Legendary investor Keith Rabois called Ramp "the best run private company on the planet," on an episode of 20VC. I had to know more. Specifically, I wanted to know how go-to-market is being architected and run so I hosted Nate Follen, Head of Business Systems Operations at Ramp to talk about AI, system architecture and building for super hyper growth.
In this conversation, we explored how Ramp builds and scales its revenue infrastructure at breakneck speed while maintaining world-class operational excellence. What emerged were invaluable insights about building scalable revenue systems, leveraging AI effectively, and fostering a culture of rapid experimentation.
The Secret to Speed: Small, Empowered Teams
One of the most counterintuitive insights from our discussion was Ramp's approach to team structure. While most companies build large centralized teams as they scale, Ramp operates with remarkably small, highly autonomous pods.
"Our team right now on the go-to-market system side is five people, including MarTech," Nate shared. "Our mandate as a company is to try to save our customers time and money. Headcount is a big cost, really. If we can build enterprise-ready scalable systems with a small team that's very accountable for what they build, that's great."
This approach echoes findings from organizational psychology research on team dynamics. Studies have shown that smaller teams (typically 5-7 people) tend to be more agile and innovative than larger ones. The key is what psychologists call "psychological ownership" - when team members feel direct responsibility for outcomes. This is also the approach that product and engineering use at Ramp.
The Power of Funnel Alignment
Most Rev Ops teams have folks dedicated by function: sales, marketing, CS etc. Ramp takes an even more granular approach that I found fascinating: they align their systems teams to specific parts of the revenue funnel and job function.
"There's really one person focused on their micro workflows and workflows that they need to work to do their jobs. That really helps focus all the efforts to being more like a product owner, a product manager, where you're actually dealing with vendors or different SaaS tools that may be external, but you're ultimately the one responsible for how the AE or BDR is efficient in their role and choosing the project work that will make them more efficient."
This creates what psychologists call "empathy through immersion" - deep understanding that can only come from being embedded in the day-to-day reality of your stakeholders. The result? Better solutions built faster.
A Framework for Build vs. Buy Decisions
One of the most valuable insights Nate shared was his evolved thinking on the eternal build vs. buy question. Rather than treating it as a binary choice, Ramp often pursues both paths simultaneously:
This approach leverages what economists call "revealed preference" - the idea that actual behavior is more reliable than stated preferences. Instead of trying to predict what will work best, Ramp lets real usage data guide their decisions.
I’m not sure how applicable this approach is for ‘normal’ companies that don’t have this type of funding and aren’t growing at Ramp’s rates but it was a really interesting approach nonetheless. I think there are lessons we can draw from this very first principles rooted approach.
The AI Revolution in Revenue Operations
Perhaps most exciting was our discussion about how AI is transforming revenue operations. Ramp is pioneering several innovative use cases:
1. Context-Aware Deal Analysis
Rather than relying on simple win/loss reasons, Ramp uses AI to analyze the full context window of deals - examining patterns across calls, emails, and actions to surface deeper insights. As Nate was describing this, all I could think was “Oh man. I want that.”
2. Intelligent Handoffs
Their systems automatically generate comprehensive deal dossiers for customer success, extracting key information about stakeholders, product interests, and critical discussion points from all customer interactions.
3. Automated Field Population
Instead of asking reps to validate every AI-suggested field update, Ramp found higher quality data by allowing automatic updates with manual override options with Momentum.io. We’re both very early customers of Momentum so it was really interesting to unpack a number of their use cases. I’ve already borrowed some of Nate’s ideas here.
Looking Ahead
The emergence of more sophisticated AI reasoning capabilities has Nate excited about the future: "If we talk again in the next year, it will be a whole new world."
I share his optimism. As AI gets better at handling the administrative aspects of sales, we may see a renaissance in true sales craft. Rather than replacing salespeople, AI could free them to focus on what matters most - building relationships and solving customer problems.
The future of revenue operations will belong to organizations that can blend human creativity with AI-powered efficiency. Based on my conversation with Nate, Ramp appears to be writing the playbook for how to do exactly that.
p.s. If you love podcasts like I do (my Spotify Wrapped said I listened to 40,000+ minutes last year) then you should try Snipd. It’s a podcast app with AI notes and clips. You can set it up so that if you double click your AirPod, it takes a “Snip” so when you’re listening and hear something you want to remember, you double click and it marks that spot and gives you a summary with the quote. I have it setup so that they go into a Notion database so I can pick stuff up from there and write about it.
It’s my favourite new product in a long time. I messaged the founder to tell him how much I loved it and he sent me an affiliate link that gives people a free month. I’m not even getting paid on it which seems like a miss 🤣 but I just want them to be successful so the product keeps getting better.
Want more content from Topline? Check out Topline podcast with Sam Jacobs, AJ Bruno, and Asad Zam