IBM and Automation Anywhere  (AA’s) today announced a collaboration (note-not a formal partnership yet) to integrate (AA’s) Robotic Process Automation (RPA) platform, used to create software bots to handle repetitive, task-based work, with IBM’s portfolio of digital process automation software, that includes IBM Business Process Manager and Operational Decision Manager. For AA, the partnership is a validation of its strong position in the RPA market, as shown in the Forrester Q1 RPA wave. For IBM, the partnership will enable clients to use AA’s RPA platform to create software bots that execute tasks within larger business processes managed by IBM’s software. Here’s our take.

IBM Can Add Smarts to AA’s RPA Platform

RPA works in a very dumb fashion today – mimicking human keystrokes and mouse movements – where all decisions must be explicitly programmed into the script. The result is that there are very few real decisions made beyond simple logic loops. Watson will be relevant down the road, but as a first step, RPA will benefit from IBM’s mature business rules engine (Decision Manager) or the embedded rules in the BPM platform. But as interesting, IBM’s content analytics, if part of the collaboration, can allow AA to keep pace with unstructured RPA intelligence from Workfusion and other RPA competitors moving quickly in that direction. RPA use cases that comb through unstructured content will be ahead of chatbot and AI-based digital workers.

Further Market Validation For RPA

RPA is beyond needing market validation.  Forrester has been writing about the market for several years But all entire market is tiny.  Some of bigger players like BluePrism  are less then 20M in revenue.  Forrester’s estimate puts AA in the 25 to 35M range.  IBM getting into the market helps.

Evidence Of Future BPM And RPA Consolidation

RPA is more than a bit threatening to BPM.  The BPM market is not growing, The RPA market is.  Sometimes RPA gobbles up a BPM project’s forecasted savings because it got there a year  ahead of the BPM effort. And there complimentary. RPA’s Bots appear as tasks that humans previously performed on the BPM process map. To modernize and orchestrate complex processes you need BPM or case management support. RPA doesn’t cut it by far.

A More Practical Analytics Focus For IBM 

The IBM/AA announcement is important from the RPA market perspective, but small potatoes in the overall IBM empire. But it may signal a subtle shift. Watson, after four years, has not delivered the expected revenue not to mention some well publicized project failures where exhausted teams of data scientists and bottomless budgets could not reach a productive state. But to IBM’s credit they have tackled some of the hardest problems our planet has.

But why not focus on some less sexy but highly productive problem areas, like those that relieve low-value tasks from humans toiling away in cubicles. This is where IBM got its start after all-basic tabulation. AA’s RPA, with IBMs Decision Manager, BPM, and content analytics can dive into IBMs traditional backyard, where the energy for  RPA is today. It’s in the “business” which refers mostly to “operations” run as shared services such as procurement, supply chain, HR, and customer service. To a lesser extent, it includes sales, marketing, and functions within independent business units. What do these groups have in common? They depend on a lot of IBM applications and technology. And with these more practical automation building blocks, you will need teams of data scientists.