Can't replace legacy software? RPA to the rescue

Sometimes the best fixes for developer shortages aren't elegant; they just work. That pragmatic lens is what many leaders will likely have to look through in 2022 as they try to pick up some quick wins by increasing investment in initiatives like low-code and no-code development and robotic process automation (RPA). As management asks employees to keep critical business processes rolling while the Great Resignation thins out office teams, the immediate appeal of RPA, in particular, will likely grow especially enticing.

Why RPA often makes sense

RPA can answer some of the major labor crunch problems facing managers in 2022, as it automates repetitive, manual computer work usually done by back-office staff. This type of work is generally the most unsatisfying part of office labor, and it's frequently performed by moderately or highly skilled people who could (and should) focus on higher value tasks.

RPA works by creating configurable bots that emulate the keystrokes and mouse clicks performed by people interacting with a software user interface. The most effective RPA bots tend to be the ones that automate those multi-step, structured tasks that are done manually again and again by workers. These activities often involve jumping from one legacy application or platform to another—sometimes called swivel-chair tasks—where a user must take data from one system and enter it again on another, non-integrated system. In other words, there's a lot of copying and pasting and a lot of clicking of mouse buttons.

Please read: Low-code/no-code tools are filling the tech talent skills gap

For example, a financial analyst may have to manually copy billable information from three or four different software platforms and then paste that into an invoice form within a legacy application every time they need to bill a customer. That type of situation would be ripe for RPA.

"Across our clientele, which consists of many household brands in Asia, the most common success stories come from the finance function," says Liu Siyong, founder and general manager of CFB Bots, an automation consulting firm with a strong focus on RPA. "The high volume of transactions in finance leads to a high volume of repetitive processing work. This, together with the serious consequences of any human error, means RPA helps finance teams not only boost productivity but also cut down risks and costs of noncompliance."

When done well, RPA can do wonders. A Salesforce study from December 2001 shows that 89 percent of RPA users are more satisfied with their jobs, and 84 percent say they are more satisfied with their company.

A short-term fix for a long-term problem?

But RPA is not always the prettiest route to automation. Most software and business process engineers would argue that using RPA applications shows a need for rethinking a software approach altogether. And many an engineer has complained that RPA bots are just Band-Aids that cover up fundamental flaws in software, integration points, and the underlying processes themselves.

They're not wrong. But the problem is that technical debt takes time to pay down, and process improvement takes even longer. You can't rip and replace everything at once. It's simply not practical.

The truth is there will always be legacy applications that don't play nicely with other applications. While the software world has stepped up its game with open API access and a stronger focus on integration, there are still plenty of walled gardens and highly customized deployments of enterprise systems that will remain siloed. And often, those systems aren't even run by the organization; sometimes, employees get access to crucial information within systems run by partners or customers that are unavailable to access via API or other types of non-RPA integration.

Please read: How composability makes IT automation simpler

"RPA can be thought of as a digital spine connecting all applications," explains Cem Dilmegani, founder of tech analyst firm AIMultiple in a recent guide to RPA. "Most large companies still use some systems built with 2000s technology. And legacy systems are not supposed to be providing integration to modern tools. RPA provides a solution to these problems. Software robots step up to bridge the gaps between systems." If you're game to get started, AIMultiple maintains a list of more than 50 RPA vendors.

So executive teams must plan for the future with software engineering and DevOps improvements that help them modernize and speed up the buildout of their digital ecosystems. Custom platforms and applications built around newly imagined business processes are truly at the foundation of sustainable technology improvements. But the business has to keep running while all of those longer range plans come to fruition, and RPA can play a significant role in triaging the crush of tedious work that's barraging crunched back-office teams right now.

"As RPA enables us to automate an existing process as is, this helps to circumvent a lot of the change management issues that arise with the typical business process reengineering," says Siyong, noting that RPA can make it possible to make early automation gains quickly, which can justify future, more sustainable transformations later. "RPA's deployment timeline is short—days or weeks instead of months or years—and is relatively low cost. The quick returns let companies better visualize the kinds of returns they can get with further digital transformation. This helps break down resistance to further digital transformation initiatives."

The outlook for RPA

There's no doubt that much of the industry is already seeing the value of RPA solutions. According to Gartner, the RPA software market is the fastest growing segment of all enterprise software. Recent projections by the Everest Group show that the RPA market will grow by as much as 50 percent over the next two years.

As Dilmegani explains, RPA has a lot of runway because of the ongoing lifecycle for creating and implementing sustainable automation and AI technology. As business leaders identify a new need, they communicate it to developers and then it takes time to engineer new applications. After the application rolls out, there's a learning curve to manage, bug fixes to deploy, and so on. While that progression unfolds, teams can use one or more RPA bots to fill the gap. Once the new technology is proven, the company weans itself off the bot and deploys the replacement.

Please read: Automation and artificial intelligence are the future of security

"I see RPA as a bit like spreadsheets. You don't want to run your critical processes on spreadsheets, but when there is a new, hard-to-define use case, [such as a] strategy team modeling whether to enter a new market, spreadsheets can be a great tool," Dilmegani says. "Similarly, there are new automation opportunities popping up before we have the time to build domain-specific, robust, externally maintained solutions. For those, we need RPA."

In this way, RPA is a healthy arrow in the broader quiver of process automation technology. In an era of labor shortages, it's a gateway toward long-term implementation of APIs, AI/ML algorithms, packaged enterprise applications, and mobile applications.

"RPA is usually the entry point for many organizations in their intelligent automation journey due to its quick time to value and low barrier to entry characteristics," explains Siyong. "Having experienced RPA, companies are more open to increasing straight-through automation involving things like cognitive automation and process reengineering."

In the meantime, RPA doesn't necessarily have to be a simplistic solution for the lowest hanging fruit. It can work hand in hand with technology like low-code/no-code tools and AI/ML to become a well-rounded option for many use cases.