For the past year, we’ve been in conversations with many of our customers who are trying valiantly to leverage the capabilities of Robotic Process Automation (or RPA) to help address time-consuming, repetitive manual processes. At so many enterprise companies, the promise of eliminating back-office manual tasks via a “bot brigade” was understandably alluring. Yet as this McKinsey & Company article outlines well, many of the early adopters are having to step back and recalibrate both their expectations and approach. As the McKinsey authors conclude, “A bot is a tool in a toolkit, just like self-serve tools, work-flow tools, lean-process maps or six-sigma methodologies. Companies need to apply these tools as part of an orchestrated action, not in isolation.”
As with every technology tool, it’s easy to get wrapped up in the bells and whistles and forget that our goal is to solve a business problem in a business context. So as usual, we need to step back and ask “what is the problem we are trying to solve?” and look to solve that problem in the smartest, simplest way possible using the tools available. As it turns out, PMG provides an enabling capability that can help support (or even replace!) an existing RPA program by providing that orchestration of an end-to-end business process, linking integration capabilities of RPA with other integrations, automation approaches, user interfaces, etc.
PMG as RPA Orchestrator
To prove the point, we recently partnered with one of our customers, a Fortune 100 health insurance company. We wanted to demonstrate the synergies between the PMG Digital Business Platform and their Robotic Process Automation (RPA) toolkit. This customer was excited about the new possibilities presented by the buzz-filled world of RPA but struggling to get full end to end business process outcomes. Our goal was to create a Proof of Concept (POC) that would pair the technologies to leverage key capabilities of each and achieve an overall business outcome:
- RPA offers the ability to interact with legacy applications that are generally only managed through screens by humans
- PMG offers the ability to orchestrate and automate a fully useful end to end process including all calls to bots, other systems, or human interactions via a configurable, codeless UI
Test Bill Request
The process we selected for the POC was the “Test Bill Request.” This is a process that our customer executes 700 times per month, as follows:
- Team members in sales or customer service send an email to a shared mailbox requesting that a test bill be run. The Sales and Service Associates must know which data is required and include it all, accurately keyed into the email.
- Those emails are converted to request tickets in a legacy workflow system
- The tickets are worked by offshore resources who then swivel to the legacy billing system, log in, and engage in a series of screens to enter the test bill request
- This request in the legacy billing system is then processed by a batch mainframe job that runs every two hours
- The offshore resource must then re-engage the ticket once enough time has passed for the job to have run. They log into the legacy billing system again, search for the test bill, and save that test file as a pdf to a folder location
- Finally, the offshore resource emails the test bill pdf to the original requester
Problems with the Current Manual Process
Currently, the process has lag between the time the email is sent and the ticket is picked up to be worked by the offshore resource. This can be hours or days. Once they enter the test bill request and pause their ticket while the batch job runs, there can be additional lag between the time that the job runs and the time they pick the ticket back up.
There is also the opportunity for invalid or incomplete data to be entered into the email, causing delays and rework as the offshore resource communicates back and forth with the requester.
Obviously, in addition to the delay, it’s costly for the manual work to be done by hourly resources, regardless of how cheap.
Finally, the user experience is sub-optimal since the user doesn’t know once they send their email when to expect the test bill back. It could be later that day or many days hence – who knows? And they have no way to find out. And clearly the user needs this information to ensure a great end-customer experience with no surprises when the actual bill is generated – we want to support that exercise.
Automating the Experience
Here is how our team automated this process:
- First, we configured a simple form in the PMG platform that the Sales and Services Associates can now use to submit the Test Bill Request. This form is searchable in PMG’s friendly user portal. The form can easily be configured to provide guidance and instructions to users, and also to integrate with other data sources for real time validation of the data entered. Also, the form can easily be configured to allow a request by one user on behalf of someone else.
- Next, the PMG workflow engine converts the data entered and validated on the form to a .csv format that the RPA bot can consume. PMG invokes the bot and provides the data in the file.
- The bot then does its work, logging into the legacy system, passing in the values provided, and triggering the test bill.
- PMG handles the “waiting period” for the mainframe batch job by using a timer step to trigger a new call to the bot once the job will have run.
- The bot then logs back into the billing system to search for the newly created test bill file and save it out to a pdf.
- PMG then pulls the file from its network location and emails it to the original requester.
Advantages to Automation
Clearly, there are significant advantages to the automated process over the initial manual process:
- No reliance on offshore resources to interact with PBS
- No manual swivel chair processes
- Eliminates back and forth between requester, billing team, and offshore resource
- Drives standardization – one repeatable way to handle this task
- Requester has visibility into process status throughout end to end process – no “black box”
Further, the solution was very simple to set up. Creating this process took less than four hours of configuration work within each system to complete.
But to achieve this outcome, the customer needed to use RPA + PMG (or other BPM platform) instead of RPA alone. Our customer purchased an RPA tool with high expectations of what it could provide for them, but has been frustrated with the pace of the automation projects. They are also unable to create end-to-end solutions. In our case above, the RPA tool had no way to quickly configure a user interface to accept the request. It had no way to expose real time status and progress to the requester. And in many cases, further integration would be desirable with upstream or downstream systems that is difficult or impossible in the confines of the RPA tool. But coupling that tool with the power of the PMG platform rendered those challenges trivial.
PMG Can Help
Our takeaway from the experiment was that many of the concerns and complaints (like in this article from HorsesForSources (HfS)) including overblown expectations and difficulty scaling, are very real for enterprise, but that RPA has a powerful part to play. We definitely agree with both McKinsey and HfS that to truly maximize the benefits of any significant technology purchase, a thoughtful program must be designed that takes varied drivers into account. However, for those looking to realize benefits from RPA quickly and get some initial ROI, pairing the RPA tool with PMG can absolutely provide significant advantages and give you something to be excited about once again.