The on-demand work model really took off when Uber made it easy for people to hail a ride from the comfort of their couches. The taxi industry was ripe to disrupt, marketplace technology was available, and the on-demand model was the right business model. Uber was so successful that the term Uber of became the zeitgeist for startups in the on-demand economy.
What followed was a surge of companies cloning the on-demand ride-hailing model for different industries from food delivery, handyman services, house cleaning, to weed delivery. Many of these jobs are rote work. Tasks involved in these jobs such as driving, ordering food, and cleaning are repetitive and can be scaled across many on-demand workers relatively easily. And given many of these jobs involve simple tasks, there are many people who can do them and the supply of workers is high.
Companies spend millions of dollars generating or purchasing insights every year. In 2012, a report by McKinsey and IDC found that an employee spent an average of 8.8 hours searching for information and an additional 8.1 hours analysing it during a workweek. That equates to around 60 days per year of people spending time generating insights rather than executing on them.