S8E13 - How AI is Reshaping PLG and User Experiences ft. Satya Ganni, CEO at Userflow & Beamer

Uh, and Beaver on the other side, it's, it's more on in the similar lines, but in, um, it's like a product adoption and engagement platform for, you know, product leaders and product marketers. Right. So, Hey, Satya, such a pleasure to have you on the show.

Um, ASP allows you to kind of post and manage your own servers and data Um, so kind of kind of as I said, it's a precursor to sass Um where you have to do customization licensing Set up all of it, but but either either on a single machine or multiple machines and deliver over the internet So kind of started there. That was my, um, first, uh, company, um, and, and since then it evolved into SAS.

So, uh, as I said, for the last decade and a half, I've been predominantly, uh, into SAS, uh, selling across different markets. So enterprise, mid market, um, uh, SMB, um, and, and it's been fascinating, uh, so far. Yeah.

So like, you know, what, what. Keeps you going. And you know, what, what makes you feel like, you know, I am like, you know, I always want to, you know, build a product and build a company. And, you know, uh, because that's, that's, that's really crazy to be like, you know, a founder of like six companies and now like seven and eight one.

Um, funny part is, you know, I was a mechanical engineer by education and a self taught programmer. So I thought, how can I marry both? Um, because software engineers has Visual Studio, mechanical engineers has CAD tools. So for a long time, I was thinking, how do I marry both? And then I came up with this idea that Hey, we need a visual studio like tool for mechanical engineers. So that was my first company, right? So I think it all starts with the problem.

Um, I think what keeps me going is, um, I think the technology space is super exciting, always, uh, evolving most, most fast. Um, and it also scales fast, uh, if you think there are right, uh, guardrails, um, it also scales fast, right? So I think these, these, these are the different things that keep me going, a lot of exciting problems to solve, um, using technology as the leverage to, um, build or, or, uh, scale. And, uh, so many opportunities, um, you know, that we have before us. Yeah.

So, yeah, like, you know what, what, what do you feel like, you know, what, what's that role Right, uh, of AI in, you know, enhancing these PLG strategies?

Um, you know, uh, you, you, you, if you're adopting organic strategies to dive, uh, you know, better we start to sign up conversion or a better, Uh, time to sign up or or a reduction in CAC. So you can use AI to generate the content, publish the content and drive visitors back, right? A second thing is even the personalization of the website or landing pages.

So for example, if you sell to two, three different user personas, and um, when you use the content to drive them to a website or landing pages, you can actually personalize the landing page. landing pages or the websites that suit to their needs, um, right? So that's acquisition. But even when you look at activation, um, as I said, if you have a large volume of users coming in, how do you cohort them appropriately to say, Hey, for each cohort, I want to accelerate the time to value.

I want to improve the activation rate. Um, similarly. So I think, I think at every step of the PLG funnel, you can use. Um, a led tools or techniques to be able to personalize and drive better conversion. So that's, that's how I would think about it, right? Um, PLD is all funnel driven. And if you can, uh, think about a led strategies at each stage of the funnel, um, I think that, that, that's what would produce a value.

And, you know, uh, then they go through the activation and onboarding, as you mentioned, which is like, again, very personalized to them or their role or their industry, like different levers that you can pull. Right. So, uh, but I'm sure like, you know, some, some, sometimes, uh, this personalization is, is driven by data, right.

Um, and data, yes, you know, that you get from, let's say, you know, intent platforms or, you know, something like an Apollo or zoom for, you know, maybe they have collect, you have collected, you know, It's a crunch base or, you know, multiple, you know, sources, right. And you never know, like, you know, like how accurate is the data, right. Because you're getting it.

So, and, and, and if you solely like very much rely on, uh, you know, this, this level of personalization, which you, uh, are collecting or you're, we are organizing it. Um, from data that of collecting from multiple sources, the personalization can go for a toss as well. Right.

So it's sometimes like, obviously it can be great if you blend it correctly, but because of like data not being right, might be like one problem, but there can be like several other problems, uh, where it can go wrong as well. Right. So where do you think in your experience where you've seen like PLG really win and in, in, in some cases where it, you know, it does fail.

I think the second, uh, uh, precursor, uh, is that you want to have some elements of, uh, uh, virality, um, in your app or collaboration, right? So for example, Uh, products like Slack or, or Zoom is a perfect example, which is what we are on, right? So I can't do Zoom talking to myself, right? I need a person on the other side of it. So that means that I have to invite you and you have to be part of the conversation.

So I think certain products have these natural ingredients, whether you call Miro, whether you call Slack or Zoom, and you know, a lot, a lot, you know, wide majority of SaaS products don't have this virality, right? So first of all, I think we have to, uh, figure out, uh, the misnomers of, uh, PNG, um, and, and, uh, do it. Second thing is, I think the inter, the con, uh, interfaces are changing very rapidly.

I think we'll see more and more of, um, I think right now we have this, uh, classic, uh, application structure model where you have a front end that's driven by, uh, you know, components and, and, uh, you guys, I think, um, it will change to more conversational interfaces. It will change to more inside. When, um, uh, conversational interfaces, right? So I think that shift we're in the early days of making that shift happen.

Um, as I said, uh, PLJ as a, you know, uh, as a standalone is, is one aspect of it, but I think you have to look at holistically in terms of your applications, the experience to the end users, um, and, and some of the metrics, I think, I think we are in a moving ship, um, here, especially with the AI coming into the mix, right? So. Still, still early days. Yep.

But you, uh, generally like, you know, as you mentioned correctly in a PLG offering, right, um, where, you know, these, these are more open tools where You know, uh, there is a sort of fireality built into the picture and, you know, and, and these tools, like all the men, the ones that you mentioned, right. Even Slack or zoom or, you know, Miro and, you know, Figma and all of these, like they have a, they have an onboarding process, right. Or any PLG onboarding process. Right.

Um, and, and, and during that onboarding process, you collect like. Uh, some information about like the user, right? Typically what's their role, like, where do they hear about you? Um, you know, what is the, what is the use case that they're trying to, uh, you know, uh, trying your product out for, right? Uh, what department do they come under? Right. And, and maybe some more right here and there.

Uh, obviously they, you have their email, so you can figure out like what company do they work for, what industry, you know, how big is the company, what industry they belong to, and, you know, some more information there. But, uh, you know, I've seen often that, um, like companies do collect this information, but, uh, you know, they, they don't necessarily personalize the, Um, the product, right.

Based on that information, just to give an example, if, uh, and for the listener as well, if you have signed up on Canva ever, right, like very easily. So they do have a very good onboarding, right? Like you sign up on Canva. They ask you like, what's the role they ask you? Like, what do you want to do with Canva? And, you know, some things and, and obviously they know what's my industry. They have all that information about my role, my responsibilities, my use case.

But still, when I land on Canva, they still like show me, uh, the, the, there is nothing like personalized in the home screen. Yes. Right. For me versus let's say you sign up, like Satya, Satya signs up or I sign up or somebody else signs up. Like all of us see the same home screen. Then what's the point of collecting that data? Right. Uh, that's, that's what makes me think, right.

So how can companies in your experience that they're like, how can companies go from, you know, just not just collecting data, but also like taking actions, uh, on using this data.

So that's where even when you log into some of the social apps, whether it's Instagram, Access, you know, I think that's a personalized experience, right? So they're using data well. To really personalize the, um, uh, personalize the experience or, or the feed, right? Um, I would say in, in, in terms of SAS, we're still in early days of, of bringing that level of personalization.

And I'm sure whether it's Canva or somebody else, they're deeply thinking about, Hey, what is the problem the user is trying to solve? How can we personalize the experience? And maybe, maybe the answer is sometimes personalization is not required. Thanks. Um, because he's trying to do an activity to derive an outcome. So let's go stay focused on that. Um, let's go stay focused on that. Right. So, so that is also, uh, possible, uh, right.

Um, but, but I think more and more you will see, uh, applications, especially the borderlines of, uh, prosumer, um, um, you know, self serve and everything else that start to span over. And, and, uh, I think majority of the PLD companies. is not limited to a geo. Um, like zoom doesn't sell in us, right? It is used by users across the globe, right? Um, across the globe.

I think when you see that huge diversity of, um, users coming from, uh, uh, various geos, various, uh, languages, et cetera, I think personalization would play a big role. So I think, yeah, the SAS apps would evolve just in the early, early days of that experience. Yeah.

And then like the feed kind of keeps changing to, uh, to, uh, a more personalized feeds as to like what you would buy. Right. So my Amazon feed would look like very different from your feed. And, you know, that's the same aspect for any B2C platform, as you mentioned. Right. And, uh, and this has been there since like many, many years, right. Even before like AI came into the picture, right.

So like now with AI into the picture, I think it should be more easier and more, you know, um, like, you know, easier to get basically more easier to get into, uh, doing something like this. So what, what is it that you feel in your opinion is stopping the companies? This is it. Um, like, because they don't have the right data, they don't have the right tools, they don't have the right, uh, you know, process or systems or like, what is your, uh, opinion on this?

And, and, uh, one is, of course, you, you have the data about the user, um, the org, the corresponding org that he belongs to, um, and, and et cetera. Second thing is, I think the actions that user do, that's where I think tools like product analytics are very, very important because.

What happens is when you get a first time user is actions in the app are very, very critical, you know, very, very critical because that tell you something that he or she might not have told you, um, in a, in a zoom call or something, right? Is he using the app? What features is he using? How deeply the feature is adopted, uh, why is he not adopting the feature? Right. Um, is it a discovery problem? You know, is it a value problem, et cetera, et cetera, right?

So again, you, you have to use that product data combined with the, um, uh, combined with his, uh, Uh, you know, behavioral data on the website combined with his formographic data, formographic means, you know, uh, again, the location, the org, um, where he belongs. So unless you combine all three types of data, you have to combine the formographic, the, you know, behavioral data that he exhibits, uh, on the website and blog and other sources or to your email campaign. The third thing is in app.

And I think you have to combine all of it to really understand what is the user trying to accomplish, right? You know, why did why did somebody from a mastercard signed up on this particular? Application what is he trying to behave? So the foundation for a lot of stuff Um, and that's where I think you probably need a data store or whatever, so that you're able to bring in all this data together and then understand the account or the user.

And then you can think about personalization or, or anything like that. So I think the secret sauce to a lot of PLG, very successful PLG companies is data, massive data. Yeah.

That, um, That generally makes all of this possible, like makes all the PLG possible. Right. So, um, what do you think are some of these platforms or what do you think some of these tools that, you know, help. Companies put together this whole PLG motion, right? Uh, we say like data and then data tools is like one aspect or one category, but then I'm sure like, you know, there's like a whole lot of other, uh, platforms that are required.

And I'm sure like, you know, get Beamer like Beamer and user flow are, um, you know, like, you know, part of that tool stack. So yeah, maybe you can share more, uh, on the tools that enable PLG.

And on the quantitative side, I would put something like product analytics, which is a must have, uh, for every B2B SaaS, uh, PMs, right? Um, uh, and then you have things like, you know, user onboarding or, you know, product adoption and product engagement platforms.

Um, the third is what I call experimentation, you know, and, and feature flagging, because you would want to run experimentation, um, within your own application, uh, to say, hey, you know, um, Are our people signing up converting better through path number one or through path number two? So experimentation is a very big piece of it.

Um, fourth thing is the feedback I think the ability to collect qualitative feedback Um, you know about the experience itself what is working what is not working etc, etc So I would classify this toolkit into four or five pillars one. There is the Usage aspect of it, which is driven from, uh, product analytics tool.

There is the experimentation aspect of it is, which is driven by, um, experimentation, feature flagging, all, all of those, um, uh, tooling, um, third is you have, uh, you know, the engagement side of the house, which is where we play, um, onboarding, you know, feedback, um, you know, uh, driving, driving, uh, adoption.

And fourth is, you know, there are There are tooling like, you know, session replay that allows you to watch user sessions and, and figure out a, which users have, have stuck through the journey, et cetera. So usually there are four or five pillars, um, that are, that are, uh, that, that are quite important, especially in context of B2C, B2B SaaS, I think B2C is a whole different world, um, because the level of. Uh, infrastructure and the tooling they require to personalize the experience.

Um, I think that that has evolved significantly. I think B2B SaaS is just catching up. Yep.

So before we move on to the last section of the episode, which is the lightning round, um, you know, I, I know we, you touched upon like, Hey, we are in the very early stages of the whole PLG. Uh, scenario. Right. So what, what do you feel like, what's next for PLG then, like, you know, in your opinion, like, okay, what's, what's the future of PLG lies in the next few years?

Uh, sometimes they might get lucky and have very high LT. So that's the fundamental limits of, you know, whether you go sales led or, or a PLG and both are phenomenal. Um, I would feel that a vast majority of the PLG companies, uh, go, go into a hybrid mode, hybrid mode, in a sense, they would have. Product led growth and product led sales. So which is, which is like a sales assistant motion, right?

You might not have sellers on the street, on the field, uh, but you might have sellers within, you know, within, within your own office or work remotely to serve the customer. So I would say 99 percent of the PLG companies. Uh, go, go into a hybrid model, right? Where you have product led growth, which, uh, addresses yourselves or users. And then you have a product led, um, sales motion where your, uh, sales and solutioning team is helping customers, uh, resolve the problem.

So I would frame that model as something like an inbound led sales assist, right? So you're still driving the needs inbound. But you're using the sales assist to do it. It's not for say classic PLG. Um, it's not self serve, but I would call it product led sales. So I think at the intersection of, uh, uh, PLG and PLS is where the future lies. Yeah.

Um, uh, and, and, and I think that that would be, uh, like, you know, I, I'm not, I've not been like very far, like Mary was following these companies, but I'm sure like they've, they've already been using AI to, uh, to, You know, drive, uh, or like, you know, precisely dark or with precision, tell you like, Hey, these are, this is the next company, uh, or prospect, uh, account that, you know, that, uh, is going to upgrade. Right.

So why not give them a little nudge, uh, and, and, you know, take them to the finish line. Right. So, yeah. Great. Great. Satya. Uh, so, you know, moving on to the lightning round, the last section of the episode. So, uh, I've already asked you like too many questions, but you know, I'm just going to ask you like three more and, uh, then you have to answer like with whatever is on top of your mind. Right. So, uh, starting with the first one, right.

What's the one thing that you wish you would have known? Yeah.

A couple of years, uh, he'd write again, maybe those people are also busy, but if I had a coach that could mentor me from, from my very early days of starting, because I took a lot of time to learn, um, uh, maybe, maybe it's my very nature. I took a lot of time to learn, but had a coach. See, uh, one thing you have to know is even people in the top one percentile in their field requires coaches. Look at Olymp.

You know, people who go to Olympics or people who are like top one percentile in their sports, they also have absolute coaches, right? So, um, are they not talented? Are they not skilled? No, I think, uh, there are certain, uh, behavioral aspects. There are certain, uh, uh, aspects that a coach can double. So I would say if I go back, uh, decades ago, uh, I would love to, you know, get, get a mentor or a coach. Yeah.

Um, I felt it's an amazing book that gives you a deeper, um, underpinnings of why we, you know, why we think fast and make decisions and, you know, sometimes why we move slow. It's an amazing book. It's a strongly recommend that everyone reads that.

Uh, thanks again for coming on and, you know, sharing all these insights with us and, you know, um, like everything about PLG and I'm sure like, You are at the, uh, you know, you're, you're building like two companies, which enable PLG, uh, to the core. Right. Um, so, um, great, great stuff there. So thanks again, once again, to, uh, to come on the show, take some time out and share all these insights.

Kind of where we serve, but, uh, really thanks for the opportunity. Yeah. Awesome.