State of RPA & Automation Nation - January 2021 Issue

Welcome to the “State of RPA & Automation Nation” where KGAutomation offers monthly perspective on how things in the automation space are and where they’re going. In our inaugural post, we look back on 2020 as a catalyst for massive changes coming in 2021. Keep an eye out for our weekly (and shorter) posts as well!

2020 has come and gone but automation trends are here to stay. As many analysts have pointed out, last year was a catalyst for digital transformation as businesses faced increased pressure to do more with less. If we’re looking at automation technology platforms from an adoption curve, KGA believes that 2020 will be the inflection point towards mainstream adoption. Over the course of the next year, prospective customers will need a significant amount of education and training on how to pivot their digital transformation initiatives to keep pace with an incessantly evolving automation player landscape and meaningfully apply the proper technology for relevant and sensible use cases.

According to Deloitte’s Intelligent Automation Survey, 73 percent of respondents said their organizations have embarked on a path to intelligent automation which is a 58 percent increase from the number reported in 2019. The key word here being “intelligent” as enterprises look to move beyond the proof of concept tinkering phase with Robotic Process Automation (RPA) and use technologies which squeeze as much value out of automation solutions as possible.

It’s important to understand that intelligent automation solutions are not necessarily technologies with interwoven AI capabilities. Does artificial intelligence have a role to play? Yes. But applying AI in a meaningful way requires human intelligence over everything else.

RPA gets a bad reputation because of the use cases it has been applied to historically. For example, there are case studies galore on RPA within the enterprise finance function for use cases such as accounts payable, accounts receivable, invoice reconciliation, etc. While RPA has had positive ROI in these areas, traditional process analysts might ask “why is there a reconciliation process at all? How come exceptions have become the majority of process outcomes?” This is why you see the advent of process monitoring, orchestration, low-code, and process mining in figure 1 above. RPA was a blunt wedge that was driven into a larger ecosystem of problems and now enterprises need to deconstruct that ecosystem to ensure solutions are designed, developed, tested, and deployed in intelligent ways which are conducive to and facilitate non-technical user maintenance.

That is all just a fancy way of saying that we all need to automate smarter, not harder.

Shameless plug:  KGA’s Organizational Design & Integrated Solution Architecture (ODISA) approach and methodology enables practitioners and teams to achieve this outcome. We’ve been customers and have learned from mistakes and obstacles so that you don’t have to. Back to your regular programming…

That Sounds Like a #2020Problem with a #2021Solution

So what are technology providers going to do to market their tech to prospective customers and convince budget owners to crack open their seemingly impenetrable and mysteriously shrinking COVID era coffers? Process mining, cloud hosted automation-as-a-service, and citizen development.

Enterprise Scale Process Mining

It isn’t terribly new, it seems pretty complicated, and it creeps people out. Process and Task Mining is meant to provide unprecedented insights into what used to be called time and motion studies as performed by management consultancies.

In the current market, vendors like FortressIQ, Celonis, and Minit are focused as standalone process mining solutions which create unprecedented insight into an enterprise’s operations and processes while traditional RPA providers like UiPath (Process Mining) and AutomationAnywhere (Discovery Bot) entered the market to identify use cases to drive RPA implementations (more on that when we discuss automation as a service).

These technologies used to mangle together audit logs and trace unique identifiers’ journeys through applications to illustrate business processes, but AI enabled tech like computer vision has allowed for rapid functional evolution of these platforms. Platforms like FortressIQ now offer its users unprecedented empirical insight into business process and behaviors in a non-invasive way with highly actionable outputs not isolated to designing and developing RPA use cases.

In the cost cutting COVID-era, management find themselves asking “what exactly does that team do?” Maybe they ask this to find heads to cut, perhaps it’s because process docs and SOPs haven’t been updated in years, maybe because internal audit is coming for compliance checks, could be to eliminate redundant technology licenses/platforms, maybe it’s out of pure curiosity. Whatever the reason, transparency into business operations has never been more opaque in recent years and getting a good look under the hood can take months if not years. By the time most Office Spacing exercises are done, processes and staff roles have organically evolved rendering the manual business analyst outputs antiquated. And that’s assuming that the anecdotal process owner feedback was useful and honest under the pressure of job insecurity.

Successful process mining programs invite participants into observation cycles and commit to using the observed data with good intention such as identifying technology gaps, connecting broken holistic processes, and offloading tasks ripe for automation and/or offshoring. Today’s process mining technology is Orwellian at first glance and has the potential to be so, but this is where human intelligence is required to ensure that value is realized from the proper application of artificial intelligence and the consequent data ocean of observed data. Security also has a massive role to play here but we’ll talk about that in a more focused subsequent post.

Customers often tell KGA that they invested in process and task mining software, but have no idea where to start with the mountain of data they’re now sitting on. For now, these technologies are incredible at identifying tasks performed, process permutations, and useful data on those things like average duration and systems touched; but that may not line up with the ROI or value goals from budget owners. This is why it’s important to enter into 2021 with specific target outcomes such as generating use cases for business process optimization, offshoring, automation (citizen led + COE-led) , and re-shoring through automation.

While having an objective process mining technology provider, their outputs need to be presented to a gateway for assessment directly into a managed pipeline which is the play current RPA providers are gunning for with their own process mining solutions. For example, UiPath Process Mining feeding use cases/ideas to UiPath Automation Hub. Enter Automation-as-a-Service right on cue.

Automation-as-a-Service

From 2016-2019, there were two apparent competing technologies: “Robotic Process Automation” and “Intelligent Process Automation”. RPA was siloed to automating ‘dumb’ tasks like entering data from Excel to an old legacy desktop application while IPA was cloud hosted and API/integration based enabling it to exist in a more solution agnostic environment. RPA behemoths like UiPath, AutomationAnywhere, and BluePrism now faced competition from nimble, tech forward start-ups like Catalytic, Tonkean, and Unqork. Then in 2019, Gartner hit us with the concept of hyperautomation.

Both camps embodied aspects of ‘hyperautomation’ with the RPA camp developing product capabilities like AI Fabric while the IPA camp began integrating with traditional RPA tech like Catalytic x AutomationAnywhere to orchestrate automation between bots, people, and systems. The line between “RPA” and “IPA” continues to get blurred each passing day as traditional digital robots become more intelligent thanks to both sides of the spectrum.

In 2021, we’re seeing product releases coming forth which not only move legacy RPA infrastructure to cloud hosting but also integrate previously disconnected products throughout the automation lifecycle. This brings a new layer to orchestration as the traditional Orchestrator solution moves beyond managing bot packages, processes, and jobs to managing the entire automation solution lifecycle and associated components as well as an enterprise’s entire automation program. For example, integrating process mining capabilities into use case assessment/pipeline and assigning those approved use cases to citizen developers or COE developers then following the typical path to production in a fully integrated ecosystem with immediate oversight. Let’s simply call this the Enterprise Automation Platform.

As we see providers compete to deliver this type of end-to-end Enterprise Automation Platform, we see endless opportunities but we also see the traditional automation COE and program going through its own transformation. My oh my, how the turntables have…

Automation COE’s will be pressured to move to provider cloud hosted infrastructure to mitigate legacy customer hosted infrastructure, maintenance, and upgrade costs as well as to take advantage of integrated products which were previously a la carte. While this will offer better governance and ease of use within an enterprise’s automation operating model, COE’s identities will be challenged as automation-as-a-service further democratizes the technology and increases the throughput potential and pace from discovery to design to development to deployment. Successful COE’s will need to lean into automation-as-a-service Enterprise Automation Platforms (and their early release bugs) to pivot towards deeper evangelism and training within their organizations.

Citizen Development

Automation technology providers have zero interest in being professional services firms (that’s why KGA exists). An increased use case pipeline thanks to automation-as-a-service means that more automation experts and development resources with applied practical experience will be needed (that’s KGA’s competitive advantage). But enterprises have the opportunity to crowdsource expertise and development from their own people (that’s how KGA leaves clients with the momentum to be self-reliant).

Traditionally, Citizen Development has been defined as providing non-technical business users with the ability to automate day to day business processes with automation tools. This effectively enables subject matter experts and process owners to automate many small tasks which account for a large ROI by sheer micro-use case volume thus enabling the COE to focus on the few highly complex and high ROI use cases.

In 2021, Citizen Development will need to go beyond teaching a few folks how to dabble in low-code/no-code solutions. This type of re-training of the workforce will need to be captured by firms like Volonte as employees need to be brought along for the enterprise’s digital transformation initiatives from day one onboarding and through ongoing career development reiterating the need for human intelligence to drive automation strategies through the elevation of human potential.

It’s easy to say that 100% of the workforce needs to go through some form of re-skilling and training to coexist with the future digital workforce, but executing on that type of a vision seems like an insurmountable task with a litany of requirements such as RFP’ing for technology partners, dealing with change management and communication plans, career framework decisions, and thoughtful learning path curation.

We see the 1-9-90 rule as a decent framework and place to start where the 1% would be the Automation COE, the 9% would be traditional citizen developers, and the 90% would be the rest of the organization having gone through some form of automation consumer training as a foundation. We’re believers in incremental progress over perfection then leveraging learnings to further define your organizations automation journey.

This framework is conducive to driving large-scale digital transformation via a cultural shift from COE-led solutioning from the top-down to enabling grass roots movements within the automation consumer and citizen developer tiers. See the previous thought on COEs’ imminent identity crises.

In 2021, automation programs need to evolve to become the enablers of the larger enterprise and acting as the broker of mass distributed information, learning & development, and governance so that long stalled digital transformation initiatives can be revived with true believers who feel and see the impact of the transformation strategy. Citizen development is the immediate vehicle for this change.

Okay, Let’s Wrap It Up…

Okay, that was a lot and we want to thank you for making it this far!

The point we’re trying to make in this inaugural blog post is that RPA is an evolving state within the federated automation nation and we’re going to see that federation become a much more seamless experience in 2021 with human intelligence at the center of it all thanks to trends such as process mining, automation-as-a-service, and citizen development.

Our clients know this and we want you as our reader to know this: human intelligence is the secret ingredient to a successful automation program even in the world of burgeoning artificial intelligence capabilities. We’re looking forward to providing you with deeper insight on how you can enable your enterprise to achieve an intentional and deliberate automation program with organizational design and integrated solution architecture at its core.