Stack Overflow’s forum is dead but the company’s still kicking

Source: Hacker News

Overview

The platform is raking in millions of dollars in revenue, with AI an ironic new source of revenue.

When Elon Musk described Stack Overflow’s plight as “death by LLM” in July 2023, he wasn’t exaggerating. After years as the go‑to resource for developers, Stack Overflow peaked during the pandemic as coders flocked to its evergreen Q&A forum. Since the rise of powerful code‑writing assistants such as ChatGPT, Cursor, Claude, Google’s Gemini and Microsoft’s Copilot, traffic has plummeted. Last month the site recorded just 6,866 questions—roughly the volume seen when the site launched in 2008.

The forum vs. the company

While the public Q&A forum appears dormant, Stack Overflow the company is still operating. Unlike Chegg and other knowledge hubs that have struggled with generative AI, Stack Overflow has found ways to monetize its massive back catalog of content.

• Revenue growth: Annual revenue has roughly doubled to $115 million despite a sharp decline in user engagement since ChatGPT’s 2022 debut.
• Loss reduction: Losses shrank from $84 million in FY 2023 to $22 million in the most recent fiscal year, helped by cost‑cutting measures, including mass layoffs.

New business model

• Enterprise solutions: The primary revenue driver is now “Stack Internal,” an enterprise offering that provides a generative‑AI add‑on powered by the site’s decades‑long archive of questions and answers. It is currently used by 25,000 companies worldwide.
• Data licensing: Stack Overflow also licenses its curated data to AI companies in a model similar to Reddit’s, which generated over $200 million from data licensing in 2024.

The company’s niche rests on the trust built by its long‑standing community and the expertise embedded in its content.

CEO perspective

In December 2023, CEO Prashanth Chandrasekar explained the shift:

“…when we saw the questions decline in early 2023, what we realized is that pretty much all those declines were with very simple questions. The complex questions still get asked on Stack because there’s no other place. If the LLMs are only as good as the data, which is typically human curated, we’re one of the best places for that, if not the best for technology.”

Large language models need extensive, high‑quality data about coding problems. Stack Overflow’s extensive, albeit aging, repository serves that need, even as many queries move to private chat windows with LLMs. The platform has become a “canary” illustrating the evolving dynamics of AI‑driven knowledge consumption.