© 2019 SourceMedia. All rights reserved.

8 guidelines to help ensure success with robotic process automation

Business leaders are increasingly attracted to robotic process automation for its ability to automate repetitive tasks and reduce labor costs. But they often overlook how RPA can improve or open new revenue streams, according to research firm Gartner, Inc.

In fact, RPA can be a critical component to effectively drive business transformation. Gartner predicts that by 2019, process automation will have limited impact on top-line growth for 50 percent of organizations that will erroneously focus on labor reduction, not on improving business outcomes.

Gartner analyst and research vice president Cathy Tornbohm recently published research focusing on eight guidelines that can help keep RPA projects from going off track. Information Management asked her to share those guidelines, and to discuss the reality of RPA and how IT leaders can position their organizations for success using RPA.

Information Management: Most of the attention being paid to robotic process automation seems to be around the areas of automating repetitive tasks and reducing labor costs. But Gartner reportedly says that most organizations are ignoring the potential of RPA to create revenue streams. What are those revenue streams and why do so many organizations miss out on them?

Cathy Tornbohm: Acquiring, retaining and satisfying businesses / consumers are a must in order to extend and protect revenue.

Examples of this include having flawless initial contact with the correct order and bill. However, many companies with legacy business practices and systems are just not very good at these types of things. Human beings only go so far: kEy$ng (get it?) accurately between systems is not necessarily one of the skills we humans have perfected. As a result, errors occur and time is wasted, business is lost, wrong amounts are charged, discounts are not capitalized on and so forth.

RPA q and a.jpg
A technician works near rack servers in the server room of the Sberbank PJSC data processing center (DPC) at the Skolkovo Innovation Center, in Moscow, Russia, on Tuesday, Dec. 26, 2017. Sberbank PJSC, Russia’s most valuable company, will boost its dividend payout to 50 percent of profit or higher, just not as quickly as some investors had hoped. Photographer: Andrey Rudakov/Bloombergg 

These are the trap doors of inefficiencies, which RPA can help organizations avoid. Obviously if organizations do not have these issues then they don’t need RPA, yet still many organizations miss out on maximizing revenue streams mostly due to lack of personal incentive of management teams (the main roadblock to process improvement).

It is hard to breakdown silos and re-organize work streams just as it is hard to redo IT architectures to make it flexible enough to move with acquisitions and divestments. So the “missing out” is relative to an organization’s specific “accidents of history.”

IM: What role can RPA play in digital transformation efforts?

Tornbohm: RPA is one tool in your tool kit that can be used if appropriate on a digital journey. Gartner says “RPA [is a] catalyst or distraction” meaning RPA might be positioned as catalyst for change or a distraction from systems in the form of a replacement or process redesign.

For example, many small banks have no other way than the UI to enter data into core banking systems, so they have people manually re-keying the data in from the website transactions. RPA is an option to perform a type of integration in this scenario.

IM: What are best practices for formalizing an enterprise automation roadmap and process evaluation questions?

Tornbohm: The first step is to find out what really goes on day-to-day in your organization. It is very surprising how many variants of processing can build up.

Use process mining, process discovery tools or consultants to figure out what you actually do in a process. Methods to do so might include extracting systems logs, or mouse clicks and keystrokes to find out how many ways an activity can happen and then eliminate the less optimal ways to automate the most common paths.

Many different tools can be used to support automation, especially ones that have best practice processes already in them. Filtering to see if RPA should be used needs to start with understanding the process in order to then understand the choices of automation available in the short, medium and longer term.

If people don’t know what they do or how they do it, they’re not ready to start with RPA. Standardized, repetitive, re-keying tasks of digital data is the optimal place to start thinking if RPA makes sense or not.

Setting up a service catalogue of bronze, silver and gold level services can help make successful deployments. This should be managed by an automation center of excellence can allow this center to support activities being converted into RPA with a variety of services, from hosting and secure systems access through to full process redesign to improve business outcomes.

IM: What are some business scenarios and processes that are especially well suited to RPA?

Tornbohm: RPA can be considered in situations where organizations have found that other integration or automation options are too expensive or too time-consuming. Use cases for RPA include instances when an organization wants to work with structured data to:

  • Automate an existing manual task or process with minimal process re-engineering.
  • Reduce or remove head count from batch data input and output tasks or data rekeying.
  • Link to external systems that cannot be connected to through other IT options.
  • Avoid major system integration projects or specific new major application deployments.
  • Replace individual "shadow or citizen IT" desktop automation with enterprise wide automation.

IM: You recently published 8 key guidelines on how to keep RPA projects from going off track. Please summarize those eight guidelines.

Tornbohm: Certainly. They are:

Guideline 1: Formalize an Enterprise Automation (EAR) Roadmap

Organizations need a plan and a roadmap for not just customer journey mapping but money journey mapping through their organization. An EAR is a great start to understand which types of automation software are available, which processes they are dedicated to, and when employees can use them.

Guideline 2: Aspire for Revenue Generation, but Act on Short-Term Results Rapidly

Most people are excited by the “low hanging fruit” saving FTE hours, but this is proving patchy in most situations. An organization needs to have greater ambitions to deliver automation at scale, which involves taking important and difficult organizational / IT decisions and making big change happen.

Guideline 3: Formalize IT Teams' Involvement as Early as Possible to Maximize Business Outcomes

IT teams need to make sure that the right tool is selected for the task at hand. If you can fix an issue in ERP in a matter of weeks at minimal cost, then do so. IT needs to help with ensuring not every un-automated task goes into an RPA tool; this will simply build up technical debt and not necessarily bring “best practices” to the table.

It is also key that RPA is deployed securely with excellent governance so that nothing changes in the allocated robots pathway when it is running a particular script/automation. IT needs to understand that RPA can be a great way to deal with “accidents of history” and to open up a conversation about the interactions with the process owners so that they can discuss when RPA or something else is the best choice - now or for future scenarios.

Guideline 4: Create a Joint IT and Business RPA Team

This team should be comprised of (1) operations “business” function leaders, (2) advisors on which tools to use and (3) individuals who know how to build and maintain scripts as well as optimize the run time of the robots.

Guideline 5: Get External Help and Build Internal Skills for RPA

There are many specialist consultants and systems integrators who are keen to help with methodologies and best practices. It is essential organizations train up their own people and practice agile quickly in initial insight deployments rather than hand it over to the externals and hope it works.

Guideline 6: Build the Business Case for RPA

The easiest ROI to count is reduction in FTE hours on the task at hand. The wider process implications are the benefits of quality and accurate process, as well as the lack of rework needed. Additionally, RPAs can help determine the legacy systems to phase out, another factor into the ROI.

Guideline 7: Communicate What RPA Means to Your Organization

It depends on what your organization needs RPA to do. If you are growing fast with inefficient back office processes or have multiple people working overtime for reasons attributed to lack of automated tasks, then RPA could be one path forward. If this is not the case for your organization and you find yourself on the verge of significant layoffs, seek specialist advice.

Guideline 8: Visualize the Cycle of a Robot Deployment

Building optimized scripts is an art and a science, so IT can help optimize scripts and build exit plans for some of the scripts when they are no longer relevant.

IM: How can an IT leader best position their organization to take advantage of RPA for real results?

Tornbohm: IT leaders can best position their organization to take advantage of RPA for real results by finding out what you actually “do around here” and helping align real business outcomes with RPA deployments. This is done through clear governance, secure deployments and an understanding of the tools capabilities and limitations.

If the tool is paired with OCR or machine learning, it can process more work and IT can maximize the process flows / use of existing applications. In Gartner’s view, the best way to think of deployments is that the bot is the digital worker and it can process many scripts/sets of instructions.

Your employees will probably have a party to welcome their new “bot” colleagues as it they will take their dull work away. Of course, they must realize the bot can have days when it fails too, so the team needs to be robust and multi-skilled to support processes in those instances.

For reprint and licensing requests for this article, click here.