Scale AI’s Sales Nearly Quadrupled in First Half
The Takeaway
• Scale AI’s revenue nearly quadrupled in first half
• Scale’s operating margin improved but company is still losing money
• Former Uber executive Jason Droege is joining as chief strategy officer
Helping artificial intelligence companies improve the accuracy of their large language models is turning into a giant business for Scale AI, which labels data for big AI developers like Meta Platforms and Google.
Scale nearly quadrupled its sales to almost $400 million in the first half of the year compared with the same period last year, according to people familiar with the figures. Scale is among the privately held firms that have benefited the most from the AI boom by providing infrastructure or services—in Scale’s case, human contract workers—to companies building advanced AI.
The financial figures provide the first glimpse of the startup’s performance since Scale, led by 27-year-old founder Alexandr Wang, raised fresh capital at a $13.8 billion valuation in May. Its investors include Accel, Founders Fund, Index Ventures, Thrive Capital and Greenoaks Capital. (Read our profile on Wang from June.)
Scale declined to comment but Wang tweeted on Monday that Scale had hit “nearly $1 billion” in annualized revenue, a measure that multiplies monthly revenue by 12.
Some of Scale’s executive team has shifted under Wang recently. The company recently told employees that chief technology officer Arun Murthy would leave the company, a person familiar with the matter said. Scale announced today that Jason Droege, a former Uber executive who last year left venture capital firm Benchmark, would join as chief strategy officer, reporting to Wang.
The eight-year-old startup, which handles the hiring and training of the contract workers, hasn’t yet turned a profit, but it managed to improve its operating margin in the first half. It spent about $1.20 for every dollar of revenue it generated, compared to $1.50 in the first half of the previous year.
Accounting only for the direct costs of its business, such as contractor salaries, Scale kept only about half the revenue it pulled in. That financial metric, known as gross margin, was just under 50%, much lower than tech investors typically look for in software firms. That’s down from about 57% in the first half of 2022.
The results show that Scale’s gross margin and revenues were slightly below what it had forecast to prospective investors earlier this year. It had predicted about $415 million in sales in the first half and a gross margin of about 51%, a few points higher than what it actually reported.
Still, the May fundraising armed Scale with a hefty balance sheet. The company had about $980 million in cash remaining at the end of the first half of the year.
Scale’s revenue spiked starting in the first half of last year, as customers building LLMs needed contractors to essentially teach the AI models by writing their own responses to questions submitted to chatbots.
Scale calls itself a “hybrid human-AI system to produce high-quality data at a low cost,” according to a presentation shared with investors in February, which The Information obtained. It employs hundreds of thousands of hourly workers to fine-tune data through a subsidiary called Outlier, The Information previously reported.
The focus on LLM customers was a pivot of sorts for Scale. It also has a similar business, largely using low-cost labor in the Philippines and Kenya, that labels data for self-driving–car companies. But that business has slowed in recent years.
The reliance on higher-paid, more-expert contractors also has helped boost revenue for Scale, as it can pass on many of those higher costs to customers.
The startup hasn’t been a slam-dunk investment for many Silicon Valley investors, who have been concerned about the company’s relatively low gross margins and heavy reliance on only a handful of large customers, The Information previously reported.