Today we’re going to talk about the innovative use of AI in driving product-led growth with Sonal Mane, Senior Director of Digital Customer Experience and Growth at Databricks.
About Sonal Mane
Growth-focused technology executive with 18+ years of experience architecting and executing multi-channel strategies that drive user acquisition, activation, and revenue scaling from inception to $500M+. Expert in building AI-powered growth engines that optimize customer journeys and fuel significant business expansion.
At Databricks, pioneered the company’s Digital Self-Service & Growth strategy, driving transformative revenue growth through product-led and AI-powered growth engines. Developed innovative GTM strategies, optimizing customer journeys with predictive analytics. Spearheaded the Digital Self-Service & User Growth organization and led a cross-functional team that achieved notable increases in user activation, Monthly Active Users (MAU), and new logo activation through targeted growth experiments.
As Global Head at Qualtrics, architected the digital growth transformation strategy managing a $130M P&L and achieving a 13% increase in retention rates. Built a global multi-disciplinary team to scale digital customer engagement programs globally across North America, EMEA, and APAC markets. Achieved remarkable user engagement, with strategies deeply rooted in growth marketing, data science, and operations.
Key Achievements:
• Established Databricks’ first digital-led customer growth funnel, integrating acquisition, activation, retention, and expansion metrics
• Built AI-led testing platform driving systematic growth experimentation
• Developed predictive user journey models accelerating SMB and Enterprise customer growth
• Created ML-based analytics solution to measure startup growth potential
Currently focused on empowering data practitioners across customer and partner organizations and scaling experiences through AI-powered automation and predictive journey optimization. Passionate about fostering high-performing growth teams and driving innovation through AI/ML and data-driven strategies.
Resources
Databricks website: https://www.databricks.com
Wix Studio is the ultimate web platform for creative, fast-paced teams at agencies and enterprises—with smart design tools, flexible dev capabilities, full-stack business solutions, multi-site management, advanced AI and fully managed infrastructure. https://www.wix.com/studio
Attend the Mid-Atlantic MarCom Summit, the region’s largest marketing communications conference. Register with the code “Agile” and get 15% off.
Register now for HumanX 2025. This AI-focused event which brings some of the most forward-thinking minds in technology together. Register now with the code “HX25p_tab” for $250 off the regular price.
Connect with Greg on LinkedIn: https://www.linkedin.com/in/gregkihlstrom
Don’t miss a thing: get the latest episodes, sign up for our newsletter and more: https://www.theagilebrand.show
Check out The Agile Brand Guide website with articles, insights, and Martechipedia, the wiki for marketing technology: https://www.agilebrandguide.com
The Agile Brand podcast is brought to you by TEKsystems. Learn more here: https://www.teksystems.com/versionnextnow
The Agile Brand is produced by Missing Link—a Latina-owned strategy-driven, creatively fueled production co-op. From ideation to creation, they craft human connections through intelligent, engaging and informative content. https://www.missinglink.company
Transcript
Note: This was AI-generated and only lightly edited.
Greg Kihlstrom:
Today we’re going to talk about the innovative use of AI and driving product led growth with Sonal Mane, Senior Director of Experience and Growth at Databricks. Sonal, welcome to the show.
Sonal Mane: Thanks, Greg. How you doing?
Greg Kihlstrom: Good, good. Let’s get started with you telling us a little bit about your role at Databricks.
Sonal Mane: Yes, absolutely. First off, thanks for having me on the podcast. And, you know, in the little bit research that I’ve done, congratulations on your 500 plus episodes. So thank you all to see the growth and, you know, the relevance of the topics. To start with intros, I’m Sonal Mane, as Greg was mentioning, and I’m responsible for driving growth with data practitioners. So think data engineers, machine learning engineers, data scientists across the Databricks ecosystem. And this includes both our customers as well as our partner organizations. And this essentially means that I get to lead out on how digital experiences get built out for these audiences. My journey in tech has been quite a ride with 17 plus years and now going and driving both product, go-to-market, as well as growth in companies such as Databricks, Qualtrics, and Microsoft. One thing that I’ve absolutely come to love is this trifecta, if I may, of AI meets product, go-to-market, meets customer experience. And it really is a place where you can sort of imagine and put in place AI-powered solutions and really build go-to-market strategies and delight your customers ultimately to land value. Prior to Databricks, I did found the digital customer success team at Qualtrics. And then at Microsoft, I was responsible for the Microsoft for Startups team and worked on the products org as well across Office and Windows. So that’s just a little bit about, you know, my journey here. And of course, you know, happy to be helping out any aspiring tech leaders in my free time. But yeah, it’s been fun working with all sizes of audiences and organizations as part of this journey. Excited to talk and share more about, you know, AI, digital growth and tech. So happy to dive in.
Greg Kihlstrom: Yeah, that’s great. And I love how you characterize the trifecta here. I think we’re going to hit all of those points here in our conversation. So there’s definitely a few things to touch on. I want to start with the combination of product-led growth and AI. So, we’ve had some people on the show talking about product-led growth before, but I think there’s a lot to explore here, particularly when we add in the layer of AI as well. But could you explain, how do you define product-led growth and why is it increasingly important?
Sonal Mane: Yes. You know, just to go back maybe a decade or two for our listeners, when I was part of the Office and Windows teams, you know, AI was this afterthought or like a separate team in a separate building that nobody really knew much about what was happening there. It’s all the cool kids and the whiz kids, you know, would be working on it. And that also meant that the way we would create product, you know, the way engineering would operate in terms of developing, you know, both features for the product, but also developing the user experience. It was sort of in a silo, right? So product and eng lived in the product and, you know, did some usability labs and customer validation, and that was it. And go-to-market teams lived mostly outside the product, which meant they would go to market, talk to customers. You know, everything from sales to marketing was all about brand and perception and prospecting and, you know, landing those big deals. And that delineation was very, very evident. And as I’ve reflected in the past, even just the past three years, with the infusion of AI now available to everyone, not only can you build your infrastructure, so think your backend schema, to your data architecture, to then building out the user experience itself, AI can now power pretty much the full stack But it can also power out-of-product growth strategies. And often the common ground and why that happens is data. So you need customer data when you’re thinking about what is the next user experience that we want to build and evolve the product to, in the same way that you need customer data to understand, OK, how do I crack this deal? Or how do I get through into this particular prospect that I’m trying to reach? What are their user behaviors and patterns? What have they used? So data being that common ground, more so than AI, has created this sort of infusion, you know, where AI is now available as a solution and creates this common ground, right? And so what that does is, be it product-led growth or sales-led growth, you know, depending on whatever strategy you’re taking advantage of, you can actually build solutions for your go-to-market teams around account intelligence and growth scores the same way that you can actually build features that are contextually available to your users. So it’s spanning this entire lifecycle of the customer. And said differently, when you think about product-led growth, what it has evolved into is this AI-powered, cross-functional go-to-market strategy. It’s no longer just, hey, we’re going to use product as a you know, driver of acquisition. And yes, that is the heart of it. So if you ask me conceptually, product led growth, you know, pretty much is where we use the product to drive both acquisition and retention, and then finally expansion of our customers. But really, it’s evolved past that to where it’s now, you know, hitting across all the different layers of an organization. The product still being the nucleus, still being that sort of primary driver of user adoption, of driving revenue, but it’s sort of taking over what one might think of as traditional sales or marketing efforts. It’s almost creating this sort of fluid layer across organizations.
Greg Kihlstrom: Yeah, yeah. You touched on the concept of contextual UX. I want to get to that in a second. But before I do that, just to talk about data, and certainly I’m talking to someone from the right company to have this conversation. It seems like we’re in a great position right now with AI, because I think you’re saying the same is we have this data is the foundation, right? You know, whether whether machines are using it, or humans are using it, you know, humans can use it and get insights from it, just like AI can can as well. But really, it’s it’s good data is at the foundation of all of that. Is that would you agree with that?
Sonal Mane: 100% 100%. And you know, any AI driven initiative, be it you know, on the product or the engineering side or the sales or the marketing side, you know, if you’re thinking about modern user experiences, if you’re thinking about, geez, how do I optimize my trial? Like every single problem, you know, really comes back to, well, how integrated is your data? Are you thinking about user engagement? Are you thinking about, you know, where do your users go even before they hear about your brand? So just everything from that sort of brand awareness, top of the funnel marketing down to then customer success and adoption and really driving that sort of expanded multi-product experience, and then closing it out with this viral growth loop that you might be building, like all of the above, right? Whatever your strategy is, data really is at the core of it. The advantage, and I know you mentioned, you know, being at the right company. I mean, data and AI is really what we live and breathe as part of our data intelligence platform. And one of the things we, pride ourselves in is this ability to work on the customer data. So it’s not really about, you know, let’s reshape or rethink, you know, what data assets you have. But really, let’s bring in and start working with your existing data as a customer. And to me, that’s really powerful, this ability to work with structured as well as unstructured data. Because as we know it, there’s lots and lots of different sources, right? Everything from a podcast like this one, to user patterns and clicks on a website. Data can be in many shapes and forms. And so solutions that really thrive in this space are the ones that help you asynergize it across different sources. And then on top of that, be able to actually understand your customer to then build that sort of product-led growth solution.
Greg Kihlstrom: And so we’re going to link to a blog post that you recently wrote in the show notes and where you’re talking a little bit more about product-led growth and AI and some other things as well. But you had touched on earlier the idea of contextual UX, but I’d love to get a definition of it as well as where’s the value in contextual UX.
Sonal Mane: Yeah, you know, if we think about a typical user experience, so when you look at, you know, Google Docs, or if you look at Figma, or even if you’re looking at, you know, if you’re using Lucid to create workflows, a lot of times, these user interfaces have mastered this whole idea of contextual UX to the point where you’re not really thinking about the user experience as much as you’re thinking about getting your work done. So the number one thing when we think about contextual UX is this idea that you’re tailoring that front end or that user experience based on the user’s context. What is their behavior? What is their persona? Are they a student? Are they an enterprise worker? Are they a developer? What are their preferences and needs? And what do they need at what point in time? It’s interesting. I was actually building out a workflow in Lucid recently. And it’s been a while since I did that. And it was super easy. There were tons of templates. So I was able to kind of draw from sample data, and then be able to, you know, real time as I think about the workflow. And this is just a very tactical example, but brings to life the point that I don’t know what the next, you know, block is that I’m going to need if I’m, you know, just rethinking this workflow on the fly. And so is it a square? Is it a diamond decision box? And so as you think about the contextual UX coming to life, really that’s what it is. You know, predicting the user’s needs, surfacing that relevant feature. So, you know, showing me that, hey, these are like your four options. Which one do you want? Right. But limiting them in a way that they’re relevant. You know, thinking about the fact that I’m real time building this workflow as I’m thinking through my diagram. And then really providing me tool tips and guidance with those templates and sample data. I know, by the way, while I’m at it, also showing me enough of contextual sort of background UX to then maybe share my content or be able to kind of provide another user to co-author it with me. So there’s all of these very simple design principles that come to life when one refers to a tailored user experience. which is essentially what contextual UX is, in the context of where the user is at. And why that’s relevant, and we can talk about how AI plays in, is A, needless to say, it improves user productivity, but it also reduces your friction to product adoption. And that’s the key. The minute you reduce your friction to adopt the product, now you have suddenly unlocked the product value. your time to value for the product has reduced, your feature utilization grows, you go from basic to advanced feature adoption. And suddenly, now this entire product has become attractive and sticky to the user. So that is really powerful in the concept and the context of, you know, the broader context of product-led growth, which is where, you know, as we think about how AI can start influencing some of those design principles, I think that’s where, you know, we can start seeing that sort of future contextual UX meets AI experience coming to life.
Greg Kihlstrom: Speaking of reducing friction, another big area, and I’ve seen quite a few examples of this even in recent weeks, of reducing friction is self-service. Generative AI certainly can play a role here. And I know you’ve written some about this as well. Can you share, you know, how do you look at generative AI and self-service and, you know, really how this can change the end user experience?
Sonal Mane: Yeah, yeah. So when you think about self-service, you know, there’s multiple layers in terms of the, you know, the offering. A lot of times and traditionally, you know, maybe three years or five years ago, we didn’t really think about this concept of digital-led services or customer success. That’s digital-led customer success or self-service. And a lot of times, even when we did, it was tied to the long tail. Today, self-service is almost the de facto of what you want for your product experience to be because there is a lot of noise. There are a lot of products on the market, especially if you’re a B2C product. And even if you’re a SaaS platform, this whole consumerization of SaaS movement is deep and pretty omnipresent everywhere. And so with that in mind, self-service has sort of become this, much like AI, almost like a basic need for how and when we think about product development. So what is self-service, right? So one is when a customer comes in or when a prospect comes in, It really starts with when they hear about the brand for the first time. What is their first landing page experience on your website? What are they reading and seeing? Everything from that first click to then activating your trial, converting your trial, and converting into, let’s say, a paid user, or then coming in with a larger enterprise license. All of these different scenarios and purchase paths can be augmented by what one might think of as a natural dynamic and context aware, again, going back to that contextual UX experience, conversational almost interface. So we’re generating everything from personalized landing pages and tutorials, guides that might be available based on who you are. I don’t want to show my data engineering personas or users the same kind of content that I want to show my data scientists. They both operate in different worlds. Yes, data is the common denominator, but with my feature set, they might care about a completely different feature set where model serving and fine tuning would be way more relevant for the data scientist. And so I should surface to him what that might look like, right? And for my analyst, I might want to show them what custom reports and visualizations might look like. And so transforming the product, but also the purchase path in a way that users can naturally go from awareness, to activation, to then actually adopting the product in a seamless way, like that entire art and science is self-service. And what you’re also trying to do as part of that is enable sophisticated troubleshooting. You’re trying to help with problem analysis or, said differently, simple solutions where you can actually answer questions for the customers on fly. And we’re all talking about agents and bots lately. And so all of those experiences, which could be digital-led, would be woven into self-service. But that’s where it’s not just limited to, hey, I have a bot, and I have an onboarding tutorial, and I’m successful. Yes, that’s great. You have a great onboarding experience that’s tailored. That’s awesome. But have you thought about personalized product recommendations? Have you thought about adaptive learning paths as part of your academy or educational website where you’re enabling your users? Have you thought about how you communicate with your end users, be it through your campaigns or be it through notifications? What are their preferences? So every single aspect, again, is touched when we think about self-service and personalization, in fact. I would say a key backbone of self-service is this whole art of predictive and personalized messaging, where you’re trying to reach users at the right time and help them, you know, get to that next step in their evolution.
Greg Kihlstrom: Yeah, and it sounds like definitely super important for that personalization and customization on the end customer experience, but also, you know, you touched on some things as far as internal, you know, because a lot of the hurdles that organizations face are those internal, you know, whether it’s process, you know, people process platform, whatever it is, sometimes all of the above. So, I mean, would you say then, you know, personalizing the experience for the internal teams is as important
Sonal Mane: Yeah, 100%. And that in and of itself is a very, very deep statement when you say personalizing the experience, because where my head goes then is, well, I’m an account executive, and I have my account intelligence dashboard, which only shows me my top 10 accounts, and what is my next best action. That’s a great example of a Gen AI powered experience. But at the same time, we can also do simple things like, especially now with all the LLMs, automating content creation, you know, Q&A within the product, test case quality assurance, we can do personalized trainings, one to many, you know, activities that include both live and recorded webinars and workshops. Lots of segmentation analysis, cohort analysis can be done on the backs of like with Databricks, for instance, Genie is one of our features that will literally let you talk to your data and ask questions as if you’re a business user without even ever having to need an analyst, you know, to build all the fancy dashboards. So there’s a ton more beyond what the experience of a day-to-day employee looks like, and then assisting those employees with all of these, you know, sort of technologies that come to life. So yes, absolutely.
Greg Kihlstrom: So last topic I want to talk about is just more about growth and helping to overcome some of those barriers to growth with AI. I was wondering if you could share an example or two of how have you used generative AI to contribute to growth and where do you see the opportunity there?
Sonal Mane: Yes, absolutely. So just to anchor back, there are a ton of different barriers And just because there are barriers doesn’t mean you wait for everything to be connected. We live in a really fast world where if your data isn’t in the best shape, or if you don’t have your customer 360, user 360 data integration, there are still things you could be doing with your existing data, though they might be limited. You could still work on ensuring that in the smaller data sets that you have, you can start building out solutions. And at the end of the day, it’s all about proving value to the rest of the organization and landing those investments in terms of both removal of that sort of tech debt, but also bringing those data silos together. So I’ll just start with that, like tons of barriers, right, including ethical, legal implications, including things like integration across silos, privacy, data security, and on all of the above. assuming that that’s all taken care of. And we have to go through our fair share of reviews with our legal team. The word personalized would always bring a ton of red flags. So the pain is real for anyone trying to implement. But if you think about really two or three simple solutions that are pretty powerful, one of them has been building out AI-powered assistants in our product. and really significantly reducing our response times, both from a support but also from a product adoption perspective. So typically, what would have been a ticket, you know, now the bot will actually serve up. That’s likely the most, you know, single most common scenario across a lot of companies these days. If you think about some of the releases we’ve had in the past few months, you know, building out our open source DBRX, large language model was a huge release for us. As a result of that, what that also meant is we could actually have an interactive customer engagement interface that could help with smart recommendations, content generation, and really power a conversation with our users while they’re in the product. So yes, for the bot, but then really bringing in DBRX as the LLM at the back end was very, very powerful. From a data privacy perspective, we eat our own dog food, so we’ll actually use things like Unity Catalog to actually ensure governance. And then if you’re thinking about what are some of the ways that you can actually personalize, it’s not just about persona development, brand, and voice. Yes, those are important, but really honing in on segments. And AI is the single biggest unlock. when it comes to hyper-personalized comms or personalized product experiences or analytics. And so we’ve really gone all out when it comes to predictive analytics and building our own platform when it comes to understanding our cohorts and how they function. So we have solutions like Amplitude and Heap in the market. We build our own. And that’s a huge investment for us to really understand that audience on the other side.
Greg Kihlstrom: Yeah, yeah. And to talk a little bit more about metrics, you know, I think AI is one of those interesting things where certainly There’s not a reason to throw out the traditional KPIs and all those kinds of things for marketing, but there’s also efficiency plays and other kinds of things that I think it enables that maybe some other technologies or features or things have not traditionally advanced. How do you look at measurements and what kind of metrics or indicators are you using to measure success of AI-driven initiatives?
Sonal Mane: That’s a great question. And done right, this is something we talk about internally, done right, AI and self-service are both invisible. Meaning if you have to think about, or if the customer has to have this thing about, oh, I was trying to do this and I couldn’t do it, then self-service didn’t do its job. So from a metrics perspective, I would say the same holds. prove that the cost to serve a customer is on par or actually much, much, much more reduced than what it would be with a human. The metric in and of itself, which is the cost to serve holds, right? But the targets that you set and the thresholds that you’re trying to hit I think those will vary significantly. So just to then zoom out, we’re monitoring things like user engagement rates. We’re thinking about what’s the time to value for new customers. It used to be a significant amount of time to where now it’s a matter of minutes based on a few recent releases we’ve had. We also look at how many issues that you can actually resolve, like I was giving that chatbot example. So how many issues or tickets could be deflected? So none of these metrics are earth shatteringly new, but what is new is we’re seeing a ton more increase in things like feature adoption or churn reduction. We’re seeing way more conversion from our trial into our sort of deep end of the trial where people will actually start going into our advanced features. So those are improvements and significant ones at that. But really from a bread and butter metrics perspective, consumption, user activation rates, monthly active users, those are very much still the bread and butter.
Greg Kihlstrom: Yeah, yeah. Well, Sonal, thanks so much for joining today. One more question here before we wrap up. We’ve talked about quite a few different topics here, but all some common threads here. What are some of the emerging trends that you’re seeing in using AI for product-led growth? And how should companies prepare to integrate these advancements to stay competitive?
Sonal Mane: Yeah, you know, like we were discussing earlier, Greg, it all starts with the data. So to me, it has always been a parallel track of sanitizing and creating the infrastructure aspect as much as building out the customer facing solution, be it through comms or through the product. So the first tip I would say is, you know, try and make sure that you have your data teams or engineering teams bought in and your support teams bought in when it comes to things like integrations across the different data silos. So that’s number one. Number two is in terms of actually thinking about implementation. The core of product-led growth is all about A-B testing. So we didn’t really get into experimentation, but MVPs or your first few pilots will likely fail. So be comfortable with that failure. Fail fast. And then the third thing is really about you know, when you start hitting those growth loops, and when you start building out, you know, end user experience, it’s very underrated, you know, in terms of getting that sort of customer interview or customer validation or product feedback loop going, a lot of people will run those usability labs, or they’ll run these segmentation surveys upfront. But keeping that constant drumbeat and having an active product feedback loop, is the single most effective way to have a successful PLG play, because you’re learning as you grow. And as your users engage, they advance. And so it’s sort of this sort of win-win solution when you try to think about improving the user experience as well as the product hand-in-hand with your customers. And the most successful companies have cracked that. In terms of emerging trends in the use of AI, like we talked about, everything from infrastructure your customer 360, your user 360, your support, you know, how customers engage and interact with your product, all the way down to, you know, your personalization of your segments and cohorts, like every single thing can be an AI-powered solution. The one thing I do often think about is, do you really need this beefed-up data science team? No, you don’t. You actually need a mix of personalities or mix of thought leaders in the room. You might need a few data scientists, a few data engineers, but don’t undermine the skills that you already have on your team. AI is something that you could also self-learn. So that’s just my two cents on the future of AI, both from a organizational team upskilling development perspective, but then also in terms of future emerging trends, we’re seeing it cut across pretty much every aspect of PLG.
