When people talk about gig work, the conversation often focuses on location-based companies Uber or DoorDash. But online gig work is growing globally and could help provide income for women and young workers in regions with fewer job opportunities, according to a new report from the World Bank.

It’s impossible to know just how many online gig workers are operating across the world. The number of global online gig workers ranges from 154 million to 435 million, which accounts for 4.4% to 12.5% of the global labor force, the World Bank estimates.

But demand for these jobs, which include things like data entry, software development, and image tagging, has shot up 41% between 2016 and the first quarter of 2023, according to the report.

Here are the main takeaways from the report:

Online gig work isn’t just for developed countries. In total, low- and middle-income countries account for 40% of traffic to gig platforms. Developed countries dominate the demand for online labor, but demand for online gig workers has risen faster in developing countries.

Online gig work opens up job opportunities for women and youth. But there are still disparities and concerns. People without access to internet or computers are left out. Women and workers in developing countries also experience discrimination when it comes to accessing work or high-paying tasks. Gig work is also inherently sporadic, without a clear path forward.

Social insurance coverage is low among gig workers, with about half of those surveyed saying they do not have a pension or retirement program. That figure is higher in other countries. “More innovation is needed in the design of social insurance products for workers with sporadic incomes,” the report’s authors write.

Regional and local gig platforms are vital in their markets, often catering to micro, small, and medium enterprises, startups, and self-employed or single-owner businesses. These smaller platforms can help employers find gig workers with similar backgrounds, or who are in the same time zone, and can be cost-effective. Still, they’re difficult to establish as viable business models due to the smaller size of their user base.