I’m going to say something that might sound crazy. Traditional SaaS is dead.
There, I said it. The playbook that built Salesforce, Hubspot, and a thousand other unicorns is being torn up and rewritten. The culprit? Artificial intelligence. Not the buzzword you see in every pitch deck, but real, embedded AI that’s fundamentally changing how software is built, sold, and used.
I’ve seen this movie before. I saw it when the cloud washed away on-premise software. I saw it when mobile ate the web. And I’m seeing it again with AI. The difference is, this wave is bigger and moving faster than anything I’ve ever witnessed. As an investor in companies like Anthropic, OpenAI, Scale AI, and Hugging Face, I’ve had a front-row seat to this revolution. As a founder who has built and sold two companies, I’ve lived the painful transition from one paradigm to the next.
This isn’t just about adding an “AI-powered” badge to your website. This is about a new breed of company: the AI-First SaaS. These companies are built differently, they go to market differently, and they have a different relationship with their customers. And if you’re a founder today, you need to understand this new world, or you’re going to be left behind.
The Old Playbook? It’s a Relic.
I remember when we were building RemoteTeam. We were a classic SaaS company. We had a great product, a slick marketing site, and a team of hungry sales reps. We followed the playbook to a T. We raised money, hired a VP of Sales, and built a predictable revenue machine. It worked. We were acquired by Gusto.
But that playbook is starting to show its age. The problem is that it’s built for a world where software is a tool. You buy a tool, you learn how to use it, and you get value from it. But with AI-First SaaS, the software isn’t just a tool. It’s a partner. It learns, it adapts, and it delivers value in ways that are often unpredictable.
Think about it. How do you sell a product that’s constantly evolving? How do you price a product that delivers non-linear value? How do you market a product that’s different for every single user? The old playbook just doesn’t have the answers.
The New Playbook: The AI-First Go-To-Market
So what does the new playbook look like? It’s still being written, but I’ve seen enough to know what the key chapters are. It’s a combination of product-led growth on steroids, usage-based pricing, and a platform-centric approach.
Product-Led Growth on Steroids
Product-led growth (PLG) has been around for a while, but AI-First companies are taking it to a whole new level. In the old world, PLG was about letting users try before they buy. In the new world, it’s about letting the product sell itself.
Think about how you first used ChatGPT. You didn’t get a demo from a sales rep. You didn’t read a whitepaper. You just started typing. And the more you used it, the more you realized how powerful it was. That’s PLG on steroids. The product is the marketing. The product is the sales.
For founders, this means you need to be obsessed with the user experience. You need to make it ridiculously easy for users to get started and to experience the magic of your product. The “time to value” needs to be measured in seconds, not days or weeks.
The Inevitability of Usage-Based Pricing
If you’re building an AI-First SaaS, you’re probably going to end up with usage-based pricing. It’s the only model that makes sense when the value you’re delivering is directly tied to consumption. I’ve seen this with so many of my portfolio companies. They start with a traditional subscription model, but they all eventually move to usage-based pricing. It just aligns incentives better.
Look at OpenAI. Their pricing is based on tokens. The more you use, the more you pay. It’s simple, it’s fair, and it scales. Scale AI is another great example. They charge per task. The more data you need labeled, the more you pay.
But usage-based pricing is hard to get right. You need to be able to measure usage accurately. You need to be able to predict and control costs. And you need to be able to communicate your pricing clearly to customers. It’s a whole new set of muscles that most SaaS companies haven’t had to build.
The Rise of the AI API
Another key difference with AI-First companies is that they are often platforms, not just products. They expose their core AI capabilities through an API, which allows other developers to build on top of them. This is a massive shift in how software is built and distributed.
Hugging Face is the poster child for this. They’re not just a collection of models. They’re a platform for the entire machine learning community. Developers can use their APIs to access state-of-the-art models, to train their own models, and to deploy them in production. They’ve become the GitHub of machine learning.
Building a platform is a lot harder than building a product. It requires a different mindset. You need to think about developers as your customers. You need to provide great documentation, SDKs, and support. But if you can pull it off, the rewards are immense. You can create a whole ecosystem around your company.
Don't Reinvent the Wheel: Leverage Cloud AI Services
One of the biggest advantages that AI-First founders have today is that they don’t have to build everything from scratch. The cloud providers have invested billions of dollars in building world-class AI services. Whether it’s speech-to-text, computer vision, or natural language processing, there’s probably an API for it.
When we were starting out, we had to build our own machine learning models. It was slow, expensive, and painful. Today, you can get access to state-of-the-art AI with a few lines of code. This is a huge accelerator. It allows you to focus on what’s unique to your business, rather than reinventing the wheel.
Building an AI-First Team
So what kind of team do you need to build an AI-First SaaS? It’s a mix of old and new. You still need great engineers, designers, and product managers. But you also need a new set of skills.
You need people who understand data. You need people who can work with machine learning models. You need people who can think in terms of probabilities, not just certainties. And you need people who are comfortable with ambiguity and experimentation.
Finding these people is hard. They’re in high demand. But they’re the ones who are going to build the next generation of great software companies.
The Future is AI-First
I’m more excited about the future of software than I’ve ever been. The shift to AI-First is going to unlock a whole new wave of innovation. We’re going to see products that are more intelligent, more personalized, and more helpful than anything we can imagine today.
But it’s not going to be easy. There are going to be a lot of challenges along the way. But for the founders who are willing to embrace this new world, the opportunity is massive. The next Salesforce, the next Google, the next Apple—it’s going to be an AI-First company. And I, for one, can’t wait to see what you build.
The Ghost of Sales Teams Past
I remember the exact moment I knew the old way was dying. I was talking to a founder of a hot SaaS startup. They had just raised a massive Series B. I asked him what his top priority was. I was expecting him to say “product” or “engineering”. Instead, he said “hiring our first 20 sales reps.” I almost fell out of my chair.
Twenty sales reps? For a product that was supposed to be self-serve? It just didn’t compute. But that was the playbook. You raise money, you hire sales reps, you burn cash, and you hope that you can acquire customers fast enough to outrun your burn rate. It was a game of brute force. And it worked, for a while. But it’s a terribly inefficient way to build a business.
Think about the cost. A good enterprise sales rep in Silicon Valley can cost you $300,000 a year, fully loaded. Twenty of them? That’s $6 million a year. And that’s before you even factor in the cost of marketing, support, and all the other things you need to make a sales team successful. It’s a huge bet. And it’s a bet that’s getting riskier every day.
The AI-First Flywheel
The beauty of the AI-First model is that it creates a virtuous cycle. A flywheel. It goes something like this:
- You build a great product with a magical user experience. This is the core of everything. Without a great product, nothing else matters.
- Users discover your product and start using it. This is where PLG comes in. You make it as easy as possible for users to get started.
- The product delivers value, and users get hooked. The more they use it, the more value they get. This is the magic of AI. The product is constantly learning and adapting.
- Users start paying for the product. This is where usage-based pricing comes in. The pricing is directly tied to the value that users are getting.
- You reinvest the revenue back into the product. This makes the product even better, which attracts more users, and the flywheel starts spinning faster and faster.
This is a much more efficient and sustainable way to build a business. It’s not about brute force. It’s about building a product that’s so good, it sells itself.
The New Moats are Made of Data
In the old world, the moats were things like brand, distribution, and sales teams. In the new world, the moats are made of data. The more data you have, the better your AI models become. The better your models become, the better your product becomes. The better your product becomes, the more users you get. The more users you get, the more data you have. It’s a self-reinforcing loop.
This is why companies like Google and Meta are so dominant. They have more data than anyone else. And they use that data to build incredible products that are almost impossible to compete with.
But you don’t have to be Google to build a data moat. You just need to be smart about the data you’re collecting. You need to think about what data is unique to your business. And you need to build a product that gets better with every new user and every new data point.
The Ethical Tightrope
Of course, with great power comes great responsibility. The rise of AI also brings a new set of ethical challenges. How do we ensure that our AI models are fair and unbiased? How do we protect user privacy? How do we deal with the societal impact of automation?
These are not easy questions. And I don’t have all the answers. But I do know that we, as founders and investors, have a responsibility to think about these issues from day one. We can’t just build things and hope for the best. We need to be proactive. We need to build ethics into our products and our companies from the very beginning.
I’m an optimist. I believe that AI has the potential to solve some of the world’s most pressing problems. But it’s not a given. It’s up to us to build the future we want to live in. A future where AI is a force for good. A future where technology serves humanity, and not the other way around.
So, yes, the old SaaS playbook is dead. But the new one is just starting to be written. And I, for one, am excited to be a part of it. Let’s go build something amazing.
Frequently Asked Questions
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
How has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.