About George
George Bonaci is VP of Growth at Ramp, one of the fastest-growing B2B companies in the US. He studied biochemistry and spent years in analytical chemistry and synthetic biology research before moving into growth marketing. At Ramp, his team is responsible for pipeline and SQLs, covering performance marketing, SEO, lifecycle, and events. Growth reports into the CTO, a structural choice that says a lot about how Ramp thinks about distribution.
🎙️ Listen on YouTube, Apple Podcast, Spotify
Most growth teams run experiments. Very few run them the way George Bonaci does at Ramp: with a scientific mindset, a tolerance for nine consecutive failures, and the willingness to test incentives directly on LLMs.
In this episode, we went deep on how Ramp structures growth, why time to results is the most underrated variable in any experiment, and what the most overlooked shift in marketing looks like right now.
Here’s what I took away:
1) Growth reports to the CTO. Because everyone at Ramp is a builder.
Most companies have growth inside marketing. At Ramp, it sits inside a technical organization. The reason is philosophical: everyone is expected to build, regardless of function. The practical benefit is direct access to engineering resources at a moment when distribution has become more valuable than product development speed. George joined because of that philosophy. It’s also why he hires scientists, engineers, and finance people onto growth teams.
2) The ICE framework has two missing variables: time to results and confidence.
Impact, confidence, ease. George adds time to results because at Ramp’s velocity, a four-month experiment is a bad experiment. He needs to know something in two weeks. Confidence matters because the people who have run a lot of experiments develop an intuition for what will and won’t work that they can’t always put into words. That gut feel is worth something. Hiring people who are right a lot, even when they can’t explain why, is part of the job.
3) When something works, go 10x. Find the asymptote fast.
Most teams that find a working channel do 50% more. George’s approach is to saturate as quickly as possible. A working channel is a temporary competitive advantage and should be maximized immediately. Wasting a little money finding the ceiling in a week is better than testing your way there over quarters. The ROI of speed is consistently underestimated.
4) Ramp tested incentives on LLMs. It worked.
The hypothesis: if you can offer incentives to humans to take a meeting or try a product, why not offer incentives to machines? Ramp ran the experiment, published the results, and found that Claude picks it up clearly, ChatGPT does not, and everything in between has its own nuance. It helped AEO visibility, employer brand, and even organic SEO as a side effect. George’s point is that marketing to machines is already happening and almost nobody is acting on it yet.
5) Ramp’s data moat is the content moat.
Ramp processes a significant share of US business spend. That data, aggregated and anonymized, is something no competitor can replicate. The content strategy is built around it. For companies without a proprietary data set, George sees three options in order of difficulty: find something unique in your customer base, run original research, or go structure publicly available government data that exists but is impossible for an LLM to crawl and digest. Most companies have more interesting data than they think. They just haven’t looked.
6) George pivoted his SEO team almost entirely to AEO.
Some run-rate SEO work continues, but the direction is clear. AEO and SEO overlap enough that the transition is not a full restart, but the bets are going on AEO. Google’s AI mode and AI overviews have validated that direction. George tells his team he doesn’t care about missing numbers this week or this month if it sets them up for a year from now.
7) The role of VP of Growth may not exist in five years.
George brought this up with his boss. His boss did not disagree. The prediction: very flat organizations where everyone is a builder and an agentic operator. Professional managers coordinating large teams through layers of hierarchy will be replaced by people who can build systems, manage fleets of agents, and provide value as individual contributors. For anyone preparing for a growth role in 2030, his advice is simple: go build something.
Find George on LinkedIn : https://www.linkedin.com/in/georgebonaci/
Enjoy,
— Jordan
