RPA state of the union

  • Facebook
  • Twitter
  • Linkedin
  • Email

Damo Vasudevan

Vice President, Digital

July 21, 2017 -

“You can fool all the people some of the time, and some of the people all the time, but you cannot fool all the people all the time."

Robotic Process Automation (RPA) started off as a scripting technology that helped a human with desktop-related actions. While the premise of freeing up the human worker to perform higher level actions has been maintained, RPA has migrated to a technology platform that works with almost human-like precision on mundane tasks like copy/paste, moving data between systems, and putting together reports. With the recent commoditization of machine learning and basic cognitive tech, robots have become both more efficient and better at learning on the job, which in turn enables them to deliver visible cost benefits almost right away. And as evidenced by the droves of customers at the recent customer events by the RPA providers, almost every customer seems to have tried the technology, with varying degrees of success. The last few years of platform updates have solved most of the basic RPA challenges paving the way for almost ALL the business units in an organization to go beyond basic experimentation to industrialization.

Every player in the space has some positioning. While AA initially focused on speed, they have spent the last few cycles addressing the scaling and integration components. BP focuses on developing components in almost an object-oriented approach to aid in reuse and create bots at scale. UIpath, on the other hand, is certainly easy to create but has a unique blend of assisted and automated robots running in parallel to solve a business problem. Workfusion, on the other hand, is betting big with their cognitive component as the future of enterprise automation, with desktop automation being offered for free. Pega, with its pre-built integration with workflow, also brings a consolidated platform that integrates workflow with automation and analytics at scale.

But what if we need to go beyond a few bots. The one thing that pops out on a bit of a deeper dive is ALL the platform providers have learnt from one another and built capability to match each other. Additionally, they have ALL also moved strategically towards machine learning and AI embedded automation since that's where they believe the second wave will require once basic automation is complete in an enterprise. And ALL are working to define strong connectors so they can very easily sit on current/legacy systems as well as provide a low-code environment for a business user to quickly create a bot.

While any these technologies is a great option to get started, the most critical questions for an enterprise leader are: 

  • Are we aligned as a company to automate and digitally transform ourselves? Digital transformation through automation has to be an integral strategy for growth
  • How should we centralize the governance of the RPA initiatives so we don't duplicate efforts across the organization? It is critical to making sure that all the decisions and analytics can be centralized so that the RPA initiative can pass the “all the people" test
  • How do we align the IT strategy with automation (e.g., to decide whether to build a new connector to SFDC or just automate the exchange with a bot)? Each IT decision will impact automation, so it is crucial that these strategies are aligned

The enterprise giants are also watching. IBM, Microsoft, SFDC, and SAP are already introducing heavy automation into their suite and are investing heavily in artificial intelligence (AI) technology integrated into their platforms. In a few quarters (yes, I said quarters, this market moves that fast) the automation problem for enterprises is poised to migrate to one that is solved using more sophisticated technologies, which may extend RPA into almost all parts of digital transformation, or on the other hand, completely kill the market. But given that the RPA players are already incumbent and looking deeply into enterprise processes, one would think they have a significant advantage as the market develops.

Will they react quickly, or be trampled by the bigger players? Or will a new entrant offer business value in a more meaningful way? Only time will tell.