Innodata Raised Its Forecast Despite Losing a Major Contract. What Are the Growth Drivers?

Despite losing a major contract, Innodata, a provider of data and tools for training AI models, has raised its forecast for the full year 2026. Previously, this loss had led to mass layoffs at Innodata and a sell-off of its shares—investors feared a slowdown in growth. The company cited the training of AI agents and robots as new growth areas. Following the release of its financial results, the company’s stock price soared by more than 9% in premarket trading on November 6, then fell by nearly 7% during regular trading hours, in line with broader market trends.

Innodata's revenue for the third quarter soared 53% year-over-year to a record $95.6 million, the company reported. This is 15% higher than Wall Street's consensus estimate.

In the third quarter, the company managed to attract five new clients and secure new orders from one of its existing clients—which made that client its largest, explains CEO Rahula Singhal. He did not name the companies themselves.

Net income for the same period increased by 24% to $10.3 million. Meanwhile, diluted earnings per share rose 21% to $0.29, which is 70% higher than expected, according to the company’s press release.

This allowed Innodata to raise its revenue growth forecast for all of 2026: from 40–42% to 42% or higher. These expectations could have been higher if the company’s traditionally largest client had not terminated one of its contracts in October, Singhal said during a conference call with analysts. The loss of this contract led to mass layoffs in India and the Philippines and a sharp drop in Innodata’s stock price. Investors feared that this could jeopardize the company’s ability to meet its 2027 targets.

Singhal said this move would reduce the company's revenue by $12 million in the fourth quarter. “The good news is that, according to the company’s expectations, this will be offset by new programs with the same client by the first quarter,” said Innodata’s CEO.

Several factors will contribute to the company's growth over the next 12 months, including physical AI (robotics with artificial intelligence), agent safety, and enterprise AI, Singhal said during a conference call.

In the first half of 2026, venture capital investment in physical AI reached $47.4 billion—more than in the years 2022–2024 combined, Singhal noted. According to him, as robots are deployed in factories, warehouses, and homes, the demand for data to train them for specific tasks will grow. In September, Innodata launched a motion capture lab to train humanoid robots. The company is preparing for a pilot project that, if scaled up in 2027, could generate $7 million in annual revenue.

Among other drivers, Innodata’s CEO cited agent training. He explained it this way: agents learn by making mistakes, but companies cannot allow them to do so for reasons of security, confidentiality, and the irreversibility of many actions. Therefore, agents are trained in an isolated environment where they can perform tasks repeatedly and learn from their mistakes—and this is precisely how Innodata plans to make money.

In the corporate sector, models are now expected to perform expert-level tasks. “We create complex reinforcement learning systems for these models and also develop their internal software,” Singhal told analysts.

Another source of growth for Innodata is security testing of AI agents. According to Singhal, as these agents are adopted in the healthcare and financial sectors, the cost of an error is rising. In several pilot projects, Innodata is testing whether an agent will circumvent rules or falsify results in order to complete a task.

Following the release of its third-quarter earnings report, Innodata’s stock jumped 9.14% in premarket trading on November 6, but the price fell when the main trading session opened—at the time this news was published, one share was trading at around $36. Since the beginning of the year, the shares have fallen in price by nearly 30%.

However, Wall Street views the company's prospects positively: its stock has received four analyst recommendations, all of which are "buy." The average price target is $123.7, which implies growth potential of more than 220%.