From Hype to Habit: How We’re Building an AI-First SaaS Company—While Still Shipping the Roadmap

[upbeat music] Hello, everyone.

Welcome to our talk on building, uh, AI-first companies. I'm Rossella.

And I'm Deepsha.

We are so excited to be here with all of you today. Uh, you've probably heard the term AI-first, uh, a gazillion times already this week. And if you are in this room, chances are you're not just wondering what that means, you're also trying to figure out how to actually make it real inside your own company.

And that's what today's talk is about. We wish we could stand here and give you the talk, the one with all the answers and a crystal clear playbook, uh, for becoming an AI-first company.

But the truth is that that talk probably doesn't exist. AI transformation is really hard. It's messy, it's full of trade-offs, and it looks different for every company. At Sprout Social, we are in the messy middle of our own AI transformation.

So this talk, uh, is a candid real-time look, uh, at what it takes, uh, to lead a SaaS company into the AI era without breaking your business, uh, the team, or your values.

We'll share what's working, uh, what's not, uh, and a practical framework, uh, uh, to guide you through your own AI transformation. But first, let's break down the buzzword. What does AI-first, uh, even mean?

I feel like AI-first is like teenage sex. Everyone talks about it. [laughs] No one really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they're doing it.

Jokes aside, what does it mean to be AI-first? It's about evolving, uh, from AI features sprinkled into the product to rethinking how you plan, build, and deliver value all through an AI lens.

More than that, it's about putting AI at the center of your strategy. But most importantly, it's a mindset shift more than anything else.

Now, if you've been in tech for a while, uh, you've seen big shifts before. Cloud-first, mobile-first, DevOps, uh, and many more. But here's the difference. Those shifts, uh, disrupted one area at a time.

AI is disrupting all of them at once. It's product, it's architecture, it's people, it's process, it's ethics, it's everything. And that's what makes it so hard, but also such a game changer.

Here's the good news. Being AI-first isn't binary. It's not like flipping a switch. It's not about throwing everything out and rebuilding your company overnight. AI transformation is a multidimensional journey.

It's an evolution. Where you are on this spectrum will vary depending on your own company. But the important thing is to know where you are and to move with purpose.

AI transformation is complex, uh, but it doesn't have to feel chaotic. At Sprout Social, we found that it helps, uh, to think about it through a simple framework. At its core, becoming an AI-first company means evolving across three key dimensions.

Strategy, what you prioritize and why; ways of working, uh, how you build, uh, ship, uh, and adapt; and people, how your teams evolve, uh, and the skills that define success.

Whether you're just getting started or already in the thick of it, uh, these three areas, uh, give you a way to make sense, uh, of the work that is ahead of you.

And we hope that this simple framework can help you navigate your own AI transformation.

So now that we set the bigger picture, let's start with the first dimension, strategy.

The very first critical shift, uh, is about how to determine what to build, uh, about evolving, uh, from an AI-enhanced to an AI-first strategy, where you are reimagining, uh, what's possible today thanks to AI.

In the past, investing in AI meant asking, uh, "Where can we add intelligence, uh, to an existing experience?" It was about sprinkling AI across your product to make workflows better.

In contrast, in an AI-first company, the question becomes, "What new experiences can we deliver that weren't even possible before?"

Being AI-first means reimagining what's possible. It's about solving problems, uh, that were previously unsolvable in ways that customers may not even imagine.

Now, dreaming about the future is super fun, but the challenge is that you still have to ship features, uh, to meet your customer needs today. That's the tension every company is facing right now.

How do you ship what customer needs today while also investing in the future you know is coming? If you over-index on the present, uh, you risk falling behind and missing the moment.

If you focus only on the future, you risk disappointing customers, slowing revenue, and potentially starving innovation of resources.

That's the core of the innovator's dilemma, and AI makes it 10 times harder. You basically need the discipline of an enterprise and the curiosity and nimbleness of a startup.

But you need both to be running in parallel. It's like steering a ship and launching a rocket at the same time. So at some point, you have to figure out how to build the startup inside your own company.

I couldn't agree more, and that is what makes this so difficult. With this fast-evolving goalfo- goalpost, if you've felt this tension, you're not alone. When a three-month roadmap starts to feel stale in three week, what is possible is constantly evolving.

How do you adjust that in the roadmap? And because of that, we're realizing that we're moving farther from deterministic roadmaps where we generally knew what we were building for months in advance.

We've had to shift our mindset that embraces ambiguity, where learning and discovery are what shape the path forward. Now the destination itself can evolve as we learn more about what's possible.

And that's how we plan. It's also evolved how we design. When earlier we could build AI into features, in an AI-first interaction, customers expect seamless, intelligent systems that stretch across workflows and even roles.

Let's think through that with an example. Let's talk about Woofwell. It's a health tracking app for puppies, and here our customer is Pineapples. Pineapples uses Woofwell to track meals, uh, activities, digestive insights, yes, that's a poop log, and also track supplement recommendations.

Each of these features in our company is owned by different teams. But in an AI-first world, when Pineapples is engaging with that app in a natural language, he's not thinking in features.

He's expecting a unified value. So when Pineapples says, "I'm feeling kind of blah," a meal-based response could say, "Hey, maybe 'cause you skipped breakfast?" A digestive-based response could say, "Yeah, your poop seems a little off today."

But both a- both answers are useful, but very narrow. Now imagine a unified c- experience for Pineapples. The answer could be, "It might be the new kibble. Your activity seems to be going down over the days, and your poop also seemed to be off today.

Maybe this new supplement that you're trying, pause it and see how you feel tomorrow." That's what AI-first really means, not just smarter features with AI sprinkled in them, but the unification to generate a broader value that you unlock incrementally through your roadmap.

But that has also evolved how we build, which means how we work together, move fast, learn fast, and stay aligned through this uncertainty. Here's what those shifts look like.

Historically, innovation often happened in a reactive way. Someone had an idea, we test it in isolation, maybe do a spike to see if something worked, and then move on to the next big idea.

It was ad hoc, and while it led to moments of inspiration, it rarely became sustained strategic driver. In an AI-first world, that kind of fragmented discovery doesn't hold up.

The landscape is moving so fast with the stakes so high that we need to treat discovery as repeatable, deliberate process, and that's what ritualized discovery means. It means building time into our planning cycles for experimentation, hackathons, and learning in various forums and formats that are visible and actionable across the company strategy.

As part of ritualized discovery, we've also embraced a mindset of MVPs for learning, not just to launch quickly, but to validate direction. We assume that not every bet will land.

Some features won't work the way that we're expecting them to. But in this case, failure is a feature, not a bug. It's what drives clarity through ambiguity.

We've also had to rethink what processes are for. Many of the processes we've built in the past were designed for predictable linear roadmaps with deterministic risks. Uh, it was a world where we generally knew what we were building months in advance.

But in an AI-first world where timelines shift, capabilities are evolving monthly, those old processes don't always hold up. At the same time, if you introduce too many new processes all at once, it can backfire.

Instead of helping, they slow us down. They become overhead. They create drag, and a few folks here from Sprout will confirm that at Sprout we've learned that the hard way.

That's why we've started treating process as a product. We evaluate it against outcomes. Does it create clarity of direction? Does it unblock teams? Does it help us make better decisions faster?

And if the answer is no, we have to reiterate or cut it entirely. Process isn't a dirty word. When it's purposeful, it gives teams the clarity and flow and becomes a process that accelerates.

Now, we heard in the amazing keynotes yesterday that execution is the moat. Speed is paramount to delivering against the evolving landscape of tech and the customer expectations. But if you move fast without clarity of direction, you end up with chaos.

And if you have clarity of direction but no momentum, you stagnate. How do you find that balance? Because in AI world speed matters but only when it's paired with direction, we've had to move from speed to smart velocity, which means building the muscle to move fast with purpose.

When we talk about prototyping quickly, building MVPs and speed, we're not just saying ship it for the sake of shipping it. We're talking about working with clarity, momentum, and adaptability.

Smart velocity is what keeps us grounded and moving all at the same time.

And now let's talk about the most important piece of it all, the people.

If strategy sets the direction and ways of working determines, uh, how we execute, the people are what makes the whole thing real. After all, culture eats strategy for breakfast, right?

And here's the truth. Becoming an AI-first, uh, company isn't just tech transformation. It's mainly a cultural transformation. And culture lives in people, how they think, how they work, how they lead, and how they feel.

That means that we need to rethink, uh, what great talent looks like, uh, in the AI era, not just in your AI team, but across the entire company. At Sprout Social, we are seeing two major shifts: how talent is evolving in this AI era and how to scale AI fluency across organizat- excuse me, the organization.

So let's start with the first one. As the world changes, uh, so do the capabilities that matter most. To be clear, the talent that we need today isn't replacing what we had yesterday.

It's just building on it. Until now, what made an AI practitioner great was deep specialization. Being a generalist or a visionary builder gave you an edge, but for most roles, those skills weren't critical.

Today, that is changing. AI depth is still essential, but now being a versatile visionary builder has become critical to success. That's why we're investing in T-shaped talent, people with deep expertise who can also stretch wide, prototype quickly, collaborate fluidly across silos, and bring

end-to-end systems to life. It's a bit like Indiana Jones. We're not choosing the professor and the adventurer. We are combining them, bringing the deep specialization of the scholar into the jungle to uncover unimaginable possibilities.

And that's why more than ever, we need people who can navigate ambiguity. There's no playbook. The technology's evolving fast. The roadmap keeps shifting. That means we need pathfinders, people who can hold to their deep expertise while forging new ways through shifting terrain.

And in most cases, the path doesn't even exist, so they have to imagine it first. And that's where visionary thinking becomes a core skill.

So this shift from shaping what's known to charting what's newly possible is what we believe defines the next generation of AI builders.

But evolving our builders, uh, is only part of the story. Becoming AI-first isn't just about the few who train models or build agents. It's about building fluency across the whole organization.

You can't go AI-first if the rest of the company is still AI-last.

So in an AI-first company, everyone touches AI, even if they're not training models or building agents. From marketers and designers to PM and care agents, we want every team to feel empowered to understand AI and confident enough to build with it.

And that's why we're investing in org-wide AI fluency to inspire people of what's possible and to make AI feel usable, safe, and real in the context of their day-to-day work.

We're supporting this enablement through things like AI newsletters, podcasts, cross-functional AI show-and-tell, and by empowering teams to use whatever AI tools help them work smarter. But fluency isn't enough.

It has to scale. To do that, we aim to build self-service pla- platform to enable product and engineering teams across the organization to prototype and ship AI-powered features without needing deep involvement from the AI team.

The goal is not to turn everyone into an AI expert. It is to create a company where AI thinking and exploration is the default, not the exception. [coughs]

Sorry.

You're good. [laughs]

Sorry. [coughs] All right. Now, we've talked about a lot of the things that have changed, and I know that's very overwhelming, but there are also some things that haven't changed, the fundamentals.

Our best AI features still solve customer problems. They're rooted in needs, not the novelty of AI, as long as they're solving customer problems. User experience, performance, reliability, trust, these are still non-negotiable in the AI-first world.

Human creativity, human judgment, and human care remains central to how we lead, how we make decisions, and how we show up for our teams and customers. So if you're leading AI transformation in your company, get honest about the trade-offs.

Invest in people, not just models and agents. Kill the roadmap if needed. Be bold. Learn out loud. And ship that weird idea.

We hope this talk served as the flashlight in the maze, and you leave knowing that becoming AI-first won't be a linear path, and it won't be perfect. But the good news is you don't need all the answers to get started.

You just need the right questions and the conviction to evolve.

We're gonna leave you with this one re- with this last reflection. The most transformative inventions in human history only became truly revolutionary when they became part of our day-to-day life.

And we're now standing at the cusp of the next greatest revolution. This is a profound privilege and an honor. Not only are we witnessing this moment, but we have the opportunity and responsibility to shape it, to guide our teams, our companies, and our communities through a transformation that will redefine how we live, how we work, and how

we connect. This isn't just a technological revolution. It's a human one. This transformation isn't easy. It takes time, grit, and the patience to navigate setbacks, doubts, and growing backs.

In fact, this is how we look like when we started the journey. [laughs]

And now, well, let's just say we learned a lot. [laughs]

So if, if it, if it feels like a long road ahead, you're not alone, but we promise it's totally worth it. Thank you so much for spending time with us today. [audience applauding] [upbeat music]