Great Automation Expectations, Did RPA Deliver On Its Promise ?

It was nearly 20 years ago when the first instances of robotic process automation (RPA) began and kicked off the business process automation industry. Over the past decade, we saw greater adoption of RPA across every industry and geography. The promise was to provide enterprise process automation, but with increased adoption, the promise of scalability and implementation beyond proof of concept left many CEOs disappointed. In short RPA had talked a big game and simply wasn’t living up to the high expectations it had set. HFS states, “RPA hasn’t inspired enterprises to rewire their business processes – it’s really just helped them move data around the company faster and require less manual intervention.”

The higher the expectation the greater the disappointment when it doesn’t deliver. Robotic Process Automation can provide a lot of value for an enterprise, it’s just not the silver bullet it presented itself as.

With over 200 RPA providers in the market today, the global robotic process automation market was 1.40B in 2019 and its expected to soar to 10.7B by 2027. Deloitte expects near universal-adoption within 5 years.

What is Robotic Process Automation?

According to the IRPA, Robotic Process Automation (RPA) is the application of technology that allows employees in a company to configure computer software or a “robot” to capture and interpret existing applications for processing a transaction, manipulating data, triggering responses and communicating with other digital systems.

Features of Good Robotic Process Automation Tool

Imitates Manual Tasks

RPA bots are far more accurate than their human counterparts for manual entry of data or copying data from one source to another or populating forms. This greater accuracy translates into increased productivity.

Scale Digital Workforce

One point of clarification here, scaling digital workers, or bots is not the same thing is scaling automation across the enterprise. Digital workers can be scaled according to their determined business rules, but scalability across use cases, business units is not guaranteed or initially successful.

Lower Barrier of Entry

Many RPA products require little or no coding requirements, so it’s attractive to business users. Plus, RPA can sit on top of existing legacy infrastructures, making integration easier.

Greater compliance

For industries that are compliant heavy, RPA can be programmed to adhere to the strictest guidelines, making it ideal for mitigating risk.

24/7 Workforce

Another attractive benefit is RPA bots work 24/7 and don’t take off for holidays or fall sick , perfect for enterprises with high volume repetitive tasks. It doesn’t interrupt human-focussed business processes.

Greater Employee Productivity

By including employees in the implementation of RPA, adoption is easier, and employees will appreciate the reduction in manual, mundane, repetitive tasks and focus on higher-value work.

Features of a Better Robotic Process Automation Tool

Multi-Tenancy: Allowing both human and digital workers to work together on the same desktop, at the same time. This eliminates the problems of siloes and drives greater efficiency.
Configure Multiple BOTs: While some RPA products let you build a BOT at a time, QueenBOT RPA allows you to build, operate and manage up to 50 BOTs
Auto Failover: Since BOTs can break, its ideal to have a platform that’s intelligent enough to recognise this and start again. Auto failover also ensures self-recovery from infrastructure failures, ensuring business continuity.
Real-Time Digital Workforce Management: This means your BOTs are never idle. You can ensure real-time management and deployment of multi-skilled BOTs to undertake many different tasks, without human intervention.

Multi-Skilled BOTs: Profile enrichment means BOTs can be assigned a specific profile which defines a process or activity. This means your BOT can be processing claims in the morning and invoices in the afternoon, so you can proactively manage your digital workforce.

Reallocation: In situations where the workload is unevenly distributed, you can reallocate BOTs to overwhelmed processes to ensure ongoing and efficient operations.

Five Ways That Robotic Process Automation Didn’t Measure Up

1. RPA didn’t automate processes, it automated tasks.
Task automation is helpful but automating a singular task or part of a process, didn’t address the issue of automating an entire process. Scalability is perhaps the biggest issue with robotic process automation. Deloitte reported in 2018, only 3% of organisations were able to scale their RPA solution. Computer Weekly reports, a recent Blue Prism study discovered 73% of Singapore enterprises had difficulty scaling. And as of January 2020, only 8% of enterprises are automating at scale. With end-to-end process automation or straight-through processing requirements, you might need to look at other automation besides RPA to address this.

2. RPA doesn’t always improve process efficiency.
A very typical example is when an organisation looks to automate a particular process but doesn’t actually know what steps are involved in that particular process, which means the less RPA can effectively improve accuracy or productivity.

3. RPA can’t fix your data problem.
RPA can only use what data is digitised and available. If enterprises have data stored in paper documents, handwritten notes, forms, files it can’t be used for automation. Typical capture technology, like Optical Character Recognition (OCR) digitises neatly formatted, readable, non-variable documents. This structured data comprises about 20-30% of an organisation’s data. What has happened to the other 70-80%? It’s not included unless it’s manually entered into a database. (Let’s hope it was entered in accurately). So, while the expectation is that your RPA product can move accounting data from local systems into your ERP, in reality it’s likely all the unstructured data wasn’t included.

4. RPA doesn’t handle change well.
Just like us, BOTs don’t typically handle change well. We typically fumble a bit to adapt to changes. RPA functions much in the same way. Changes in the way data is formatted or read can be an issue. If the RPA product can’t read or understand it, its left in the great black hole where it disappears. And BOTs break. And speaking of change, the popularity of RPA also worried employees, whose concern was that a robot might replace them.

5. RPA isn’t learning.
We come to expect with Machine Learning, the technology gets smarter as it repeats tasks, but it’s not quite the same with RPA. You must provide the instruction; RPA will carry out the task(s). When it comes to making decisions or learning your processes, even with complex business rules, RPA is just not that smart.

6. RPA Implementation is not so simple.
Forbes council article explained setting RPA up is not so simple. Requiring upfront and ongoing investment of money and resources comes as surprise for organisations. With the hard coding requirements, it takes data analysts or other subject matter experts to set up robotic process automation, even as a proof of concept for use case testing.

Questions to Ask Before You Invest in Automation

While RPA can be viewed as a short-term solution to fix or automate certain tasks, a longer-term vision is required for sustainability.

What should be automated?
You can automate any number of tasks within your organization, but what makes the most sense? We’ve seen finance and accounting, specifically invoice automation as being a prime candidate for a use case but drilling down to specific business challenges is required. It could include moving financial data from a local server into your ERP.

How are you going to automate it?
This is a tough question and depending on the complexity, it might require a few tools to achieve automation. For example, do you know all the steps in the process? Is that documented? You might have to use a Process Discovery tool to map out all the steps. What about the data you want to use? What format(s) does the data arrive in? Paper? Emails? Does your staff have to manually key in information? Or is manual entry outsourced? Do steps in the process require higher levels of decision-making? You might need to consider an AI solution that includes Machine Learning to tackle this.

Does this automation investment align with our long-term business objectives?
Boosting productivity may not include reduction of headcount, but reassigning valued employees to higher-value work, such as customer service can prove to be a better business decision.

The Future of Robotic Process Automation

Even with some shortcomings, RPA remains a highly consumable technology. According to Deloitte, 72% of organisations will have started their RPA journey by 2020. In addition to the increased adoption, the dollar investment in RPA will continue to grow, Forrester predicts that the RPA software market will hit $2.9 billion by 2021—up from just $250 million in 2016.

And the global pandemic has only accelerated this digital transformation adoption. Enterprises might rush to implement RPA as employees work remotely but that can cause more confusion and frustration for staff.

The Addition of AI to RPA

So instead of just relying upon robotic process automation as the only tool in the enterprise toolbox, companies embed AI technology to make RPA work better and become smarter and shorten the training times.

Forrester states that AI is now an integral part of RPA software in their recent report, Ignore the AI Hype to Select the Right Robotic Process Automation Tool. Just a few years ago, there were very few RPA providers that included AI technology in their offering. Now many RPA providers include AI, either through partner providers or developing the technology in house. With the competitive marketplace, customers can choose RPA providers that are experts in their specific industry such as healthcare or financial services.

According to Forbes, as the product matures and with greater adoption over time, RPA will become a commodity. With this greater adoption, smaller organisations will be more likely to purchase it, as it won’t require IT investment. And generate new specialised positions, “market need will arise for a new service provider: robotics-as-a-service (RaaS) operator.” These specialists will focus on specific market segments. Meanwhile, enterprises look to make RPA tools more efficient, such as data digitisation of unstructured documents and data. Or outsourcing operational tasks to other lower-cost providers.

In Ignore the AI Hype to Select the Right Robotic Process Automation Tool, Forrester predicts RPA will move under the AI umbrella, that AI will drive the smarter automation, maintaining the digital workforce. As user communities grow, there will be more sharing of data sets to shorten implementation time. And AI providers such as Google and AWS will jump into the automation marketplace.

From Product to Platform: Intelligent Automation

Intelligent Automation combines robotic process automation with artificial intelligence to accelerate business processes for digital transformation. Intelligent Automation is a combination of various automation technologies to bridge the gaps with RPA. The AI component, along with Machine Learning helps to understand and learn your processes, becoming smarter over time. The Integrated Automation Platform takes it one step further with the inclusion of a multiformat data ingestion product that enables enterprises to use more of their data for their automation programme.

The Future of Automation

Predictions include growth of pure-play RPA providers via strategic acquisitions, from cloud to task-mining to artificial intelligence. This evolution of competitors will be available to address the broader enterprise need for automation. Even with this broad product portfolio, automation software providers will still need to create the right automation recipe for enterprise customers.
It’s an opportunity to build software that allows clients to customise their automation tools, especially as their automation journey changes over time.

Analysts agree that while the adoption of Robotic Process Automation accelerates, there’s no one tool to address transformation in its entirety. So, it’s best to think of RPA as not a platform per se, but a tool in your automation arsenal needed for enterprise automation.