Stay up so far with free updates
Simply log in Artificial intelligence myFT Digest – delivered straight to your inbox.
Artificial intelligence startups are generating revenue faster than previous waves of software firms, in accordance with recent data. This data suggests that transformative technology can also be creating strong firms at an unprecedented rate.
According to an evaluation of payment information by fintech group Stripe, leading AI groups generate tens of millions of dollars in revenue inside a 12 months – significantly faster than comparable non-AI technology groups over the life cycle of a start-up.
The findings come as investors raise questions on the economic advantages of generative AI and the likely returns on Big Tech's planned trillion-dollar investments in computing infrastructure to support the technology in the approaching 12 months.
However, the info suggests that young AI firms have more momentum than other highly touted technology startups of yesteryear, particularly within the software-as-a-service category.
Stripe, itself a $65 billion Silicon Valley behemoth, collected data on the annual revenues of the highest 100 private AI firms using its payments platform as of July 31, 2024, in comparison with a comparable cohort of promising SaaS startups -ups (as of July 2018). .
Stripe's customers include OpenAI, Anthropic, Mistral, GitHub and Midjourney, in addition to many other of the best-known AI groups.
The data showed that it took the AI startups within the cohort a median of 11 months to succeed in $1 million in annual revenue after their first sales on Stripe, in comparison with 15 months for the previous generation of SaaS Company.
AI startups which have reached over $30 million in annual revenue have reached this milestone in 20 months – five times faster than previous SaaS firms.
However, a report from Goldman Sachs this month raised concerns concerning the profitability of AI firms because “today's AI winners aren’t any longer capital-light firms,” citing the numerous costs of computing infrastructure to operate and train AI models are required.
The Stripe data reflects how AI startups — lots of that are the most recent iteration of SaaS firms — are adapting to those market dynamics by constructing experimental products that customers are willing to pay for.
ChatGPT, OpenAI's AI chatbot, launched in November 2022, became the fastest-growing consumer application in history when it reached 100 million users inside two months of launch.
OpenAI has developed a subscription service for businesses to access ChatGPT, which has helped the corporate's annual revenue reach $3.6 billion, in accordance with individuals with knowledge of the group's funds. However, the corporate also burns well over $5 billion a 12 months because it invests in training recent models.
“Unlike previous generations of software firms, AI firms pay significant computational costs from the beginning and are due to this fact under pressure to speed up monetization,” said Emily Sands, Head of Information at Stripe.
The demand for generative AI – software that may generate text, code, images, audio and video, amongst other things – can also be global. According to Stripe data, around 56 percent of AI firms' revenue got here from overseas.
This demand advantages groups that produce AI images and audio, corresponding to London-based unicorn company ElevenLabs, which makes AI language software, and German AI language translation company DeepL.
“In countries like Singapore (and) Iceland, we see that greater than 3 percent of the population actually buys from these top 100 AI firms,” Sands said. “It’s a really global phenomenon.”
Stripe itself is currently training AI models using its vast data, which incorporates greater than $1 trillion per 12 months from billions of transactions and tens of millions of companies, to develop more personalized checkout and payment processes.
The speed of monetization reflects the power of startups to bring recent products and features to market based on rapidly changing AI models from OpenAI, Anthropic, Google and Meta, forming the idea for applications corresponding to transcription and coding assistants .
Byron Deeter, partner at Bessemer Ventures, which invests in SaaS firms, said the issue with larger software firms lies of their older technology architectures and inherent slowness, while startups have a proposition to quickly improve productivity.
“We see many (AI) firms go from zero to tens of tens of millions of dollars in revenue in a couple of years,” Deeter said.
