The AI Pivot Nobody Saw Coming: How G2 Became ChatGPT's Most-Cited B2B Source
Written by: Taylor Cromwell
Photography: Provided by G2
Summary
- Three-generation bootstrap: Abel's grandfather sold pumps before manufacturing them in post-war Germany. Abel built BigMachines over thirteen years (sold to Oracle), SteelBrick in two years (sold to Salesforce for $360M), then spent a decade on G2.
- The founding insight: After struggling to help customers discover BigMachines' niche software, Abel launched G2 in 2012 to bring Yelp-style peer reviews to enterprise software buying—replacing expensive analyst reports with crowdsourced validation.
- The risky parallel bet: With G2 at just $2.3M in funding and minimal traction, Abel started SteelBrick in 2014—running both companies simultaneously. SteelBrick sold in two years while G2 quietly built its moat with 500,000 reviews by 2018.
- The AI inflection: When ChatGPT launched in 2022, G2's decade of work—3 million reviews, 180,000 products, 2,000 categories—became exactly what AI needed. G2 is now the fourth most-cited source on ChatGPT, ranking above Wikipedia for B2B software queries.
- From disruptor to infrastructure: G2 evolved beyond competing with Gartner to becoming AI's trust layer. Reaching 100M+ buyers annually, Abel now envisions AI agents negotiating software deals in real-time using G2's verified data as the shared source of truth.
Introduction
Chances are, the last time you searched for software (whether on Google or ChatGPT) you landed on something powered by G2. Its data now underpins how companies decide what tools to buy.
What you see on the surface doesn’t show what’s underneath: more than a decade of building and rebuilding. Behind that infrastructure is a founder who’s lived every version of startup life — euphoria, exhaustion, and reinvention.
By the time Godard Abel launched his third company, he’d already built one over thirteen grueling years, sold another to Salesforce in two, and was still chasing an idea that refused to leave him alone. That idea became G2.
What began as a “Yelp for business software” is now the fourth most-cited source on ChatGPT, ranking ahead of Wikipedia for B2B software queries. A decade later, G2 has become the infrastructure AI relies on to understand how businesses buy..
Today, as CEO of a global marketplace used by millions, Abel is still guided by that instinct and by a belief that software buying itself is on the verge of another transformation.
Act I: The Entrepreneurial DNA
The Rubble and the Vision
In 1946, Abel's grandfather founded Abel Pumps in Essen, designing pumps for coal mines restarting Germany's economy. With no capital to manufacture, he sold pumps first, then scrambled to outsource production. One pump at a time, payment before production (bootstrapping before Silicon Valley gave it a name!)
Young Godard absorbed the lessons: build from nothing, sell before you manufacture, trust that execution will follow vision. "I grew up around entrepreneurship," Abel recalls. "My dad was always working hard — it felt gritty, but that was normal."
The Parallel Path Never Shown
In high school, after the family had moved from Germany to Pittsburgh, Abel and a friend launched Ultimate Car Care, a neighborhood car-washing business. It taught him the thrill of earning a customer’s trust.
That spark reignited in a very different setting. After graduating from Stanford Business School in 1999, Abel found himself in the middle of Silicon Valley’s first boom.
“I think if you had any entrepreneurial bones in your body, if you were in California at the end of 1999, like you were gonna start a company.” He was fascinated by how Michael Dell had transformed PC buying with an online configurator. “You could choose your memory, your hard drive, your operating system and magically order it online,” he says.
That idea — configuring complex products digitally — collided with a conversation back home over Thanksgiving. His father, still running the family’s industrial business, doubted that the internet could ever reach manufacturing. “You can’t sell pumps online,” he told his son. “Every order is custom-engineered.”
Abel saw an opening. “Somehow with my dad, I said, hey, let’s build BigMachines. And his pumps were very big, hence the name.”
The idea was ahead of its time: a web-based platform that let manufacturers configure, price, and quote complex products online. But the execution was grueling. “I was 28 and naive,” Abel admits. “I had no idea what I was getting into.” It took thirteen years of rebuilding and persistence before BigMachines became the leader in what’s now known as CPQ software. Oracle acquired the company in 2013.
After the sale, the quiet was disorienting. For the first time in years, he had nothing to build. “Friends would say, ‘You’ve made it — why aren’t you happy?’ But I missed having a team, a mission.”
As he reflected on what hadn’t worked at BigMachines, he felt frustrated by how painfully slow and opaque the enterprise software buying process had been. It had taken nine years for BigMachines to appear in a Gartner report, twelve to be named a leader.
Meanwhile, consumers were already using Yelp, TripAdvisor, and Amazon reviews to make everyday decisions.
“If you were buying a $50 dinner, you could read hundreds of reviews,” Abel recalls. “But if you were buying a $500,000 software system, you had to call a paid analyst.”
That gap sparked his next idea: a crowdsourced marketplace where business buyers could discover and validate software based on trusted peer reviews. “There was no site like G2,” Abel says. “That was really the vision.”
If you were buying a $50 dinner, you could read hundreds of reviews. But if you were buying a $500,000 software system, you had to call a paid analyst.”
— Godard Abel, co-founder and CEO at G2
Act II: Reinvention
The Spark for G2
By 2012, Abel was restless again. BigMachines had taught him how to build, but it had also exposed a structural flaw in the enterprise software world. “It was very hard for our customers to discover our software,” he recalls. “We made kind of niche software for manufacturers — eventually called CPQ — but most people hadn’t even heard of that category. It was very hard for customers to find us.”
That tension planted the seed for G2: a platform where software buyers could rely on authentic peer reviews instead of paid reports. “We had a bunch of other ideas,” Abel says. “But this one just felt divine. It had energy. We all aligned around it because the industry needed it.”
He pulled together a trusted team from BigMachines — Mark Myers, Matt Gorniak, Tim Handorf, and Mike Wheeler — and set up shop in his basement. “Stacy [Godard’s wife] and her friends would say, ‘Who are these mangy guys coming over every day?’” Abel laughs. “We were reading The Lean Startup and said, ‘Let’s just build an MVP and launch at Dreamforce.’”
By September 2012, they had something live…barely. Abel knew G2 would take time to build the trust, volume, and verification required for network effects. But with no revenue and slow adoption, it was a long bet.
And that’s when he did something few founders would dare: he started another company while G2 was still struggling.
In 2014, two years into G2 with little traction, Abel started SteelBrick with co-founder Matt Gorniak. G2 had grown to just $2.3 million in seed funding, struggling to monetize while Tim Handorf served as CEO. Meanwhile, Abel saw an immediate opportunity in CPQ software; the same space where his first company, BigMachines, had taken twelve years to succeed.
“I remember sitting in a dinky office hallway in Highland Park, above a bad Mediterranean restaurant,” Abel recalls. “I’m funding two companies, and at some point I’m going to jeopardize Stacy and the family. Have I just gotten over my skis?”
The gamble revealed pattern recognition at work. While G2 needed time to build its review moat (by 2015, they had just 50,000 verified reviews coming in at 3,000 per month) SteelBrick could capture immediate market demand.
In just two years, SteelBrick achieved what took BigMachines over a decade. Salesforce acquired it for $360 million in 2016, with Abel spending a year as SVP learning Marc Benioff’s scaling playbook.
The early metrics seem quaint now. By September 2016, G2 celebrated hitting 100,000 user reviews. Monthly sessions reached 400,000. They'd raised $7 million in Series A from Pritzker Group Ventures.
Every review collected, product catalogued, and category mapped was creating structured data that didn’t exist anywhere else. While Gartner published quarterly PDFs, G2 was building real-time, machine-readable intelligence.
By 2018, when Abel returned as CEO after Steelbrick's exit (with Handorf moving back to Chief Product Officer), the foundation was set: 500,000 reviews, 58,000 products, 1,200 categories, 2 million monthly visitors. The acquisition of Siftery added tracking data from 30,000 additional products, creating what Abel called a "third pillar": beyond discovering and buying software, G2 would help companies manage their stack.
When ChatGPT Changed the Game
November 2022 was the inflection point. ChatGPT launched, and within weeks, Abel noticed something remarkable: AI was citing G2 more than almost any other B2B source.
The platform's position was unique. While competitors scrambled to add chatbots, G2 had spent a decade building exactly what AI needed:
- Three million verified reviews providing nuanced, contextual feedback
- 180,000 products and services catalogued with structured metadata
- 2,000 software categories mapping the entire ecosystem
- Daily Grid updates versus quarterly analyst reports
- Free access enabling unlimited AI training
"We thought we were building Yelp for business software," Abel reflects. "We were actually building AI's memory."
The Answer Engine Revolution
G2's response wasn't to compete with AI but to become its infrastructure. The shift from SEO to Answer Engine Optimization (AEO) required fundamental changes in how they structured data.
The results were immediate and stunning. G2 became the fourth most-cited source on ChatGPT overall in 2025, ranking above Wikipedia and other enterprise companies for B2B software queries. The platform appears in approximately 5,900 Google AI Overview keywords. On Perplexity, G2 ranks ninth among all cited sources.
"Last quarter's analyst report is already outdated," Abel explained. "AI needs real-time data rooted in authentic customer voice."
The technical innovation went deep. G2 restructured its entire data architecture to be AI-readable, creating verification systems that large language models could trust. While traditional analysts published static reports, G2's dynamic data updated continuously, providing AI with fresh intelligence every day.
The Product Transformation
G2's AI integration came in three strategic waves:
First, the launch of G2.ai — an AI assistant built on ChatGPT that revolutionized software discovery. Instead of requiring users to know specific categories, buyers could describe problems in plain English. "AI is just a better interface," Abel explained, after a decade of trying to build guided workflows. G2.ai could surface niche "long-tail" apps users would never find through traditional browsing.
Second, the acquisition of unSurvey brought conversational AI to review collection. Three-minute voice reviews replaced twenty-minute forms. The YCombinator-backed technology also powered AI-led custom research, delivering voice-of-customer insights in days rather than months.
Third, the partnership with Profound created an AI Visibility Dashboard, showing companies exactly where they appeared in AI responses. This became a roadmap for improving AI presence. Companies could see which prompts triggered mentions, how they ranked against competitors, and what actions (like collecting recent reviews) would boost their AI visibility.
We thought we were building Yelp for business software. We were actually building AI's memory.”
— Godard Abel, co-founder and CEO at G2
Act III: The AI Era
Disrupting the Disruptor
The original mission to disrupt Gartner's analyst oligarchy seems almost modest now. Today G2 has become the trust layer for algorithmic commerce, no small feat.
The contrast with traditional analysts is stark. While Gartner charges upwards of $100,000 for annual subscriptions, G2 provides free access to anyone. The platform now reaches over 100 million buyers annually through direct visits and syndication to AWS Marketplace, Azure, and other platforms. G2's Grids refresh daily while analysts publish quarterly. Most importantly, G2 aggregates millions of implementation experiences rather than relying on expert opinion.
The numbers tell the story. In the past year alone, G2 added fifty-six new AI categories and 4,871 new AI products — increases of 75% and 119% respectively.
The Data Solutions business, initially aimed at investors, now serves consulting giants like McKinsey, Bain, and BCG. Non-investor customers represent over fifty percent of new Data Solutions deals, spending twice as much as traditional investor customers.
The Network Effect in the AI Age
G2's 2023 partnership with AWS proved prescient. By 2025, AI-generated comparisons based on G2's reviews appeared on 2,500+ AWS Marketplace listings. When Reddit users discuss software, they increasingly find G2's verified data integrated into company profiles. The platform had become omnipresent — exactly what AI omnipresence required.
"G2 is everywhere you go for software," became more than marketing. Through syndication deals and API integrations, G2 data appeared wherever buyers searched. The compound value of ten years of reviews created an irreplaceable moat. Every review made the next AI answer more accurate. Every verification made the system more trusted.
While competitors rushed to add AI features, G2 had become AI infrastructure. The September 2025 partnership with Profound formalized this position, helping software vendors understand and improve their AI visibility across all major language models.
The Future Abel Sees
"I predict that one day we will have software buying agents talking to selling agents, so that a complex process like selecting enterprise software could be done by agents, almost in real time," Abel explains. “A buyer’s AI could describe exactly what the team needs — ‘I want a CRM that integrates with HubSpot and costs under $50 per seat’ — and a seller’s agent could instantly respond with a verified fit. The negotiation, validation, and trust scoring could all happen in seconds.”
It’s not a far-off vision. The pieces are already visible inside G2. G2.ai handles the discovery layer, matching problems to solutions in plain language. Verified reviews feed a real-time trust score that agents could rely on. The Profound partnership makes those interactions measurable: companies can already see how often their brand appears in AI-generated answers with G2 as the source, and why.
Abel sees that as the next frontier: a transparent marketplace where intelligent agents trade on verified human experience. “For that to work,” he said, “there has to be a shared layer of truth. That’s what G2 can be.”
It’s a fitting continuation of the family story. In 1946, Wilhelm Abel sold pumps one at a time from a bombed-out workshop in Essen. In the 1970s, his son brought that same precision engineering to the U.S. market. And now, in 2025, Godard Abel is building a digital infrastructure where software can be bought and sold instantly, with trust built in from the start.
The Vision Realized
"I want G2 to be known for shifting power from a single analyst to real users' voices," Abel says, leaning forward with the same intensity he had in that basement thirteen years ago. "And as the most trusted data source in the age of AI."
It's happening faster than even he predicted. G2's data flows freely through every major AI system, shaping millions of software decisions daily. The disruption Abel envisioned is complete, but in a way nobody could have foreseen.
The peak remains the same. The route keeps changing. But for G2, the climb itself has become the destination, a continuous ascent toward becoming ever more essential in a world where software, AI, and human judgment must somehow learn to trust each other.
So what’s next for G2? After crossing $100M ARR, Godard is looking to finally conquer Everest. “It’s a personal dream of mine to build a public-scale company and someday ring the bell. G2 is continuing on its growth path, and we look forward to a successful exit when the time is right.”
Spiraling Up: The Journey to Become Unicorn
HubSpot for Startups and LinkedIn collaborated on Spiraling Up: The Journey to Become a Unicorn—a three-season documentary series that told the story of various startup founders at different parts of their respective journeys, scaling their businesses to unicorn valuation (a private investment-backed company worth $1B or more). Each season featured one or two unicorn founders. Godaard's was the unicorn story from season 2 (released summer of 2023). Watch the full episode below.
Author: Taylor Cromwell
Taylor is a writer, interviewer, and founder of Creator Diaries, a newsletter digging into how real people turn ideas into income. Her editorial work spans HubSpot, beehiiv, Just Go Grind, Confluence VC, and more — covering founder strategy, content marketing, and the business side of the creator economy. When she’s not writing about building businesses, she’s probably wandering an antique market, walking with a podcast, or scheming her next countryside escape.
