Scaling robotic process automation (RPA): Are you ready to manage your digital workforce? Bookmark has been added
Scaling robotic process automation (RPA): Are you ready to manage your digital workforce?
As organizations adopt robotic process automation (RPA), many are focused on identifying the tasks and processes to be automated, building the proof of concept projects, and testing build-operate-transfer (BOT) capabilities in discrete applications. One area that may not be getting enough attention is preparation in managing the bots once they take their place alongside human workers en masse.
September 5, 2018
A blog post by Gina Schaefer, managing director, Deloitte Consulting LLP.
Fundamentally, a bot is software—a unit of capacity, a digital worker. Yet it cannot be managed like just another software application, especially when deployed by the dozens, hundreds, or even thousands. Effective digital workforce management will depend on insights from at least five key areas:
- Bot performance. After a bot is deployed and work assigned to it, how will you know if the work has been completed? Is the bot delivering expected results? What happens if it gets stuck mid-task? Visibility into bot performance will be important—visibility over and above technical status that many of the vendor tools provide today.
- Automation program performance. Beyond individual bots, how is your automation portfolio performing overall? Many organizations focus simply on the "bot count" without monitoring the efficacy of the investment. Important, high-level metrics across the portfolio should include hard and soft dollar returns on investment, full-time employee (FTE) savings, transaction volumes, automation utilization, and maintenance-related results.
- License management. As your bot workforce grows, how will you level the workload across the bot population? Optimizing a digital workforce without necessarily having to scale licenses and structure involves visibility into license utilization rates for the RPA program overall, by functional area, and according to prescribed time periods such as by month or hour.
- Bot-human interaction. How will you gather input from human workers to make improvements to automation? Effective bot management should include a feedback loop whereby information about automation initiatives is shared with "the business" and they, in turn, provide insights to help troubleshoot bot issues and elevate bot performance to the next level.
- Service management. For companies that use a centralized automation services model to manage bot deployment and operations, how will service levels be measured and reported on? Can service utilization and performance be tracked to identify improvement opportunities?
Visibility into, and management of, digital workforces is an emerging imperative. Tools are being developed to create visibility and governance over automation portfolios. But the bottom line is that you can't effectively manage a digital workforce as if it was a collection of apps. Bots are more like their human counterparts than software applications, so a leading practice for automation programs will be a digital labor management strategy that includes capturing, analyzing, reporting, and acting on a range of data—digital exhaust—being generated by your bot and human workforces.
Developing such a strategy and properly equipping your automation team to execute on it deserves more attention than it may be getting today.
A webinar series on data science, artificial intelligence, and design thinking
Catch up on the latest