In a memorable scene from the movie “The Founder” about the origin of McDonald’s, the McDonald brothers plot the layout of their restaurant in a life-sized mockup drawn in chalk on a parking lot.

This example of process optimization was certainly “lean,” but it involved no software whatsoever. Today, in contrast, optimizing business processes almost always means automating them – at least in part. And when we say automation, we mean with software.

Just what software, however, is an open question, as today’s frothy software marketplace has spawned several contenders. From the business process automation or BPA of the last decade to today’s robotic process automation or RPA to the latest entrant, digital process automation or DPA, information technology decision makers have a plethora of options to choose from.

Be warned: This is a clear-cut case of caveat emptor. With the help of the big IT analyst firms, the providers in these overlapping categories have stirred up massive confusion. Let’s clear up the biggest misconceptions.

Myth No. 1: Digital process automation is more ‘digital’ than business process automation

Now that DPA is all the rage, the BPA vendors of old have all relabeled themselves DPA. But has anything really changed?

Peel away the marketing hype, and you end up with some notion of “digitizing” a business process. Only digitizing is the wrong word: We digitize analog information by putting it into ones and zeroes, but that’s not what we’re doing with our processes.

Rather, what vendors mean by digitizing a process is really digitalizing a process, which means reworking it as a process that software can automate, at least in part – which is essentially what we meant by BPA all along.

In fact, there’s nothing particularly digital about DPA, or anyway, any more digital than BPA has always been.

Myth No. 2: Today’s process automation is about taking humans out of the process

Automation has been with us since the Jacquard loom in the early 19^th century, and for every wave of innovation, people fear that automation will put them out of their jobs.

To be sure, a few people always end up losing their positions, but for the most part, people end up with different ones. Weavers became loom operators, and so forth.

Today, there’s a common misconception that our software is so smart and so powerful that process automation will finally take people out of the loop – only that’s not what’s happening. Just as with the early days of the Industrial Revolution, people are still involved.

As the technology improves, it takes over increasingly sophisticated tasks from human operators, thus giving people more interesting, valuable work to do. Process automation is more about empowering people than putting them out of work, just as it has always been.

Myth No. 3: AI solves all the problems of process automation

Perhaps we simply need smarter software, like the robots in RPA, right? Just one problem: They’re not AI. In reality, RPA isn’t smart at all.

The idea of RPA is for software to mimic a human’s interaction with some old application or two, by clicking buttons, filling in forms, cutting and pasting information, and the like. Such automation can be quite sophisticated, but for RPA to bring AI to the table, it needs to be cognitive RPA – another thing entirely.

However, today’s cognitive RPA has only a few narrow strengths. “Enterprises are beginning to employ RPA together with cognitive technologies such as speech recognition, natural language processing, and machine learning to automate perceptual and judgment-based tasks once reserved for humans,” according to a report by Deloitte.

The problems with RPA, however, aren’t that the tools aren’t “smart” enough. Instead, the challenge is more about resilience – dealing with largely unexpected changes in the IT environment.

Adding cognitive capabilities to RPA doesn’t solve these resilience issues, you simply end up with smarter technology that is still just as brittle as before.

The end result: a surprisingly narrow set of use cases where RPA – or even cognitive RPA – can provide substantial business value. “You should consider using RPA if you have a large legacy application as part of a process that functionally works, has no bugs, doesn’t need new features, [and] doesn’t require developing additional applications to support the process,” warns Keith L. Murphy, solution architect at OutSystems.

Perhaps the problem is more with RPA than with AI itself. If we expand our focus beyond RPA, then, can AI help us with process automation generally?

Not so much. The state of the art: several products on the market now that use AI to offer “next best action” advice that essentially acts as an autocomplete for building workflows. As a human assembles a workflow, the AI suggests the next step.

We all like to make fun of autocomplete. Do we really want it telling us how to run our businesses?

Myth No. 4: Digital process automation and low-code are separate markets

In many ways, DPA is the next generation of BPA. But peel back the marketing verbiage, and you’ll find that most of the DPA products are actually low-code platforms.

Low-code represents a visual, model-driven approach to building software that requires little or no hand-coding. Many low-code platforms use a “drag-and-drop” approach to building apps to support the construction of workflows. After all, what better way to construct a flowchart for a workflow than dragging boxes and connecting them with lines?

In fact, many of the low-code products on the market began their lives as business process modeling (BPM, which can also stand for business process management) tools of one sort or another. Evolving a BPM tool into a low-code platform thus adds the automation part of the story, as the product of using such a platform is a working app, not simply a model of a process.

Where, then, do DPA tools fit into this equation? In essence, DPA is another name for a low-code platform that began as a BPM tool – or at least, a low-code platform that specializes in BPA. “For digital business, the decades of challenge to enable continuous improvement for ‘process’ and the aspiration to do so with no code remain,” explains Jesse Shiah, chief executive of DPA provider AgilePoint Inc. “The challenge will only exacerbate when one created not tens or dozens, but hundreds to thousands of DPA applications.”

Why, then, do we have separate DPA and low-code markets? You’ll have to ask the big analyst firms. My theory: The low-code market was getting too crowded, so Forrester split off DPA so they could feature process-centric low-code platforms separately from the others.

Meanwhile, Gartner completely missed the mark by insisting that all these various platforms and tools were really all part of the application platform-as-a-service (aPaaS) market – even though many of them run on-premises as well as in the cloud.

Myth No. 5: The distinction between ‘wide’ and ‘deep’ digital process automation makes sense

Forrester may have a better grasp of what’s going on in the low-code and DPA markets than Gartner, but it still struggles with which buckets to sort the various vendors into.

This struggle recently led them to the notion of “wide” versus “deep” DPA. “Wide” referred to the low-code DPA vendors and how their tools are well-suited for building larger numbers of simpler applications, while “deep” designated those companies that are better-suited for building large, complex applications.

Just one problem: These two categories aren’t separate at all. True, some low-code vendors focus more on larger apps while others focus more on smaller ones, but most of them are perfectly well-suited for the other type.

And in any case, the notion of “large” is beginning to lose its meaning, as the best practice approach to such apps is to build them as assemblages of modular components – components that a “wide” DPA platform would be perfectly adept at building.

The future of process automation

The current state of the art, it seems, consists of vendors struggling to understand the evolving process automation value proposition their customers require, while the analyst firms are mostly at a loss as to how to describe these intertwining, emerging markets.

The real wildcard in this story, however, is AI. Today the role AI plays in process automation is mostly hype, but tomorrow, expect AI to recognize the best way to automate a process as well as it can recognize photos of cats – which, by the way, it does quite well today.

With all of this discussion of automation technology, however, one fact remains steadfast: Business processes are mostly about humans and the work that they do. Process automation technology will continue to improve as it has for centuries, but the real story here is a human one.

Jason Bloomberg, a leading IT industry analyst, author, keynote speaker and globally recognized expert on multiple disruptive trends in enterprise technology and digital transformation, is founder and president of agile digital transformation analyst firm Intellyx. The firm publishes the biweekly Cortex newsletter, advises companies on their digital transformation initiatives and helps vendors communicate their agility stories. Bloomberg, who can be followed on Twitter and LinkedIn, is also the author or coauthor of four books, including The Agile Architecture Revolution. (Disclosure: AgilePoint is an Intellyx customer. None of the other organizations mentioned in this article are Intellyx customers.)

Image: geralt/Pixabay