Meet Rahul: The Marketer Who Got Banned Before He Learned to Build Content Engines
We sat down with Rahul Bathija, a former Divisional marketer, to talk about why most founders start their content engine in the wrong place, what five years at Figma taught him about building content that compounds, and how he's applying the same playbook at Greptile.
Most founders assume their first growth problem is a channel problem. Rahul Bathija spent a decade learning it's actually a sequencing problem.
Rahul is Head of Growth at Greptile (YC W24), an AI-powered code review platform that recently closed a $25M Series A led by Benchmark. Before that, he spent five years at Figma during one of SaaS's most compressed hypergrowth phases, building the content and SEO engine that helped the company expand from core UI designers into the entire design market. His earlier reps came from Divisional, where he trained across paid, email, and landing page optimization with real clients.
"If you have design and you have engineering, marketing is the last step and understanding that art of distribution is quite fascinating."
That curiosity has shaped everything he's built since. In this conversation, Rahul walks through the content philosophy he developed at Figma, how he thinks about finding the 'aha' moment before choosing a channel, and what changes when AI compresses a five-year growth playbook into one.
How Getting Banned on Instagram Put Rahul on the Path to Figma
Rahul didn't plan to become a content-led growth specialist. He stumbled into it by failing at something much simpler.
In his second year at Western University, he tried to sell soccer keychains online by spamming every soccer hashtag on Instagram. His account got banned in a week. That failure cracked something open. He realized marketing wasn't "put up a billboard and hope it works." There was actual craft behind it, with real frameworks and real leverage, and he wanted to understand how to best apply it.
That curiosity led him to Divisional, where he spent a year working across paid ads, email sequences, and landing page tests with real clients. The work was deliberately varied. Some experiments worked, many failed, but each one sharpened his understanding of what users actually cared about and how to move them.
That foundation sent him cold-emailing his way into Silicon Valley. He landed a conversation with someone at Figma who brought him on board. The team was small (four people on the growth side), and the problem in front of them was one Rahul would keep coming back to throughout his career: a cold-start problem.
New Figma users landed on a blank canvas and didn't know what to do with it. The product was powerful, but that power wasn't obvious until someone showed you what you could make with it. Rahul's answer was to find the specific things that people wanted to create and build direct paths to those outcomes. That instinct, to isolate the 'aha' moment first and build backward, became the through-line connecting everything he's built since.
Why the First Content Decision at Greptile Had Nothing to Do with Channels
The lesson Rahul took from Figma wasn't about any particular channel. It was about the order of operations.
At Figma, the team identified 10 use cases, including wireframing, prototyping, and high-fidelity app design. They built individual templates and landing pages around each one. Most didn't produce meaningful results.
"Of those 10, it was only 3 that drove the majority of the results that gave us the conviction to go forward. And when we scaled that up to 300, it was actually 30 that scaled it."
The wireframe kit was the clearest proof. Someone searching for wireframing help would land on a Figma page, find a ready-made template, and get activated in the product immediately. No blank canvas required. The template solved the cold-start problem. The landing page made it findable. The organic traffic compounded.
Once the wireframing funnel proved out, Rahul built a top-of-funnel wireframing article above it. Then he connected both into a broader UI design bucket. Over five years, the concentric circles kept expanding: from core use cases to adjacent topics like color theory, palettes, and typography. The keyword "gold color" still generates 50,000–60,000 monthly visits for Figma today. A long way from wireframing, but the same underlying logic. Start at the most specific, highest-converting 'aha' moment. Exhaust it. Then expand outward, one circle at a time.
At Greptile, Rahul applied the same logic from day one. He built an examples page showing real Greptile catches in open-source repos, with the full back-and-forth between the AI and actual engineers on the record. In one case, Greptile flagged an issue in an Nvidia repo that their engineer pushed back on. Greptile clearly documented the problem and held firm, which prompted the engineer to review again, acknowledge the mistake, and fix it.
The content requires no pitch or claim. It just shows the product doing its job.
"When any discerning engineer sees that, it's almost the same proxy as showing the product in action and getting someone to the 'aha' moment as quickly as possible."
From there, he built upward. He knew exactly who the audience was.
"We often say our ICP is somebody who reads the Anthropic changelog."
The next layer was the State of AI Coding Report: a data-driven look at how engineering teams are shifting because of AI. Stats on PR volume, code quality, market consolidation. Content their audience was already looking for, not content designed to sell them something. It landed on Hacker News. Their ICP shared it.
Bottom of funnel: the 'aha' moment, documented. Top of funnel: the content their audience actually wants. Both layers go into a content brief before any execution begins. The brief maps out design, content, engineering, and distribution channels together.
"It starts with a really good brief. That was my biggest learning from Figma: how do you write a clear explanation for why you're doing it and what are the components."
At Greptile, the goal is to get from idea to published output inside a week. AI has helped collapse the hardest bottlenecks. Rahul uses Claude Code and Cursor for data scraping and site-building, work that used to require full engineering sprints. The Repositories project, a real-time index of trending open-source repos, got shipped over the winter holidays. What would have taken months now takes days.
Design judgment and nuanced copy still need human craft. But freeing up engineering time means the team can spend their energy on what actually differentiates the work.
Looking Ahead: Hire for Obsession, Ship Cool Content
Rahul is actively looking for two hires at Greptile right now: a growth engineer to build web projects and product-adjacent experiences, and a developer relations person who lives inside the content their audience already consumes, including AI releases, infrastructure outages, and open-source moments worth writing about.
Both hires follow the same logic: find the function that requires genuine obsession. Staff it with someone who will obsess.
The broader bet underneath all of it is simpler than any framework.
"In a dev tooling space, if you put out cool content, people actually just manage to find it themselves, because they're so naturally curious."
That's not a paid media thesis or an SEO infrastructure play. It's a belief that the best growth marketing for a developer tool is content so genuinely useful that the right people find it on their own. Build it well. Post it in the right places. Trust the audience.
The Figma playbook took five years to fully mature. Rahul believes the same engine at Greptile, built with clearer frameworks and AI compressing every execution cycle, could get there in one or two. The principles haven't changed. The clock just runs faster now.
Divisional connects founders with fractional marketers who've built these kinds of content-led systems before, and know how to build them for your stage. Let's talk.