Doculabs hosted its third robotic process automation (RPA) summit a little earlier this year. The good news is that we now see several early RPA adopters succeeding with their efforts.
The most successful have centralized their RPA efforts and formalized processes and methods. We talked about two potential roadblocks: fixing broken robots and measuring costs and benefits.
Rapid growth in the use of RPA.
At the Summit, there were eight separate Doculabs clients from the banking, investment management and insurance sectors. Some organizations had been working on robotics projects for the prior 18 to 24 months. In two cases, the clients had nearly 100 robots in production. This growth of RPA was dramatic compared with our first and second summits, which we held only 13 and six months earlier, respectively.
The preponderance of RPA efforts focused on back office repeatable functions. In one case, at a major bank, robots automated the entry of data into four disparate systems. These were simple tasks associated with the company’s “know your client” process, repetitive jobs that the average worker would not want to perform anyway.
In another banking case, robots now lookup data from different systems, all in support of client loan applications. The lookups from record-keeping systems include credit checks and credit card payment histories.
These mundane data lookups take away numbing, repetitive tasks, enabling humans to focus on more critical analysis and decision-making.
Two factors that assure RPA success.
There are two key takeaways from those organizations that have begun to embrace RPA. These organizations have centralized their RPA efforts, and they’ve formalized the process and methods of creating new RPAs.
1. Centralize your RPA efforts
Several organizations finally are putting significant resources into their RPA efforts. They have a centralized team managing the effort. In one case, this includes a Center of Excellence that shares RPA development and management resources—some 15 people—between five business units.
2. Create new robots using a factory-style operating model.
This organization, in effect, turned RPA into a factory-style operating model, using a model borrowed from manufacturing. The strategy: combine business analysts and developers. That mix of resources—in concert with agile development methods—seemed to have done the trick for this insurance client.
Thus far, RPA isn’t causing layoffs.
But here’s one surprise: No one is getting laid off because of RPA. Instead, RPA has freed capacity. Many employees who were doing rote, repeatable tasks, now work in more strategic areas. For several organizations, that extra capacity coincided with more internal demand for resources fueled by business growth.
No one was sent home; no task was outsourced to India.
Indeed, for these companies the focus is now on overall digital transformation, especially when it comes to e-signatures and the automated filling-out of forms.
Post implementation RPA issues that need to be resolved.
1. Fixing broken robots.
What Doculabs has been working on with clients are post-RPA implementation issues. These robots, it turns out, are more fragile than anyone anticipated. That’s where a rock-solid support process needs to be sorted out. And that includes implementing solid information governance regarding the affected data, including setting up a business continuity plan.
2. Understanding the costs and benefits of RPA.
The other unresolved issue is how to recover the full costs of an organization’s RPA effort. Companies now must understand licensing and development costs, especially when there’s a dedicated team involved. Who pays for that infrastructure support? And what are the right KPIs for a large RPA effort?
In spite of these open questions, I prefer to dwell on the good news our summit participants delivered. In previous summits, RPA was met with a healthy dose of skepticism. That is no longer true. We had several organizations tell us about successful RPA developments and implementations. It’s working. Companies now have an RPA track record. This is real stuff!