 Well, good day and thank you for joining us here on theCUBE. John Walls here, bringing you to this conversation as part of the AWS startup showcase. And we're joined by Justin Bauer, who is the SVP of product for Amplitude. And Justin, good to see you today. How are you doing? I'm doing great. Thank you for having me, John. Oh, you bet. Now a pleasure looking forward to it. You know, personalization, that's where everybody's talking about these days and how do we better personalize our digital presence, our digital products? How do we get much more acutely aware of the end user at the end of the day and grow? I know that's what Amplitude's all about. So maybe if you just give us a 30,000 foot perspective on that, about your thoughts about personalization today and how Amplitude tries to affect that. For sure. So I think, first off, personalization matters because it actually works. I think we live in a world where as, you know, we're drowning in content and distraction and it's been proven that customers respond better to digital experiences that are more personalized, that are more relevant for them and frankly, just save them time. And the nice thing about this is not only do customers benefit, but companies do too. We actually see that a big impact on a company's bottom line if they're able to deliver a more relevant customer experience to them because that leads to better engagement, better retention and higher loyalty and lifetime value for those customers. So, well, let's just go right to an example then. I know you work with a lot of different people. There's anybody in particular that stands out. Maybe give us an idea of a case study here about what practices you put into place, the kind of evaluations that you do and ultimately the service that you're providing that allows them to increase sales and get a little more stickiness with their customer. Yeah, that's great, that's great. So, I think one company, customer of ours we're working with right now on this is actually Chick-fil-A. So, people probably familiar with Chick-fil-A, their mission is to be the most customer caring company in the world, which I love. And personalization is critical to that strategy because it helps them create a more relevant and seamless experience for their customers. And the experience itself in the app is actually pretty simple, which is the magic of personalization. So, you open the Chick-fil-A app, you see a list of menu items and those items are relevant to you based on your previous behavior. After you order your entree, you're then offered a list of personalized sides. And then after that, a list of personalized drinks. And the great thing is that as new items get introduced to the menu by Chick-fil-A, you see the ones that are most relevant to you based on predicted affinity and all of the machine learning that we're doing in the background. And so really now Chick-fil-A is actually, they're able to deliver a customized menu for everyone that automatically updates based on your behavior, your preferences. And I think the real beauty of this is that they're able to configure all this by a marketer through a simple UI. This did not require an army of data scientists or engineers, they're able to use the Amplitude platform to build out this entire experience for their customers. Right, because I mean, it seems like there'd be an enormous amount of analytics that you have to apply here, right? Because you got all this structured and unstructured data, you know, it's all over the place, right? And a lot of times people worry even though they have on hand. And so you got to help them sift through all this, right? So let's talk about that process a little bit for somebody who's watching and thinking about, well, that's all sounds well and good, but how do you kind of automate this? How do you make it so that we don't have to invest a lot in a team dedicated solely to, you know, sifting through our data and making it valuable for us? Yeah, I mean, I think that's the beauty of Amplitude, actually offering this in that that's actually our original first product, product analytics, that's what we've done. So we've actually made an out-of-the-box system that can read from all your different data sources. So whether those be your product sources, marketing channels, data that sits in your data warehouse, but it's not just piping that data. We then combine that into a unique identity profile for that customer across all those different touchpoints and also have out-of-the-box data governance so that you can make sure you maintain the quality of that data profile over time. And then that gets fed into our, we call our behavioral graph, it's our database that's actually built to both understand and predict future behavior. And so all of this happens effectively out-of-the-box for a customer. They don't need to do any of this themselves. We're managing all this for them. And then what they experience is an analytics application so they can analyze that user behavior, understand kind of what the drivers of different things like engagement retention are and then use that to actually personalize the product experience. And you mentioned machine learning. Talk about that aspect of this and how much more capability you have now because of what NL can deliver. And in some ways it adds complexity, but also obviously delivers exponentially, I would think, in benefit and value at the end of the day. Yeah, for sure. I mean, it's just not possible to do one-to-one personalization without machine learning. I think that's actually, when we talk about the benefits and the advantages of personalization, it's probably even worth taking a step back. Like there's a lot of different types of personalization. I think when you wanna do behavioral personalization, where you're truly getting to one-to-one experiences, you have to use machine learning. You compare that to maybe demographic personalization, which is actually, I think, when most companies talk about when they're doing personalization, they're actually doing demographic personalization. That's like, are you a male or a female? What's, are you living in a city or a suburb? But the reality is that light segmentation, it's not really that effective. Do all women who live in a city behave the same? Obviously not. And so we want instead to use behavior because your past behavior is the best predictor of your future behavior. And you need machine learning to be able to actually come up with for an individual what is their likelihood, propensity, to actually engage on any piece of content, of which think about for, you think about Chick-fil-A, how many different items they have on a menu. You can think about, like we work with a content company that has millions of different articles and they want to figure out what's the right article to put in front of you. That's just not possible to actually analyze that by hand nor actually orchestrate that in real time without actually leveraging machine learning. And so that's the exciting thing that's happened with new advances in supervised and supervised learning models that we can actually do those in generalizable ways for our customers. We've talked a lot about behavioral. So that's obviously metrics you can track, right? I saw something, I clicked on something, I acted on something, I bought something. These are all very measurable activities. On the other hand though, as you know, in the consumer space, a lot of it's emotion. Two, I make decisions based on my feelings or my thoughts or whatever. Can you do any kind of unpeeling of my motivation in this? Almost like empathetic investigation so that you have an idea of what social views I'm emanating or I'm sending off. So, hey, yeah, we can get John this way too. Yeah, so I think a lot of it is, I mean, we're talking a lot about the science of product development for sure and how do you bring personalization leveraging data. There is then the art of actually understanding like what are the emotional states that users are in like this isn't to say that the ability to personalize the product means that you're not to actually bring in the art as well. Like it actually is both about the art and the science coming together. And so you still need to, like you're still gonna talk to your customers, you're still gonna understand them and kind of what their different needs states are. But this is then taking what you have which you've built as a great product and then how do you optimize that? That's why we call it an optimization system and actually deliver the best experience based on that customer's behavior. So just to kind of flip this a little bit then what are you doing amplitude? What are you doing that hasn't been done before? I didn't understand that a lot of people think personalization just has a great horizon has a lot of great promise. Oh, but we're not there yet. I mean, what haven't we delivered on yet that you think amplitude is improving on and refining this capability? Yeah, so I think there are a couple of things there as to why we haven't fully seen the promise of personalization deliver. And I would say we're really starting to see that chasm emerge where there are some companies that you think of Netflix like obviously Amazon and others who've been really successful here but they've done it through armies of people. What hasn't happened is a self-serve way of doing this so that it does not require massive investments in technical resources. And so what we've solved for three things. One, we've already talked about it, but it's just so true. Like this actually in and of itself is not an ML problem first. It's actually a trustworthy data problem. Do you actually have the behavioral data that you can trust? Can you actually capture that across the entire customer journey? Because you can't personalize a journey if you don't even know what your users are doing to begin with. So you have to start there at that foundational level. And that is a big part of our secret sauce is that we've built a database specifically cater to helping you understand that journey of that customer across all the different platforms and channels that they do. That's not easy to actually unify behavior in that fashion and allow you to analyze that in real time. So that's the first thing we did is build that database. So that's number one. And that's just the foundation. You have to have that. Like I said, I think so many companies fail because they think we can go higher ML engineers, but if you don't have the foundation it's not gonna work. The second thing isn't necessarily technological. It's more cultural, but it is really critical. And I think our analytics applications helped a lot here. Which is you gotta break down the silos between marketing, product, engineering and data science. You actually have to have all of them working together to really be able to fulfill the promise of personalization because you have to be aligned and what's the outcome we're trying to drive. Like that's actually how I literally can walk you through like how the actual product works. But the first starting point is what are we trying to accomplish? Like in the Chick-fil-A example, it is we want people to buy more than one item. Okay, so that's your goal. Like you have to get alignment that that is the goal. Cause if everyone's arguing about different goals it doesn't matter what ML model, like the model needs to know what we're trying to actually focus in on. And so how do you bring people together and you do that through shared understanding of data? Like you do that through what we call it a North Star. Like we're aligned and what is the North Star that we're focused on and can you measure that? And that's analytics is focused in on that. And then when you have both of those you've got behavioral data, you understand the journey of a customer, you're aligned in the goals and outcomes you care about, then you can leverage machine learning to actually deliver that personalized experience. And the advances that we're making there are actually doing that in a generalizable fashion. So that does not have to be custom built for every single use case. And our models are now able that we can run a model basically every hour to update for a customer. And that scales horizontally. Well, I know Chick-fil-A certainly has a track record that is inarguable, right? And you've had a lot to do with satisfying that appetite for success. So Justin, congratulations to Amplitude. It's been a real pleasure speaking with you and thanks for the time today. Of course though, it's been great. Thank you for having me. Excellent, speaking with Justin Bauer, the Senior Vice President of Product at Amplitude and you've been watching the AWS startup showcase here on the Q.