← The Official SaaStr Podcast: SaaS | Founders | Investors
The Official SaaStr Podcast: SaaS | Founders | Investors
SaaStr 788: Going Multi-Product in the Age of AI with Webflow, Rubrik, Zoom, and ProductBoard
00:24:55
SaaStr 788: Going Multi-Product in the Age of AI with Webflow, Rubrik, Zoom, and ProductBoard At SaaStr Annual's AI Summit, we asked product leaders from some of the fastest-growing SaaS companies to share their insights on navigating the AI...
Loading summary
Transcript33 lines
- [00:01]
Podcast Host
Welcome to the official Saster podcast where you can hear some of the best Saster speakers. This is where the cloud meets up today on the Saster podcast.
- [00:12]
Annika Gupta
What may not be obvious about AI is I truly think that it's going to fundamentally disrupt enterprise SaaS. I don't think that people will be content with the modes of operation and the way that you consume and interact with applications today. And as we've thought about our own products, we're like, well, how do we insert this new way of operating? I actually don't think that that's going to be a side by side. I think fundamentally that's going to completely take over the way that SaaS applications are architected and everything is going to be a personalized experience and a personalized workflow that enables you to come in and surface the job that actually needs to get done and do it and then automate it in the future.
- [00:57]
Mahesh Ram
Hey everybody.
- [00:58]
Podcast Host
Thanks to the 10,000 of you who came out to Saaster Annual this year. We had a blast and big news. We'll be back in May of 2025 or in May of next year. That's right, SAS train will be a bit earlier next year. The 13th to 15th of May 2025. We'll still be back at the same venue in the SF barrier in our 40 acre San Mateo County Event center campus. And tickets are never cheaper than right now. So grab your tickets@saster annual.com with my code, Jason50 for an extra discount on our best pricing. That's Jason50@saster annual.com. see you next May at Saster 2025.
- [01:35]
Rachel Wolin
Please welcome to the stage Rachel Wolin, CPO at Webflow, Annika Gupta, CPO at Rubrik, Mahesh Ram, Head of AI Product and Zoom Review Accelerator at Zoom and Hubert Palan, CEO at Productboard. I'm so thrilled to be here with this group of AI all stars and thank you all for coming out to our session. My name is Rachel Wolin. I'm the Chief Product Officer at webflow. In case you don't know what webflow is, webflow is a website experience platform that gives anybody developer superpowers for some of the biggest brands in the world like orangetheory, Dropbox and more. And so today we're going to be talking about going multi product in the age of AI. And I'm joined by Anika Gupta, who's the Chief Product Officer at Rubrik, Mahesh Ram, who is the head of AI at Zoom, and Hubert Pilan, who is the CEO at Productboard. And so today we're going to really get into the messy details of what it means to build AI products. And so we're going to talk about how do you navigate adding new products, how are we disrupting our own products and how are startups just like you all building AI products? And so with that, Hubert, could you introduce yourself a little bit?
- [02:54]
Hubert Palan
Hi, I'm Hubert pon, founder and CEO of ProductBoard. We're a AI powered product management platform. If you want to confidently decide what to build next and align everyone around your roadmap.
- [03:06]
Rachel Wolin
Thank you, Hubert. And so I would love to hear a little bit more. You know, you have kind of this insider view of what kind of AI products people are actually building at Productboard on your platform. And so what are you seeing in the market? How are product teams like big and small using AI to disrupt themselves? And then how are product teams using AI to understand their customers better?
- [03:30]
Hubert Palan
Yeah, great, great question. There's a, we have like 6,000 companies building on top of product board. I'd say it's kind of probably obvious that AI is essential in the SaaS world because if you are building product management platform, if you're building Zoom collaboration type of platform, if you're building data security platform, there's a lot of use cases that are built on top of the data that the products are ingesting or that you're creating inside the product. And so it's probably a no brainer that summarization, getting sense of trends and insights, generating new, new types of data with AI on top of that is very valuable. The harder part to decide actually is to what is it that you should build? What AI you should build, what is the essential AI? And this is where it comes down to traditional product strategy, which is what is your core value proposition? What is it that competitors are doing so that you stay competitive? And I'll give you an example. In our case we were a product management platform and we were deciding what is it that we're going to build? Should we build an AI editor? People were saying, hey, you should really build like AI Transcription and so that, you know, Zoom doesn't have to transcribe, you can do it. And we said, well, wait a minute, what is our core value proposition? Our core value proposition is take the insights, take customer conversations and help the product people figure out what is it that should be on the roadmap so that it's really deeply understood and supported by the, by the customer insights. And so we said we're going to put all the other things on a back burner. And we're going to focus on this core value prop and we're going to build a really good AI powered linking of insights to the product features that you have on the roadmap and really help there. We saw a lot of, we see a lot of people though making a mistake where they're just like built ton of use cases kind of haphazardly. They go after everything that's easy. And then they're surprised they don't have a highly differentiated wedge in the market. And we see it also ourselves. There's a lot of startups that are coming at us and trying to sell to us what they built, which I think that is an important consideration. It's like, what is it that you want to build versus buy? There's a lot of tech that has been built and I'm sure that Mahesh, you have more experience with that. But that is something that I would also suggest to be very thoughtful about. Like can you sustain a business around the AI capabilities that you're building versus just building something that is a feature but is a very difficult thing to sustain in the long term as a business?
- [06:08]
Rachel Wolin
Yeah. Thank you. I love the kind of AI powered versus AI native versus AI first. I think it's helpful to understand that. Mahesh, can you introduce yourself a little bit as well? I think you have some, some interesting thoughts here.
- [06:22]
Mahesh Ram
Sure. Nice to, Nice to be here. And by the way, we're Zoom is a very happy product board customer. So thank you. Hoover. Thank you. So I joined Zoom in 2022 when Zoom acquired my conversational AI company Solvi, which is today known as Zoom Virtual Agent. In our contact center suite have been. I headed up the AI product initiative to launch AI Companion. Hopefully many of you use AI Companion. Now it's in 12 products at Zoom and which released in September of 2023. The last 18 months have been purely focused on taking Zoom into the generative AI age and driving generative AI across the entire platform and product suite.
- [07:05]
Rachel Wolin
So I think what's really unique about you is you have like these two lenses, you have the former founder lens and then also now you're this exec at Zoom. Um, just focus on AI. What, what do you think is different about building an AI first company versus kind of integrating AI power into an existing company. How do you think about that? And then maybe just a follow on where do you think AI is in the enterprise adoption cycle?
- [07:33]
Mahesh Ram
So I, I think to the founders in the room, I think, you know, when you're thinking about building a company based on generative AI, I think you have fundamentally unique opportunities because you're not, you don't have to deal with legacy application, legacy code. And so when we were building Solvi, nobody was talking about generative AI. I think we built the first large language model on BERT that was in, in our space ever, in 2017, 2018. And I think it really opened up the opportunity to really rethink an existing way of doing business without any antecedents, without any debt or business or technical debt. So that's a very liberating thing, I think, for founders. So for the founders in the room, I would say the number one thing I would say is completely think about the workflow you're intercepting with your AI. Don't think about the AI. So if there are 19 steps to doing something in the world today and you as a founder can use AI to take 19 and take them into 4 and save people time and money, that's what you should focus on. AI will become a way, a means for you to do that. So that's, as a founder, right, you have that unbelievable opportunity to create something from the ground up, purpose built for the problem. At Zoom, we had a different. When I came into Zoom, you know, massive business, you know, one of the most well known brands in the world, but had no gen, had very little generative AI. We were doing virtual backgrounds, we were doing other things. So Eric's mandate to me was basically like bring us into, you know, make everybody at Zoom a generative AI thinker. And that required a mindset change. But there's nobody better to do that with than Eric Wan. Because what he, first thing he said to me was, he said, he said, I don't care what it costs quality. He said, if you deliver the highest quality with the best user experience, I will take care of the cost problem for you. And so it was very liberating to hear that because then you don't have to worry like, oh, do I take the cheapest solution? Do I do that? But then you still have to bring everybody in a company of thousands and thousands of people along with this journey. And I think we didn't like immediately create a center of excellence and have everybody create this R and D middle thing. We just said, everybody is going to ship, we're going to get into a ship first mentality again, user centric, user focused, what's the flow? So think about what Zoom does. You're in a meeting, there's, what do you want to do? You want to focus on the meeting. So how do we get a note taker to be better? How do we do, you know, ask me questions if I join the meeting late, how do I in a chat channel, how do I catch up if I've been on vacation for a week? Whiteboard. Suppose I'm a bad whiteboarder. Can I just use my words to create whiteboard phone calls? You know, can I do voicemail generation? All of these things are things that we already do in our daily lives. Can we do them better, faster with generative AI on top of the platform we already had? So it's a quite different phenomenon when you're putting, when you're taking an existing business and turning it into generative AI and building something from organically generative AI. So I think that's kind of the two lenses or the two frames. So hopefully.
- [10:46]
Rachel Wolin
Yeah, I love that. I also think it's amazing to get to have like a front row seat with Eric and you know, kind of at a generational company where he is bringing that kind of like founder energy back, especially around AI. All right, Anika, so really excited to hear your perspective and maybe you can introduce yourself a little bit as well.
- [11:08]
Annika Gupta
Yeah, absolutely. Great to be here. So I'm the head of product of Rubrik. Rubrik is a data security platform and what we do is we help large organizations, especially in regulated industries and the government, recover their data and applications after they've been hit with ransomware attacks. It's a very exciting space, but unlike other companies, there's not always an obvious place where AI is going to fit into the portfolio. And how are we going to do it, given that we do work with very conservative organizations that are very risk averse and are using us in their most critical time of need, which is usually a cyber event.
- [11:49]
Rachel Wolin
Yeah, I'd love it if you could tell the audience a little bit about how you're thinking about disrupting and future proofing your core products. And then like, what do you. With AI, of course. What do you think the audience can learn from that experience? Yeah.
- [12:04]
Annika Gupta
So the way that we've approached our AI initiatives and kind of similar to what my other panelists talked about is really looking at where can we provide the most value to save people time and to deliver better experiences for them. So AI for us is not a separate product. It's actually embedded in all of the products that we have and we've looked for unique avenues for us to deliver a better experience. So you'll see on this chart we have a lot of Different AI areas that are integrated with our core product. Some of them are around how do we provide, how do we enable our customer support team to provide a better customer experience and resolve tickets and issues faster? Given the complexity of every customer's environments and the situations that they're facing, other places are looking into our own product and saying, well, if we add a AI assistant, which we call Ruby, into our product, how can we help people troubleshoot faster? How can we help IT professionals who may not be security professionals respond more confidently and quickly to cyber events and figure out what's the best next step to take? The biggest challenge in all of this has been there's a lot of different ideas, but how do we make sure. Back to Mahesh's point around quality. Quality is super, super important. People are not going to want to take recommendations around security and what's happening within their data infrastructure from an AI agent unless they're confident that what we're telling them isn't is going to help them make their, their environment better and not worse. And so there's a lot of risk associated with engaging AI and using it. So we've had to be really thoughtful around the entirety of how we design the user experience to surface information and allow people to take action, stringing together workflows without enabling them to take actions that might be destructive to, to their environment or their organizations.
- [14:04]
Hubert Palan
Can, can I, can I come in? Just like, it's interesting to hear that we are all talking about adding AI on top of a system that already has a strong value proposition and workflows. And that is a very important thing to realize because you, you can't forget that it's not just the AI magic that you still need to build the workflows and the structures that you plug the AI into. And so don't confuse, you know, AI first with AI only. And that's a, that's a big, big danger there.
- [14:36]
Annika Gupta
Yeah, I love the point. I think there's also, if you think about AI in your product, you can almost. The way we've approached it is, let's think of it, new way of interacting with the product. It's not a product in and of itself that we're monetizing or is a separate feature, but rather if you were to reimagine your product and instead of having people go between dashboards trying to figure out what to do or how to do it or what actions to take, you actually just reimagine the entire experience from what would a personalized workflow really look like? And how can I deliver value in a truly different way than what is traditionally allowed in SaaS applications?
- [15:14]
Rachel Wolin
Yeah, I love that. I feel like we think webflow about AI is just yet another developer superpower that we can give to our customers. I also think Ruby is maybe a more trusted Clippy. I kind of, I kind of like this.
- [15:30]
Annika Gupta
We use that analogy internally and some people hate that because they have bad experiences from clippy in the 90s and.
- [15:36]
Rachel Wolin
Some people love it. I mean Clippy was just ahead of their time, you know. Okay, moving on to the lightning round.
- [15:47]
Hubert Palan
Can I, can I, can I say before we go there, one other thing that I remembered since this is about advice and actionable and a multi product, right. We are actually going to launch the Insights experience as a separate product. And a big part of it is because we are a seed based product and the AI value doesn't really scale that much necessarily with the seed. I mean, it depends what, what is the use case. But if you're analyzing large volume of insight and a lot of customer conversations, it scales with the volumes. And so it's a very interesting thing to think about the packaging through the lens of the use cases. But then also, you know, I'm not suggesting to do cost based pricing or marketing professor from, from Berkeley Business school wouldn't like that. But what I'm saying is that you need to consider that when you're thinking about do I just put it into the product as a feature or do I launch it as a separate offering on its own?
- [16:41]
Rachel Wolin
Yeah, I think that's an interesting point. I think we're kind of starting to move out of this era of just AI making you more efficient and then moving into the era of AI delivering value and delivering like explicit customer value that you can price on in a different way and package differently into your product. So I think that's a really interesting point.
- [17:00]
Hubert Palan
Mahesh, you probably. This is a tough question for you guys soon, huh? How do you package it?
- [17:06]
Mahesh Ram
So we're thinking about, I mean if you think about your day, like I always think to myself, I don't get up in the morning and think, oh, my biggest initiative today is to have a meeting, right? We're having a meeting or I'm not in a, or let's get in a chat channel and chat, or let's make a phone call or all the things that you. Or let me create a whiteboard today, right? All of these things are towards a larger objective. And so when you think about what the larger objective is, your frame of reference for what the product and feature needs to do completely changes. Because if the goal is to, let's say that the project team is working on shipping a product in 60 days, there's going to be phone calls, there's going to be chats, there's going to be whiteboards, there's going to be customer conversations, there's going to be information that's coming in from outside coming. How do I keep that synthesized? How do I collaborate on that? And let's say someone joins my team a month in, how does that person catch up? Right? So if you think about the problem statement, it's very different than the product statement, right? So I think that's where AI is so powerful, right? Because what I can do is actually bring that person up to speed much faster for having missed a month of that project than any single human being on that team could do. Because I can actually go. And all that collective memory, all that knowledge can actually be harnessed by AI, can get you up to speed. I'll give you one small example. That's one of my favorite features in Zoom Team chat is I have a team in Japan that works overnight on Japanese stuff for the AI. Okay, The Japanese capabilities. They're chatting all night long. I get up in the morning and I just say, summarize this chat channel for me in English. And I immediately know what they did. And I ask questions against it. I say, what were the two things that they need my help on? And it answers my question. I can then go back to that team in English, ask, answer the questions, and they can then translate that into Japanese and get the answer from me asynchronously. That's really practical example, right? And so agentic applications will just take this even further where they'll synthesize all these products and allow me to prepare for the day. And my day involves all of these things, but with a larger objective, which is to deliver the product on time. So I think that's where we're thinking about it, is single workloads deliver a lot of value and then put it together into synthesis.
- [19:23]
Rachel Wolin
Okay, I think we have time for one question. I'm going to zoom through. Here we go. Last question. What do you see in AI that may not be obvious? Like what. What are you seeing now that you're building and you're seeing across all these different companies? You're seeing your customers problems. What do you think is not obvious right now?
- [19:45]
Annika Gupta
I think that what may not be obvious about AI is I truly think that it's going to fundamentally disrupt enterprise SaaS. I don't think that people will be content with the modes of operation and the way that you consume and interact with applications today. And you know, as we've thought about our own products, we're like, well, how do we insert this new way of operating? I actually don't think that that's going to be a side by side. I think fundamentally that's going to completely take over the way that, that SaaS applications are architected and everything is going to be a personalized experience and a personalized workflow that enables you to come in and surface the job that actually needs to get done and do it and then automate it in the future. So I actually think that's the future of enterprise SaaS as opposed to like what exists today.
- [20:35]
Mahesh Ram
I'll go next. I have a few things. So one, one is local AI. I think local AI is going to change fundamentally how we work with AI. Local AI meaning AI PC, the devices having AI capability, the ability to run small models on very small devices. What that does is, if you think about it is the initial thing is, oh great, I can do certain things on my local machine, but think of security. Maybe I just want to summarize all my emails, but I don't want to send it to the cloud. Think of like defense, security, government, all the places where a local model can be secured and can actually create value. Think about hybridizing local models with cloud models where maybe the local model can get 90% of the fidelity, but maybe it can go to the cloud to just get what it needs. So maybe it has your assets that are locally secured and can do that. As an example, we're running virtual backgrounds on local AI PC. If you're sitting in a coffee shop and running a Zoom meeting, it doubles the battery life of your machine by having that run locally. So that's a very practical application of local AI. But there's so much more coming in that area that as all of us start getting AI PCs into our homes, into our work, that's number one. Number two is I think all of us are going to have personal AI assistants. We're already starting to see that, right? With the perplexities and the gleans and everybody else trying to get into the consumer side and then we have the work ones with AI companion from Zoom or Copilot or whatever. Something has got to put this together because I wake up in the morning and I think I have to do this personally. Maybe I have to go to the bank or I have to do something Drop the kids off or do something and then I also have to do my work. How do we bring these things together? I think there's a real opportunity to bring these lives together. And the third one I'll say, which I think is Eric has said this is this idea of a digital avatar of ourselves. And I know it sounds scary, it sounds creepy, but I think if you think about some practical applications where I can have a representation of myself doing certain things. If the three of us need to schedule a meeting together, could we have our digital avatars, our digital selves actually try to coordinate, find the times, schedule a meeting, put it in our calendars, remind us and when it runs into a problem, intercept and ask us. I think there's, I think that's going to happen.
- [22:43]
Hubert Palan
This is super practical. I'm going to take us a little more philosophical and that is, I do keep things thinking about the divide that this super fast advancement in AI is creating just in general in the society and the knowledge workers. And you said, you know, we're gonna all have these assistants and that's probably for this crowd here. But what is gonna be the impact on even my ability to talk to my grandparents about what's happening in the, in the world is already limited. And so I, I do think about how, what kind of a responsibility we have as, as the players in this space to make sure that we don't like running super far away from the majority of the society and create a divide is going to be very difficult to bridge to get everybody, you know, safely home.
- [23:30]
Mahesh Ram
I just want to say one more thing though. All the founders in the room, I raised money in 2008 when the market completely crashed and it was the grimmest time to be a founder and it was really difficult. 2009 to 2012 were the best years we ever had. So if you're out there, like stay persistent, stay, stay in it. I'm telling you, like this is hard, but you, there'll be some big winners in this room. So I hope that you stay persistent and go, go hard and get what you're deserving.
- [24:00]
Hubert Palan
So plus one.
- [24:09]
Mahesh Ram
Hey everybody.
- [24:10]
Podcast Host
Thanks to the 10,000 of you who came out to SASTER Annual this year. We had a blast. And big news. We'll be back in May of 2025 or in May of next year. That's right, SAS train will be a bit earlier next year. The 13th to 15th of May 2025. We'll still be back at the same venue in the SF barrier in our 40 acre San Mateo county event. Center campus. And tickets are never cheaper than right now. So grab your tickets@saster annual.com with my code Jason50 for an extra discount on our very, very best pricing. That's Jason fiftyasterannual.com See you next Monday May at Sastra 2020.