Forget AI. Work automation is the game-changing tech you can use right now.

Martin Stewart - Jan 30, 2023

According to the 2022 State of the CIO survey, 46% of CIOs are pushing to increase operational efficiency (up from 34% in 2021). The pressure is on for IT leaders to do more with what they’ve got. 41% of CIOs said they aim to transform existing business processes through automation and integration—indicating that automation is a key strategy for achieving higher efficiency. A report from KRC Research found 78% of business leaders and 53% of employees believe they could save 2 to 3 hours of mundane work every day with work automation.

Demand for simple automation tools outstrips AI

41% of organizations are already using some type of work automation tech (56% plan to do so in 3-5 years). But only 13% are using AI-driven tools now (with 44% planning to do so in the same timeframe). That means that in the next half-decade, 57% of organizations will be using AI-driven automation and 97% will be using non-AI automation.

So why is the appetite for process-driven automation overshadowing AI-driven tech?

1. Process-driven workflow automation is easy

With simple, process-driven work automation what you’re doing is automating the work people are already doing. They know what needs to be done—and how it should be done. They know it works. With workflow automation, there’s no training or upskilling required. No lab coats or boffins required. Workflow automation is within reach of the people who do the work.

But AI-driven automation requires specialist tools and training, throwing up a barrier to adoption. Business and IT leaders aiming to boost efficiency of work organization-wide can see that automation tools that everyone can use gain more traction than complex AI tools that require IT assistance.

2. Routine work is repetitive and doesn't need AI to drive it

Business leaders and employees want to automate mundane work for two reasons:

High volume repetitive tasks are typically low-value to the organization, yet take up significant blocks of employee time (and energy)—with little variation. Simple logic is sufficient to handle the variables in the process.
Repetitive tasks are boring. Employees would rather be working on challenging, high-value work that makes a difference.

By nature, routine work is repetitive. Repetitive work is bad for people, but great for machines. By nature, most routine work is also relatively simple. That means it’s easy for a person to model work as a process for a workflow engine to execute.

3. The business need is for predictable outcomes, not constant optimization

Work automation is predictable and trackable because people create the workflows and the tech follows the logic to execute actions. Work automations happen in the same way every time. Visual workflows make them easy to create and easy to understand. Workflows created by people are deterministic. That means there’s no risk of unintended variation.

Conversely, AIs can introduce unexpected variations—variations you don’t know about, don’t understand, or don’t want (and are often discovered after the fact). An AI’s process is non-deterministic—meaning you don’t always know what the AI is doing. That could mean that risks are being introduced. And it can take a long time to discover and train all of the potential risks out of an AI system. Over time, an AI can optimize a process to make it better, faster, and cheaper. But in most cases, using an AI to automate high-volume, low-value routine tasks is like using a sledgehammer to crack a nut.

Takeaways

Workflow-driven automations are easier and faster to implement because the people who do the work can automate the work—whichever team they’re in. By contrast, AI-driven tech requires specialist skills and tools, so they naturally fall within the IT department’s remit. This brings three problems: more work for the IT team, slower progress for business units, and high risk of incorrect implementation of the process.

That’s why—for most organizations—simple, drag-and-drop workflow automations are a better fit: a more adoptable, more sustainable initial solution to the problem of efficiency.