Robotic process automation (RPA) had a big year in 2021. A blockbuster IPO, mergers and acquisitions and talks of upcoming events set the stage for an even bigger 2022. Market growth for RPA is projected to expand from US$1.7bn in 2020 to $9 bn by 2025 according to IDC.
Despite all the movement, there are signs that something strange is happening in the RPA category. For instance, the Sapphire Ventures CIO index reports that RPA is one of the top technologies that companies are hoping to spend less on in 2022. Troubling data suggests the software is not living up to its transformational hype and IDC reports that the overall enterprise automation market potential of $40bn dwarfs the $9bn projected for RPA, and Forrester points out that the category is plateauing.
These signs may be the reason top RPA vendors are feverishly expanding their offerings beyond task automation which was their initial core product. After all, the task orientation of RPA is anchored to bots mimicking human actions to find efficiencies.
If you think about it, this approach is another form of outsourcing, known as ‘botsourcing’. Transformation-minded enterprises are recognizing that they need to look beyond task automation to enterprise-wide opportunities and RPA vendors are trying to catch up.
RPA and enterprise transformation
To understand why RPA is not leading to transformative outcomes across the enterprise, we need to look at the long history of the technology. It emerged from the automation of software testing, where recording point and click tasks was fundamental.
Often referred to as screen scraping or macros, it is still great today for modernizing legacy ‘green screen’ tools enabling automation across on-premise hardware footprints. In the context of the modern world, however, the bot-based architecture faces challenges.
RPA does not scale well
Often companies struggle with broad automation initiatives based entirely on RPA. RPA bots capture user interaction (UI) tasks and they become more fragile as more get deployed. Significant time is required to maintain a large volume of bots.
Even if a platform has bolted on application programming interface (API) connectors, optical character recognition, artificial intelligence or process mining capabilities, the bolt-ons bring another layer of maintenance. ‘Maintenance-hungry’ automation impedes the pace of deployment and scalability.
It is difficult to govern across the enterprise
Success at the enterprise level requires solid governance, but RPA platforms inherently wrestle with organization-wide governance problems, both in IT and the wider business. Bots offer the advantage of quick efficiencies, but as their number increases IT-driven governance and security best practices become harder to implement.
RPA can turn into technological and economic debt
RPA is the best fit for quick tasks where something repetitive needs to be done quickly. The collection of bots will grow and so too will the maintenance needs. In essence, the army of bots demands another army of bot managers. While the bots were supposed to take a load off the humans, the humans end up needing to take a load off the bots.
General maturity cycles
Technological maturity follows predictable innovation cycles. The marketing of new technology regularly produces overenthusiasm for its capabilities. This is true across all industries, from the enterprise to the home kitchen.
In the wake of World War II in 1947, the 750-pound ‘Radarange’ was introduced to the American market for $2,000, promising incredible cooking convenience at the push of a button. It was not until two decades later that the microwave caught on and began selling millions of units per year. One thing the futurists got wrong in the 50s and 60s however, was the prediction that microwaves would replace professional chefs.
Advances in technological maturity do not come from copying a human process, but rather from something completely different. The transformation of the restaurant business was not the added convenience of microwave ovens or robot chefs, but rather modern refrigeration and supply chain innovation.
The end product is not what the futurists projected, but it is ultimately better for the consumer. Chefs gained the ability to import refrigerated goods from all over the world to delight customers with top quality cooking.
If we return to the world of enterprise software, we see organizations buying into task automation because the futurists tell them that is where the value lies. When the time comes for their contract renewal, however, they are choosing to spend less on RPA. This is a signal that businesses require something completely different to achieve the transformation that they need.
In a recent conversation with Daniele Tonella, former CEO of AXA Tech and UniCredit, he made a great observation that “digital transformation has been so frequently confused with buying technology and it never works”.
I fundamentally agree. I often tell business leaders that digital transformation is about transformation first, digital second. Digital transformation is the product of business transformation and how technology is leveraged to achieve outcomes.
Automation maturity models all too often focus on one technology, like an ‘RPA maturity model’ when the future of enterprise automation involves five more steps beyond RPA (automating tasks). It can be visualized in the complete model of enterprise automation maturity below.
- Automating business functions – Data movement and integration of systems to fully automating select functions, such as recruiting candidate management or VM provisioning.
- Automation across applications – Automating entire ecosystems of applications surrounding the foundational ones of the company, such as the customer relationship management, enterprise resource planning or APIs.
- Automating business processes – Automating end-to-end workflows across departments in a seamless way.
- Automate business decisions – Driving business decisions based on pre-defined or learned criteria.
- cognitive automation – Autonomous business operating with self-learning capabilities.
Enterprise automation maturity model (Source: Workato)
The entirety of this maturity process is centered around the user and it takes place across key parts of the organization, what I call the five towers of automation. These are customer experience, employee experience, supplier and partner operations, front-office systems and back-office systems. As organizations reach greater maturity, the user experience across all of these towers should improve and benefit.
Have we outgrown RPA?
With a clearer view of automation maturity, we can examine a few signs that an enterprise has outgrown RPA task automation:
- Scalability impedance – The enterprise has purchased RPA but have not been able to fully deploy across the enterprise, as the pace of deployment was hindered by fragility.
- Compliance and governance conformance – Multiple teams developed RPA bots with no enterprise-wide compliance or governance, impeding the extent of bot deployment.
- Maintenance demands – The business has deployed hundreds of bots, but is spending more time maintaining and fixing what was deployed than spending time expanding the program.
- Underwhelming efficiency outcomes – Bot-economics are not yielding the efficiencies for the business that were promised.
- Stacking up other tech – In addition to RPA, the enterprise is stacking up other technologies to expand automation capabilities beyond RPA’s limits.
Some who recognize the shortcomings of RPA have been thinking about what RPA 2.0 or even 3.0 will look like. In the context of automation maturity, this view misses the point. Microwave 2.0 or 3.0 did not transform the restaurant experience, it was something completely different – refrigeration.
The team at Morgan Stanley’s research arm point out that transformation-minded companies have outgrown the patchwork of technologies, including RPA, that passes for automation today. In this new era of automation, we are seeing enterprise automation emerge as a completely new category that will usher organizations into new levels of maturity, driving greater outcomes, improving customer experiences and fueling unexpected innovations in business automation. It may be a big year for RPA in 2022, but the decade belongs to enterprise automation.