Enterprise Automation & RPA: The Highways & Local Roads of Automation

In 1919, a young US military officer volunteered for one of the first cross-country road trips taken in the United States: a public relations exercise for the military with a convoy of 40 vehicles.?

Today, a road trip sounds unremarkable. But in the early days of automobile travel, it was an intense challenge. The route in 1919 was a mix of roads paved by novices and muddy horse paths. Conditions were so bad that they needed a group of motorcycles to scout ahead for road quality.

Decades later, the young military officer - named Dwight D. Eisenhower - would compare this experience to the Autobahn he saw in Germany during World War II, and go on to champion the American Interstate Highway System when elected President.

It is human nature to apply old thinking & approaches to new problems. Paving horse paths made terrible roads for automobiles - yet that was the state of cross-country travel in the US until the 1950s. When policymakers debated the new Interstate system, many proposed two-lane highways, which were conventional wisdom at the time. But Eisenhower pushed for a multi-lane system like what he saw in Germany because he knew how effective it was.?

Although the interstate system in the US accounts for 1% of all roads in the country, it makes up 25% of all traffic - and is directly responsible for $1.6T in economic value today.

Different types of roads → different approaches to automation?

Transportation is a great analogy for the processes we interact with in organizations every day. I’m not the first one to point that out. On occasion you will hear the phrase “paving the cowpath” (or horse path) in IT circles, and it communicates the same lesson Eisenhower learned a century ago: the old way is not always the best way.

This resonates in the automation market, where Robotic Process Automation (RPA) is popular. RPA relies on bots to automate tasks performed by a human clicking on a screen, so RPA automations mirror the manual work of individuals. While many vendors encourage customers to rethink processes before automation, the architecture of RPA is dependent upon human actions in order to define the steps in an automation.

I’m not saying RPA lacks value, but rather, it is an extremely limited view of the value of automation. In the early days of car transportation, regional communities - in their limited view - saw local roads and paved horse paths as transformative. Similarly, the rise of RPA brought automation to individuals or small teams looking to automate tedious day-to-day tasks such as processing invoices. But when it came time to take transportation to a larger, more transformative scale, a different approach was required. And the same is true for how organizations approach automation.

Although valuable for the automation of mundane individual tasks, RPA is not architected to handle the broader needs of the organization - just like local roads cannot form the transportation backbone of a region or nation. This is where enterprise automation comes into play. If RPA equates to the local roads or paved cowpaths of automation, enterprise automation is the regional or national highway system.

Since the nuanced differences in automation technologies can be challenging - even for experts - this transportation analogy is a useful way to break down the differences between enterprise automation and RPA and the challenges they are designed to solve.?

RPA = Local Roads

We all know our local roads like the back of our hands. They are familiar, useful, and sometimes illogical. Having spent the last few years living in Boston, I can attest to the third point. Local roads are designed to:

• Handle the lowest volume of traffic at the lowest speed
• Remain at the bottom of the transportation standards hierarchy - some are even unpaved and aren’t required to adhere to common standards
• Optimize access to property that highways and freeways can’t access

This optimization for access vs. throughput or standardization means if an important local road goes down, it causes problems for the surrounding neighborhood - but the entire system of the city and country can keep running.

RPA is almost the perfect parallel. It has been around for many years, after its initial popularity for software UI testing back in the 90s. By definition, RPA is designed to:

• Automate a task flow that a specific human does to save time on tasks: “Build, deploy, and manage software robots that emulate humans actions interacting with digital systems and software.”
• Remain at the bottom of the standardization hierarchy - RPA automations aren’t built according to common standards or tests but rather typically done in siloes by individuals or small teams.?
• Optimize automations involving data/systems that aren’t highly trafficked (don’t have APIs - RPA is great for automations involving documents, legacy systems)?

If an RPA bot is implemented as intended, and it breaks down (which happens often, known as “bot fragility”),the impact is limited to a small team. If it is implemented in a broader way, supporting processes that are meant for superhighway-level traffic - this bot fragility can translate into organization-wide breakdowns and a painful, unending cycle of monitoring and maintenance (performed by humans).

Enterprise Automation = Superhighways

Highways support a huge volume of economic value for regions and countries as a whole. Relative to local roads - the volume of traffic is dramatically higher. Only in catastrophic circumstances, like epic winter storms or plane crashes on freeways, do we get a taste of what it would be like if that volume of traffic were forced to be supported by local roads. (I’ll give you a hint: it’s really bad)

Highways are designed to handle millions of cars for decades - from the size of the highway down to the paving material, everything is standardized and tested to ensure consistent performance day in and day out for the population.

Just like with the maturation of our transportation system, we’re seeing a similar maturation in how companies are thinking about their own approaches to automation. “Most companies with automation are working on tactical or department-level projects involving RPA, low-code application development, and low-code workflow automation projects,” notes Morgan Stanley research analysts in an article titled The New Push for Enterprise Automation. “Now, the push for organizational resilience and real-time decision-making is further advancing automation initiatives. The next step: standardizing automation and moving toward enterprise-wide projects.”

These enterprise-wide projects are numerous: order-to-cash in finance, employee onboarding in HR, account-based outreach and messaging? in marketing/sales, and more. In each case, they involve multiple highly-trafficked operational systems, different teams, and cross-functional standard processes that occur all the time.

These end-to-end processes are the beating heart of the business - and like major highways, any problems encountered reverberate throughout the organization. It is for this reason that enterprise automation is becoming increasingly important - because the impact is massive (both positively when done well and negatively when done poorly or not at all). Businesses run on end to end processes, so at Workato we pour everything into reliability, governance, security, and other enterprise-grade requirements.

But the potential for enterprise automation does not end with the major processes. Thanks to an approachable, low-code/no-code approach, there are endless stories of companies using enterprise automation in creative ways across the organization - from marketing processes to building collages out of thousands of employee photos.?

Why Both are Needed

In a healthy economy, both highways and local roads have a role to play. In the enterprise, both RPA and enterprise automation also have a role to play. Tasks and end-to-end processes are all worth automating.

The market has been scrambling to provide bolt-on solutions to meet the new demand for enterprise automation through mergers and acquisitions. But bolt-on is not the same as built-in, especially when it comes to enterprise automation. Eisenhower discovered that building a highway over a local road is a misguided exercise. Building local roads off of highways, on the other hand, makes a lot of sense. I believe that the businesses that win in the next decade are already building their highways, and the results will only become more clear as digital transformation takes full effect.