DEV Community: Dan Coulter

Why Vertical SaaS Companies Will Win the AI Race

Dan Coulter — Tue, 15 Jul 2025 19:58:53 +0000

As AI transforms the enterprise software landscape, a fascinating paradox is emerging: the companies best positioned to win may not be the ones with the most sophisticated AI technology, but rather those with the deepest understanding of specific business domains. This insight, shared in a recent conversation between a16z General Partner Martin Casado and Box CEO Aaron Levie, challenges conventional wisdom about AI's impact on vertical SaaS.

The Domain Knowledge Moat

The conversation reveals a crucial realisation about vertical SaaS companies that many in the tech industry have historically underestimated:

"I mean, I would say most vertical SaaS companies I see, the technology's trivial. But the understanding of the domain itself... It's not about the technology. It's the fact that somebody else has figured out the business all of that works. Like they have 10 people from the farm industry that is like sitting next to the engineer and be like, this is how you should do the clinical trial workflow. And that becomes so much of the IP."

This observation highlights a fundamental shift in how we should evaluate software companies in the AI era. Whilst horizontal AI platforms might seem more impressive from a pure technology standpoint, vertical players possess something far more valuable: contextual intelligence about specific industries and workflows.

Why Vertical Players Have the AI Advantage

The domain expertise that vertical SaaS companies have accumulated over years of working within specific industries becomes their competitive moat when building AI agents. As Casado notes:

"Now, that translates fine to agents, but I still would then bet on that vertical player doing that as opposed to somebody prompts their way into chat to BT to build a FDA compliance agent. I would still largely bet on compliance agent.ai to do that over the pure horizontal system that has no particular domain kind of expertise for that."

This suggests that in the race to build industry-specific AI agents, companies with deep vertical knowledge will outperform generalist AI platforms, even if those platforms have superior underlying technology.

The Persistent Need for User Interfaces

Another compelling insight from the discussion challenges the popular narrative that AI agents will completely replace traditional software interfaces:

"I still think that there's a relationship between some amount of GUI and the agent and the APIs, because again, like you don't want it every day of your life go to a blank empty screen and say, what's our revenue today? You just want a dashboard at some point and just shows you the revenue."

This observation suggests that the future of enterprise software won't be purely conversational. Instead, we'll see a hybrid model where:

Structured queries are pre-built into dashboards and interfaces
Ad-hoc requests are handled through AI agents
Routine information is displayed through traditional UI elements

The implications for vertical SaaS companies are significant: they don't need to abandon their existing UI/UX investments to embrace AI. Instead, they can strategically integrate AI capabilities whilst maintaining the familiar interfaces their users rely on.

AI as a Decision-Making Partner

The conversation also explores how AI is beginning to influence executive decision-making processes. Levie shares a particularly striking example:

"I just spoke with a very, very legit company, household name... We're at the board level for every decision they ask the AI for, like, basically, more information for the decision... this founder was telling me it's literally better than half of my board members."

This anecdote illustrates how AI is moving beyond operational tasks to strategic decision support. For vertical SaaS companies, this presents an opportunity to position their AI capabilities not just as workflow automation tools, but as strategic advisers that understand the nuances of their specific industries.

Practical Implementation: The Box Example

Levie provides a concrete example of how Box uses AI internally for earnings preparation:

"I already use it for, let's say, our earnings calls, where we'll do a draft of the initial earnings script... I just load up the earnings script and I'll use a better model and say, give me 10 points that analysts are going to ask about this. And like, how would I improve the script?"

This use case demonstrates the practical value of AI in enterprise contexts: not replacing human judgement, but augmenting it with comprehensive analysis based on historical data and pattern recognition.

Strategic Implications for Vertical SaaS Companies

Based on these insights, several strategic principles emerge for vertical SaaS companies looking to integrate AI:

1. Leverage Your Domain Expertise

Your deep understanding of industry-specific workflows, regulations, and best practices is your primary competitive advantage. Use this knowledge to build AI agents that understand context that horizontal platforms cannot easily replicate.

2. Don't Abandon Your UI/UX Investments

The future likely involves a hybrid approach where traditional interfaces coexist with AI agents. Continue investing in user experience whilst strategically integrating AI capabilities.

3. Position AI as Strategic Support

Consider how your AI capabilities can move beyond operational efficiency to provide strategic insights and decision support that leverages your industry expertise.

4. Focus on Contextual Intelligence

Rather than competing on raw AI technology, focus on building AI systems that understand the specific nuances, regulations, and workflows of your vertical market.

The Path Forward

The conversation between Casado and Levie suggests that the AI revolution in enterprise software won't be won by the companies with the most advanced neural networks, but by those who can most effectively combine AI capabilities with deep domain knowledge. For vertical SaaS companies, this represents a significant opportunity to strengthen their market position by leveraging their existing industry expertise as the foundation for AI-powered solutions.

As we move forward, the companies that will thrive are those that recognise AI not as a replacement for domain expertise, but as a powerful amplifier of it. The future belongs to vertical SaaS companies that can build AI agents which don't just understand technology, but truly understand business.

This analysis is based on insights from the a16z podcast featuring Martin Casado and Aaron Levie, discussing how AI is transforming enterprise software and the strategic advantages of vertical market focus.

How F1 Thinking Can Improve Your Software Development

Dan Coulter — Sun, 06 Jul 2025 15:09:49 +0000

With the Silverstone Grand Prix this weekend, I've been thinking about how Formula 1 and software development share more in common than you might expect. As someone who's spent just as much time debugging production issues as watching Verstappen defend into Copse, I've started to notice some patterns-especially around how elite performance is built and sustained over time.

This isn't a direct metaphor-your dev team isn't changing tyres in two seconds or hitting apexes at 300km/h-but the mindset behind a high-performing F1 team offers useful lessons for building and shipping software that lasts.

Thinking Beyond the Next Race: Building for the Season

In F1, success isn't about a single win, it's about delivering consistently across an entire championship. Similarly, in software, the best teams don't just optimise for short-term feature velocity. They build systems that are resilient, maintainable, and scalable over the long term.

A great current example of this mindset is James Vowles, Team Principal at Williams Racing. He's taken on the challenge of turning around one of the most historic teams in motorsport-not by chasing flashy results, but by systematically improving the underlying culture, technology, and operational structure. He's focused on getting the fundamentals right: data systems, leadership structure, internal tooling-things that don't win races overnight, but that are crucial for sustained success.

That philosophy applies directly to software teams. The best ones aren't necessarily the fastest out of the gate-they're the ones that build a foundation to scale, iterate, and adapt.

Refactor regularly rather than relying on huge rewrites
Invest in dev tooling and CI/CD pipelines
Prioritise team health and onboarding, not just deadlines
Choose stable frameworks (like Laravel) with predictable release cycles
Build processes that survive scale and team changes

Clean Pit Stops: Tight, Predictable Deployments

A well-drilled F1 pit crew isn't just fast-they're predictable. Everyone knows their role, the process is rehearsed, and fallback plans are in place. You want that same confidence in your deployment pipeline.

You're not aiming for a sub-three-second deploy-but you are aiming for something that:

Has a clear checklist
Can be repeated with minimal variation
Surfaces issues early
Has a clear rollback path
Involves the right people at the right time

Data That Drives Action

F1 teams run on data-but more importantly, on useful data. They don't just collect telemetry for the sake of it; they use it to make decisions between sessions, on strategy, and even mid-race.

In software, we often set up logging and observability tooling, but don't always take the time to refine it. What metrics actually matter?

Metric	Why It Matters
Request response time	Directly impacts user experience
Error rate (especially 5xx)	Reflects system stability
Deployment failure rate	Indicates quality of release processes
Slow DB queries	Common root cause of performance issues
Background job queue time	Affects async UX like emails, notifications, etc.

Handling Incidents Like a Safety Car

In F1, when things go wrong, a safety car or red flag forces teams to adjust quickly. Some use it as a chance to regroup or pivot strategies. In software, production incidents serve a similar role: they test how well your team communicates and responds under pressure.

Assign an incident commander who leads resolution
Use a dedicated channel for communication
Have clear escalation paths and rollback criteria
Do a post-incident review focused on process, not blame

Handled well, an incident becomes an accelerant for improvement, not just disruption.

Communication Under Pressure

During an F1 race, you'll hear engineers communicate with intense clarity-no waffle, no delay, just relevant information at the right time. The goal isn't brevity; it's utility.

We should aim for the same, especially during deployments or incidents:

Establish pre-defined roles for who communicates what
Use tools like Laravel's Notification system or Slack bots to automate alerts
Create visibility dashboards instead of chasing updates
Appoint someone to act as "race engineer"-bridging the business and technical side

Strong communication isn't just a nice-to-have-it's critical infrastructure.

Marginal Gains: Continuous Delivery Over Big Bangs

Championships are rarely won by a single bold move. They're won by stacking small improvements, week after week. That's the heart of the "marginal gains" philosophy you see in top-tier engineering cultures-and in F1 paddocks too.

For software teams, that might look like:

Upgrading Laravel versions incrementally
Optimising queries one controller at a time
Refactoring tests during code reviews
Improving onboarding docs after every new hire
Reducing friction in the CI/CD pipeline with each sprint

These 1% improvements compound. They don't slow you down-they're what keep you in the race.

Final Thoughts: The Championship Mentality

Whether you're tuning an F1 car for Silverstone or deploying a Laravel app, the mindset that delivers results is the same:

Play the long game
Build strong foundations
Learn from failure
Prioritise clarity
Never stop improving

James Vowles didn't walk into Williams expecting instant results. He's focused on culture, infrastructure, and sustainable progress. That's the exact mindset any engineering leader should embrace.

So as the F1 circus rolls into Silverstone this weekend, it's worth reflecting on how those principles play out in your own team. Are you building just for the next sprint-or for the next season?

Because in software, as in racing, the difference between finishing and winning often comes down to the systems you build behind the scenes.