ChatGPT: The "Low-Key" Research Preview That Became the Fastest-Growing App in History

Most products that change the world get a big launch, a stage, a countdown, a press event. ChatGPT got almost none of that. It was released quietly, on a Wednesday, by a company that didn't even bother tracking its own website traffic that day.

A Research Lab, Not a Product Company

ChatGPT's story begins with its parent company, OpenAI, founded in San Francisco in 2015 as an AI research lab structured around a non-profit entity, later joined by a capped-profit subsidiary. Its founders included Sam Altman, Greg Brockman, Ilya Sutskever, Elon Musk, and several others, who together pledged an initial $1 billion toward the venture. Musk stepped down from OpenAI's board in 2018 but remained a donor. In 2019, Microsoft made its first major investment in the company, committing $1 billion, a relationship that would grow substantially in the years to come.

For most of its early life, OpenAI operated the way research labs typically do, publishing papers, releasing models to developers, largely unnoticed outside AI and academic circles. That changed on November 30, 2022.

A Launch With No Fanfare

ChatGPT was released not as a flagship product with a marketing campaign behind it, but as what the company itself described as a low-key research preview, built on the GPT-3.5 model. There was no major press event, no giant marketing budget, and reportedly so little internal expectation of interest that OpenAI wasn't even tracking its website traffic closely at launch. According to Sam Altman, it was, by his own account, a quiet, sleepy day for a small startup.

That quiet didn't last. Within five days, ChatGPT had attracted a million users. Within about two months, more than 100 million people had tried it, making it, at the time, the fastest-growing consumer application in history. The surge was driven almost entirely by word of mouth, people sharing surprising, funny, or genuinely useful conversations they'd had with the chatbot, rather than by any coordinated advertising push. In the aftermath, OpenAI's valuation reportedly climbed to around $29 billion, and in January 2023, Microsoft followed its earlier investment with a second, multi-year commitment reported at $10 billion.

From Free Preview to a Business Model

As ChatGPT's user base exploded, OpenAI began building out the infrastructure of an actual product business around it. A paid tier, ChatGPT Plus, offered enhanced access for a monthly subscription, alongside developer API access that let other companies build their own products on top of OpenAI's models. In November 2023, at its first developer conference, OpenAI unveiled "GPTs," customisable versions of ChatGPT that users could build and, eventually, share through a planned GPT Store.

Growth wasn't without turbulence. On November 17, 2023, OpenAI's board abruptly removed Sam Altman as CEO, stating that he had "not [been] consistently candid in his communications with the board." The decision triggered immediate backlash from OpenAI's own employees, and within days, Altman was reinstated as CEO. The GPT Store's launch was quietly delayed in the aftermath as the company worked through the fallout. By December 2024, OpenAI had introduced an even higher-tier subscription, a $200-per-month ChatGPT Pro plan, reflecting the company's growing ambitions around premium, professional-grade access to its models.

A Dominant Position, Then Real Competition

For a period after its launch, ChatGPT held an overwhelming share of the AI chatbot market, reported at around 87 percent in January 2025. That dominance has since narrowed considerably as rivals scaled up their own offerings, with reports indicating ChatGPT's share had fallen to roughly 65 percent by around September 2026, while Google's Gemini grew from about 5 percent to more than 18 percent over the same period.

A Notable Reversal on Advertising

For years, Sam Altman spoke publicly about his discomfort with advertising inside AI products, at one point describing ads in AI as "uniquely unsettling" and calling advertising a "last resort" business model for the company. That position shifted in early 2026. In December 2025, ChatGPT briefly began surfacing unsolicited in-app suggestions mid-conversation, including one widely shared instance recommending the Peloton app to a paying Pro subscriber, a feature OpenAI's Chief Research Officer disabled after user backlash over the following weekend.

Then, around the time of Super Bowl LX in February 2026, OpenAI made its Super Bowl advertising debut, even as rival AI company Anthropic ran its own Super Bowl commercials directly mocking the idea of advertising inside AI assistants, reinforcing its own Claude product's ad-free positioning. Reports indicated Claude's App Store ranking rose sharply in the days following that campaign. Just hours after Anthropic's ads aired, OpenAI announced it was beginning to test advertising within ChatGPT itself, rolling it out to a subset of free and lower-tier "Go" users in the United States, with ads clearly labelled as sponsored content. At the time, only around 5 percent of ChatGPT's roughly 800 million weekly users were paying subscribers, a gap that reportedly factored into the company's shifting stance on monetisation, alongside a reported $1.4 trillion in committed infrastructure spending through the early 2030s.

The Marketing Strategy: Let the Product Advertise Itself, Until It Couldn't Anymore

What makes ChatGPT's early growth genuinely remarkable from a marketing standpoint is how little conventional marketing had to do with it. Rather than investing in advertising to build awareness, OpenAI released a free, immediately useful product into the world and let curiosity and word of mouth do the rest, a rare case of virality outperforming any planned campaign in both scale and speed. That free-access model functioned as its own funnel, hundreds of millions of people experienced the product firsthand before OpenAI introduced any paid tier at all, by which point trust and habit were already well established. Only years later, under mounting commercial and infrastructure cost pressures, did the company begin experimenting with a more conventional monetisation lever, in-app advertising, a shift its own leadership had previously resisted, and one that competitors were quick to use as a point of public contrast.

From a Quiet Wednesday to a Global Habit

ChatGPT's story is ultimately a reminder that the most explosive product launches don't always come with the loudest marketing, sometimes they come from simply putting something genuinely useful into people's hands and letting them tell each other about it. What began as a modestly framed research preview, on a day so unremarkable that no one thought to measure it, ended up reshaping how hundreds of millions of people search, write, and work, one shared conversation at a time.