What is hyper-automation?

Hyper-automation is a new term, and, as with any new term, its meaning can be fuzzy. I've been doing (hyper-)automation for the last 10 years, so here is my definition of it (although, I'm not pretending for academic precision here):

Hyper-automation is a way of boosting work productivity of an organization by replacing manual tasks with software automated routines wherever it's possible and makes sense. Unlike traditional software automation, hyper-automation targets many more manual tasks because the cost and effort of automation has been significantly reduced thanks to no-code/RPA software and the proliferation of APIs.

As you can see, hyper-automation, in principle, is not really different from the usual software automation: you take a manual task or business process and automate it with software. The productivity gains come from the removal of slow and expensive manual work. However, what's different is A) the tooling, and B) the scope of automation. Let's look closer at that:

Tooling

Regular software automation typically follows 2 paths. Let's call them "coding" and "scripting":

• "Coding" - is basically custom software development. Automation software is designed from scratch using a low-level programming language like Java or C#, and a programming framework (such as ASP.NET). The result is unique and rarely reusable. Due to the high cost of custom coding, it's only suitable for automation cases, where the benefit of automation is so huge that it's sufficient to compensate for the high cost of coding.
• "Scripting"—automation is done by using a high-level programming language of a software platform with some high-level abstractions of the platform already baked into the language. Typically, you can find it in enterprise ERP or CRM systems. Examples: ABAP scripting in SAP or Apex scripting in Salesforce. The cost and effort of automation in scripting are lower than with coding but still high.

Hyper-automation continues the trajectory of reducing the cost & effort of automation and uses a new path, "designing". Typically, "designing" means creating visual, graphical workflows that operate with even higher-level abstractions. Workflows represent a programming paradigm that requires absolutely no or very little coding, which makes them suitable for people with less technical skills. That has two important consequences:

1. "Designing" significantly expands the share of workers that can be involved in work automation.
2. "Designing" enables self-service automation, where people who know best the subject area also have the skills and means to automate it.

Scope

"Designing" simplifies automation by an order of magnitude and lowers the bar to the point where significantly more manual tasks can be automated. These tasks weren't automated because it didn't make sense previously—the cost and effort of automation were too high and wouldn't be compensated by the benefits of automation.

From a practical perspective, hyper-automation results in a massive increase in work productivity thanks to automation of even smaller tasks, such as using an approval workflow, sending a report on schedule, or setting up alerts and notifications. Organization-wide productivity gains come from thousands and thousands of small-scale routine operations automated by knowledge workers and business professionals in a self-service manner. While the benefit of automating one small task might appear negligible, the compound effect of automating as many tasks as possible (and reasonable) leads to higher work efficiency across the whole organization.

Of course, hyper-automation isn't limited to small tasks. Using elements of hyper-automation can significantly simplify and reduce the cost of automating large-scale business processes as well. For instance, certain internal web APIs can be built using visual workflows, instead of coding. However, since hyper-automation deals with higher abstractions, it means that it's less customizable (e.g., in terms of the user interface). From that perspective, hyper-automation shouldn't be viewed as a replacement for "coding" or "scripting" but rather as a complementing cost-reducing technology.

What's the difference between automation and hyper-automation?

I suppose, at this point, you, dear reader, can already answer this question. Here is my version:

The usual software automation, due to its high cost, is used to automate business processes where a lack of automation would cause significant (or even business-threatening) inefficiencies. So it's a relatively small number of critical high-effort work operations that must be automated.

Hyper-automation takes care of the remaining manual tasks for which automation using regular means would've been prohibitively expensive or cumbersome. Due to the large number of such still-not-automated work routines, the aggregated effect of their automation unlocks another level of work productivity in the organization.

Having said that, you probably shouldn't use a hyper-automation platform for purposes that can be automated with an existing business application (e.g., CRM) unless its functionality is clearly overkill. However, hyper-automation tools can be very effective in "gluing" together different business applications and creating automations that span multiple systems.

Random remarks

[1] A bigger share of workers involved in automation means broader automation governance is required.

[2] Machine learning (ML) can definitely be helpful in certain automation cases, but, generally speaking, ML is orthogonal to hyper-automation. There is no principal difference if ML is used for coding/scripting or hyper-automation. It's just one more data service with an API.

[3] Since enterprise applications used by an organization can be both cloud-based and on-premises, hyper-automation workflows should be able to cross the border between the cloud and on-premises. This may pose a challenge for purely cloud-based hyper-automation platforms.

[4] There are no-/low-code automation tools that don't use workflows. Instead, the required behavior is configured using all kinds of checkboxes and droplists. I'm not sure if such tools should be included in the "hyper-automation" category.