The crowd in the cloud

The crowd in the cloud

Over the past two decades, offshoring has revolutionised business. Now cognitive automation threatens to sweep the model away altogether. Mark Harris meets Shamus Rae, Head of Innovation and Investments for KPMG in the UK, to discuss the future of outsourced labour, and how the cloud may play a role.

Partner, Head of Innovation and Investments

KPMG in the UK


Also on

the crowd in the cloud
  • Robots could significantly shrink $300bn outsourcing market
  • The future of robotic process automation is about bring more 'human' intelligence to robots
  • In five years, teams could be made up of employees, crowd specialists and cognitive robots, all creating and collaborating together

Over the past two decades, offshoring has transformed business models, allowing enterprises to move significant segments of their operations to locations with lower labour costs. But, critical as this labour revolution has been, it’s offshoring where the impacts of cognitive automation will be felt first.

“Labour arbitrage, or offshoring, is going to die very quickly,” says Shamus Rae, Head of Innovation and Investments for KPMG in the UK. “The next wave of robotic process automation, which will be about bringing more intelligence to robots, is coming down the line very quickly.”

Robot outsourcing
Businesses that outsource work to countries with cheaper personnel – currently a $300bn market – will find that a robotic system might be one-third the price of an offshore worker, according to a study by the Pew Research Center. And as cognitive automation systems mature, they will increasingly compete with local workers, too.

“In the UK, you can imagine that over the next decade, there will be 10 to 15 million jobs that will be rotated because of this technology coming through,” he says. “That will mean millions of people trying to maximise their income when jobs are scarce.”

A threat to crowdworkers
One place they are likely to head is the cloud. In the US, there are already more than three million so-called crowdworkers, people working from home on digital platforms like Amazon Mechanical Turk, CrowdFlower and Upwork. Much of the work they do today is similar to the tasks that robotic process automation is being developed to tackle: the collection, categorisation, and transcription of data.

But there is no reason why on-demand digital workers cannot be professionals – or at least do the work that professionals have traditionally done. Online education providers are using virtual professors to mark work, law firms are outsourcing document processing, and some researchers are even experimenting with using untrained crowdworkers to analyse medical imagery.

“The problem the crowd has is the lack of trust gained from an established brand,” says Rae. “To make a policy decision about pricing in Latin America or have someone look at your medical X-ray, I suspect you’re not going to go to the crowd. Businesses will want to have some brand association that allows them to believe in a result.”

Making a seamless transition
KPMG in Australia operates a platform called Marketplace that gives on-demand access to professionals trained in financial modelling, analytics, project administration and book-keeping. “It effectively creates a crowd environment of staff for our clients to access,” says Rae. “They can say, I just want this one task done and I need these skills.”

A platform like this could provide a seamless transition into a robotic system as the technology improves. At first, says Rae, expert workers will use the cognitive automation system as they would any other digital tool: “You can imagine us creating a bot for tax service and having a crowd with the right skills checking its work, or answering questions that can’t be answered by the bot.”

Robots learning on the job
The robot would gain experience from its many interactions – potentially thousands or more each day – and gradually improve its responses. People would be involved every step of the way, directing the way the robot learns. “You can imagine a world in five years’ time when teams are made up of company employees, crowd specialists and cognitive robots, all creating and collaborating together,” says Rae.

A more dynamic working environment
While customers’ experience would barely change, businesses could eventually deliver high levels of service with far fewer personnel. “As people get displaced out of larger organisations, small and medium-sized enterprises (SMEs) will pick up the slack,” says Rae. “The crowd system will become more and more important, and will be a way for SMEs to pick up specific skills.”

Come the cognitive automation revolution, corporations will slim down, the SME sector could become more vibrant, and many more workers – human and robot alike – are set to be labouring in the cloud.

Connect with us


Request for proposal



KPMG’s new-look website

KPMG’s new-look website