeah,
Sheetal 04:28
So I was born in Austria, and I've lived in Saudi Arabia, Japan, where I learned English. India, Panama, Colombia and the US.
JJ 04:43
That is fascinating. Wow. But But I love your story because it it's it's an example of how no matter who we are, what our background is, we can bring elements to the world of product management, right? So you learned really early to be agile and to, you know, be flexible and to meet people and build relationships and communicate and, and connect the dots. I love how you said that. So it's fascinating. I always love to hear about people's backgrounds, and you have a really, really cool and we could spend the whole episode talking about that. That's fascinating. Yeah, maybe we'll do that over coffee someday. Okay, so so, you know, obviously, you have a lot of great experience in building machine learning driven products. So what's your take on the future? How do you see the future of ml and digital products playing out?
Sheetal 05:40
Yeah. So future of ML, right, and ml AI all the rage right now. I think that there is lots opportunities, you know, as we're building these digital products, we're collecting lots and lots of data about what's going on. And that's, that's going to power our future digital products, right. So everything from there three main things I wanted to touch upon. One is like personalization of user experiences, and predictive analytics, we're just in the infancy of that their future products are going to be highly personalized, and, you know, micro personalized to you, JJ, your experience will look very different than my experience. And so yeah, hyper personalization is secure, or will be coming soon. And number two is, you know, the focus on ethics and privacy, right? They were just touching on these surfaces, right now, how about, like copyright issues, and all of that. So, and privacy is going to become more and more important as we think about, you know, which data should be used in these ml models, and which data should be out of bounds, right. So that will become very important. Number three is virtual experiences, right? So, you know, everything from augmented reality to, you know, how we shop, right? Like, we're, we're seeing some of that, right, where you can place a couch in your empty living room. But how that will enhance the shopping experience is going to become more immersive, more engaging, more personalized, all of that. Having said that, I think, you know, at this point, we're in the pivotal time period where AI is not always the answer, right? So as product leaders, we'll have to figure out, like, how do we balance that, right? Where it's really exciting to go deep into the tech and try and try these different things. But how do we, how do we make sure we're not over engineering? And it's, it's not it just because it's exciting to do right now. Right? So so we'll have to figure out that balance. And I can talk more about that, like, essentially thinking about, like, what is the cost of integrating these ML into your product? Right, is it? There's the people cost, but then there's also the computing costs and retention costs of training these models? And yeah, so we will have to balance that, are we seeing really seeing the return on investment? Or do we need to wait, so when is the right time to make that investment for your particular product is going to be important, so
JJ 09:02
Spot on, and so exciting, right? Hyper personalization, and ethics and privacy and virtual experience. I mean, they're exciting. They're, they're, they're something we're gonna have to keep learning about. As, as we go. And as we, as you know, communities and societies and companies, and, you know, the community of product and tech, but but I couldn't agree more with that last part about, you know, AI is not for everything. ML is not for everything. And sometimes, it doesn't even make sense, right? It doesn't make sense from a strategic or financial standpoint. And I think that's one of the things that I'm watching is how do product teams and by that, of course, I mean, you know, product management, design and engineering and, you know, all of all of the folks that are involved in this, how do we make the best decisions? And, you know, it's really nothing different than this is the market need and now let's go you know, try to Do you find the best solution for it, it's no different than what we've always done. It's just that, you know, as you said, it's, you know, this is like the shiniest object we've been around in many years. And so we kind of always want to, you know, make sure our, our products have that have that new thing. And, you know, we can say that, but sometimes that's not the case, and it's not needed. So I think that's a really, really important part, especially for leaders, especially for, you know, the folks who are kind of architecting the solutions, we really need to have some kind of boundaries on, you know, what's real, and what's just, you know, the shiny object syndrome. So, you know, I want to I want to turn a little bit because I, you know, your, your background in building Amazon subscription products, give kids Plus subscription product is so cool to me. And so Amazon is a, you know, a great company, obviously, and they're known for customer obsession, I use some some of the things that I've heard about Amazon and some of my speaking, just because I think it's, it's so cool. So, you know, knowing that, that Amazon is really about customer obsession, and doing things for the customer, how, how has that influenced your approach to product management and product development?
Sheetal 11:15
Yeah, so I was a big proponent of the customer, even before I go to Amazon, I think it was just kind of something that, like, attracted me to Amazon to right, because I was always thinking about the customer. I talked about my upbringing, right? Like, for me, it's really about, you know, using technology to solve users needs. And, and that was just my own personal philosophy. And then, when I came to Amazon, it kind of gave me new vocabulary, you've, you've probably heard of the working backwards method that Amazon employs. So working backwards is basically like thinking about what like new reality you're trying to create. That's rooted in the customer insight, and needs of the customer. And then just working backwards from that big vision that you have, right? So the way we do that, Amazon is essentially, we write a PR, a press release. And that press release, we just imagine we like, we have not built anything, we have not written even a single line of code, no user stories, nothing. But we just imagine that we are ready to launch. And we're writing a press release to the public about what that product is, is doing, what the new capabilities are. And it's a rah, rah piece. And it's, it's, you know, we we make sure we put a customer anecdote in there. So like, customers who have tried this product, say, blah, blah, blah, about it. And so it's, so that helps us like bring, bring a connection of the customer at the start of the development, right or not started the door, but at the start of the vision of the idea, right? So even just that at the idea phase, we are routing it to like one person, and the only that we're solving. So like we might have a quote from from you, JJ, right, like, okay, so JJ said this about this product and how it's changing the way that she is conducting her podcast now. If it was a solution about that, for instance. So yeah, so it's all at the start of it, just working backwards from it. So the PR is a big part of it. And then and then we have a expansive list of questions that we answer in an FAQ format. So these frequently asked questions can be from the external audience, right. So like, what would customers ask about these about this product and how how we are answering them. And that helps us refine the idea. But it could also these questions will also be from like our internal stakeholders, right. So what are some key key questions that another, another product area might have that maybe is overlapping? So for instance, like I was in kids, plus or work. So like, what does the prime team Prime subscription team have have to say about this idea? How are they going to be impacted? How are they going to be involved, and all of that. So we clarify all of those kind of thinking and decisions through this process. And it helps us like make sure that what we're thinking, the idea is grounded in Insight, number one, number two, we've thought through what levels of complexity they're going to be internally. And number three, like how the customer behavior might have to change or what the customers questions are going to be. And like, we're anticipating those and answering those during the visioning. And then yeah, and then after that, like, I feel like, that's when you're going from zero to one and you're innovating something big and new, right? So we talk about Amazon, like two way doors and one way door. So when you're starting something new, it's really the first time you're gonna do it. So it's a one way door, you want to really think through fully what the idea is flesh it out. But then a lot of the, a lot of the decisions that we make in business are two way doors. So you can you know, in an agile environment, you can iterate and make changes based on what customers are saying. So that's like when, when we're in development phase, right? That's, that's where we would like respond to customer's needs by changing based on how customers are interacting with our products. The
JJ 16:45
Press release in the working backwards is, is one of my favorite techniques, I think it is so, so powerful, to have us thinking about the outcomes that we, you know, want to see the behaviors that we want to modify or change or, you know, whatever we're trying to help the customer do or achieve. And, I mean, we literally start with that picture that vision, and I think it's so, so powerful. And I love that and I can imagine that, you know, everything kind of flows from there, right? And, you know, the FAQs and you know, you know, everything, I'm also a really big fan of identifying those one way and two days, two way doors, because I think, you know, I do talks and one of my chapters in my book is about judgment and decision making, essentially. And, you know, we get so paralyzed by the ambiguity around product management, that, you know, sometimes we forget that many of the decisions we make are not that big a deal, right? I mean, it's not that we want to fail at them and not do the right level of diligence. But sometimes we overthink things. And the fact is, for a lot of us, you know, working on the software products, for example, I don't think so we can, we can buy back, right, and we can walk back through that door and change in so I love that I think that's really great, great perspective of kind of how you've, you know, brought your perspective to Amazon, you've you've also, you know, learned from the culture there as well. So, that's awesome. So, um, you know, of course, you work on b2c subscription products. You You did a really amazing one. And so I always, I'm always fascinated by kind of the nuances of B to B versus B to C products. I think a majority of my time has been at b2b, b2b, excuse me, but I've worked at you know, some some b2c as well, but tell me, tell me your experience, like tell, let's talk a little bit about that, like, how do you see the nuances or kind of the most important pillars when you're building b2c subscriptions? Like, you know, those types of products like how does that customer driven focus and how does your your kind of, you know, approach fit well with that b2c subscription model?
Sheetal 19:08
So b2c subscription products, I have to say, comparing b2c and b2b, b2c is harder, because customers are fickle, right? And you businesses have like this buying process that's more longer term and more relationship based. So there's more more stickiness in a sense, right and more appetite for problems that may occur in the products whereas customers and customers, they are. They don't have that appetite, right? They, they want the best solution out there. And if there's a free version, they're not going to open their wallets. So I think you know, I love the complexity of b2c subscription products because I feel like it's harder and I enjoyed, like the harder problem solving aspect of it and driving that change. But to answer a meta question about like, what's, what's that one thing that I think about is like value, the value you're delivering to your customers, right? So if you're delivering value, you're going to retain the customer, there's obviously going to be external factors, right, like, the person who lost their job, they're like trying to rationalize their money. So like they're going to do, they might unsubscribe. But essentially, if you're delivering value, barring any external factors, impacting the customer, you're going to retain them. So that's my meta thing, right? So when when customers come in, like making sure that they're getting to their first aha moment, and by that I mean, they're looking, they're looking, they're coming to your product to solve a need. And the first aha moment is basically the first time they see the value of your product, and they have like, either completed a task or done something that is really what the reason why they came to your product, right? I think that's, that's number one. So having a really clear sense of how you're delivering value to your customers, and then making sure that, you know, if you think about like segmenting your customer, making sure that all segments of your customer base are receiving that value, right. So if there's a significant portion of your customer base, that's not getting enough value versus like what they're paying, then they're at risk of getting churned. Right. So having that detailed view of like, what your customer profile looks like, how much value are they getting? Are? Is that value changing month over month? Or is it staying kind of the same? I think that is, that is number one in my mind.
JJ 22:24
You know, I want to tie back what you said earlier about kind of the future of of machine learning and AI and because I can see, you know, all three of those right, hyper personalization, ethics and privacy and virtual experiences, really tying into the impact of the future of b2c subscription products. Do you see it that way? Do you? I mean, what do you think when we tie that together? And say, Okay, not that, you know, ml? Is the the end all be all for everything out there. But do you think there's going to be a big impact when it comes to those consumer products?
Sheetal 23:02
I mean, I think ML is going to impact all industries, right? So subscription products, are also going to be impacted, right? So the way that I think about personalization, right for subscription products is, is going to be similar. For instance, let me give you an example. I am a Spotify Premium user. So like I pay for that. So that's a subscription monthly subscription. And recently, I logged in and it's it told me, Hey, we were as a premium account holder, you're you get 15 hours of free listening to audiobooks. So, to me, that's like, a big delight factor that Spotify has given me. Right? Right. In a way that's going to deliver value to me if I like audiobooks, right. And I didn't have to do anything extra for it. So how does that change in a way that is personalized? would be would be the difference? Right? Like maybe they already based on my history of like, the songs I listened to, or the podcasts I subscribe to, based on that they've already kind of recommended books that I would like, right, right now I have to go and pick my own, but maybe they do already have a curated list, right? So like, any friction that in the product that's there for where the product relies on customer taking an action, right like me going and finding the audiobooks I will like, right? In this example, for instance, all of that friction, there's an opportunity for that to go Oh, wait, right. So it should already have curated set of books. And this is just for me to just pick up. And so I get that time back, right, whatever, 1015 minutes I spent and trying to figure out, Oh, that's great. Let me go and see what book I want to read. Right. So those, those are the things that ML can help solve from a personalization and prediction perspective.
JJ 25:29
You know, it's such a great example, because it's, it's essentially, and I think so many of us, myself included, you know, get so enamored with, you know, AI and ML, for example, at this point in kind of history, but whatever these new new things are, and we think of the big things we think of, you know, the world changing things that, you know, the applications that can, can help, which is fine and fun, and somebody is going to find those, right, but, but I, I think like you, while those things will come about that as well. It's about finding those small hassles across the customer journey that can now be improved or eliminated with, you know, something in the background. And that's gonna, that's gonna make a really big difference. So, you know, I love that example. Because just the fact that you get these, you know, 15 hours without doing anything is awesome. But then you're like, hey, actually, if it was already showing me some books that I wanted to just click on and start reading right away, that's even better. That's right. Yeah, yeah. The other thing I think about as you were telling that story, and as we were talking about Amazon, because I think I spend more on Amazon than any human should, should. My queue and people, right, but my cats love all the boxes that show up in our house. So they, you know, they get a new toy every time. But, you know, one of the things about consumer products is, you know, especially digital consumer products, is that every product we use, influences our expectation on every other product, right? And so when you've got a company like Amazon, building something that I mean, I can buy anything in the world in two clicks, and it gets to my door, before I even, you know, turn around, you know, that is just an experience that a lot of people excuse me, a lot of businesses can't replicate, well, I expect it, I expect it from the other apps from the other companies, etc. You know, you know, same thing that you were just talking about with you expect, you know, you know, either free content or whatever it is, right. And so, the more and more, you know, as consumers, we engage and interact with these products, the more that that whole ecosystem influences, and as product managers for those products, we've got to be aware of that, that our quote unquote, competition isn't just direct competition, but it's the experiences that others are building as well. Absolutely. The big question I always, always kind of wrap up with, you know, what do you what do you advise for us? Because we like to learn from each other and product. But before we get to that, the last question I have, is kind of how you go about using data to help your decisions and your team's decisions. Because, you know, we've talked a lot about some of the things that can influence consumers and, you know, you know, the outside influences they've been laid off, or they're trying to, you know, change their budget. You know, I don't know that we ever know that. That's why somebody turns so how have you gone about using data in a way that that helps you, you know, figure out what's driving consumer behavior one way or the other? Whether we're, you know, driving growth or driving churn, or anything else? Are there any kind of real, real highlights of the way that you use data to drive your and your team's decision making?
Sheetal 29:00
Yes, I will tell you three things. One is, I think this probably doesn't come as a surprise to to you, but like, you have to use a combination of quantitative and qualitative data. And the reason is, because if you only rely on qualitative data, which means like you're talking to your customers learning about their daily lives in figuring out how this product is going to solve their needs and whatnot, you you will go away, and then they will go and do whatever they're it's really hard to change the behavior. Or, you know, they'll say that, Oh, yeah, this prototype will solve our knee but they won't use it or they'll try it and something happens and they'll stop. So the best or the The other way that you learn about the customer is quantitative, right? So you, which means you use the data based on how they're using your product. So when when they're using your product you are like capturing all this data, and then you figure out, like, what they're doing what, where are the friction points? Where did they abandon your product? And then you use that knowledge to go back to them and understand, like, what happened, right? And obviously, you're not solving this for one person. So there's a lot of nuance to that, right? Like how you go about marrying the qualitative data with the quantitative data. But there's, there's lots of ways for you to like, figure out what's going on there. And, yeah, so so that's my biggest thing, right? Like you, you can't use just quantitative. There, you'll see like, some people have an inclination to only use quantitative some people have an inclination to use only qualitative and kind of depends on if they're right brained or left brain. But where product leaders really excel is when they're able to use both in a way that's complimentary, right? So that's, yeah, like, for me, number one, the biggest thing, right, and there's an art to it. So and we can, we can have a whole episode about this, just this one point. But I will leave it at that. And then the second thing I wanted to talk about was, you know, data, people have this feeling that you can use data to say anything and tell any story. So the way that I earn trust with my teams is, I make sure that I have a very clear decision criteria on how I'm making my decisions. And there's a lot there's transparency in it, right. So I start that from the onset, and I don't try to I don't try to like massage the criteria later to say okay, but now, no, no, no, we I still wanted to do this, even though the data told me no. So. So having that clear decision criteria from the onset, helps build trust with your team. Right. So that's like, number two, a really big important point I want to make sure I share with you. The third one is, you know, which will, you'll love this. You there's so much data that you can get into analysis paralysis, right. So the act of moving fast, can give you more signal than trying to get the perfect solution out there. Right. So once you have 70% of the data, that's my rule of thumb. And something that I talked a lot about in with my teams, once you have 70% of the information to make a decision. Just go right, let's just go and do it. Because the next 10 20% of data that you're going to gather, to make to refine your solution, slight and make it slightly better is not going to get get your bang for the buck, if you will, right. Like the act of moving fast, then you can iterate on it on your product. And it will it will get you in the same place faster if you just move fast once you have a good amount of information.
JJ 33:49
Yeah, I love that so much. All of that advice is so spot on. And yes, let's definitely do a whole new episode on decision making. Because how do you how do you make decisions, how to use data for it, and all kinds of data, and move fast to make those decisions? So really, really great advice. So my my final question to you is, you know, again, there's so many great resources, so many great people to learn from. What do you suggest what what are some resources that you have found, that have helped you in your career and maybe some of the folks that you've coached and managed and led in your career, any specific resources that you've liked?
Sheetal 34:31
Yes, I am someone who learns from books the best. So there's an there's abundance of books out there to learn from so I will leave it at that but one thing that I will share with TJ with you and your listeners is think about like what is the best way you will learn right? Is it Look, is it through shadowing other people? Is it through taking courses? Is it through actually doing, doing the work, right. And I have found that different people learn differently. So if you can figure out what you're doing dominant way of learning is, then you can make sure that you're optimizing your time spent on that. Right. And there's an I have some favorite books, thinking in bets. As a one I read recently, Thinking Fast and Slow, I think you're seeing the theme of thinking, never split the difference. Yeah, so I love reading books. I like to read not just product, or business books, but I also like to read memoirs a lot. I read Born a Crime, which is by Trevor Nova. So it just, I feel like learning from other people how, how they're growing up, and like, how their how their history is impacting what they believe in now, helps me understand diverse people. And that diversity, I feel like sometimes people, when they're building products, they, they only think about how they're using the product. And when when I think about building that arsenal of knowledge about like, oh, there's different view all these different people with different backgrounds, it helps me think about my users in a different light, light, that there are different segments, there are different way needs and motivations and behaviors. So I have to think about those as thinking about what my product is,
JJ 36:56
that's so so right. And I have found that the more we open ourselves up to different perspectives that, you know, we do learn better, and we do make better products. And I also completely agree with you that that each of us needs to to figure out which you know, what is our our best or most favorite learning mode, and, you know, focusing on that, and, you know, maybe I listen to podcasts, but you don't like podcasts or you don't like to read an AI, you do like to read whatever, there's no one way to learn. So I think that's really good advice as well. And all of those books you mentioned are fabulous. So we will link to all of those as well. And I've just loved this conversation I've learned loved the kind of learning different things from you and hearing about your experiences and how it's, it's kind of all tied together. So Sheetal Raj, Paul, thank you so so much for joining me and for sharing your wisdom with us. This episode. I've really, really enjoyed the conversation. Thanks for joining me.
Sheetal 37:59
Same here. Thank you for having me.
JJ 38:02
And thank you all for joining us on product voices. Hope to see you on the next episode.
Outro 38:06
Thank you for listening to product voices hosted by JJ Rorie. To find more information on our guests resources discussed during the episode or to submit a question for our q&a episodes, visit the show's website product voices.com And be sure to subscribe to the podcast on your favorite platform.