November 27, 2017
Automation done right: 10 critical steps to forge path to success (Part 1)
By: David Steuer

By now, anyone who has started down the path of robotic process automation or RPA knows it’s not as simple as just plugging in a bunch of bots and instantly receiving business results.

In fact, for some who have gotten automation wrong—and there are many ways to do so—the results have been suboptimal, wasting precious time and money and, in some cases, even damaging the brand.

At Accenture, we put automation to work by using automation to free up more than 40 million productive hours over the last two years and help these employees eliminate repetitive tasks and move into higher-value roles as we continued to grow. We learned a lot from this effort and are using the assets and lessons to help other organizations achieve their objectives.

But what exactly does a business do early in its automation journey to forge a path to success?

We’ve identified 10 themes to get automation right; the first four, detailed below, revolve around how you plan, start and govern automation. In my subsequent posts, I will focus on technology and execution.

1. Pick the right process to begin the journey

I’ve seen two big mistakes in this area. The first is picking a process that isn’t singular; it’s actually multiple processes. For example, one client wanted to implement automation for onboarding new customers. That’s far too big and complicated place to begin. We worked with them to break the larger process down into bite-size activities or tasks, addressable by automation, while still keeping a focus on the end-to-end process.

Secondly, find something that provides visible value without being too complex. In the onboarding example, perhaps automate credit checks of individuals, or verifying whether a person is an existing customer. Seek something large-volume yet relatively easy and quick to start. 

2. Don’t apply automation as a cure for a “sick” process

Automation is not a band-aid—it can’t fix a process that’s already broken. Robotic process automation, a rapidly growing category of automation, should almost be considered as a last resort. The first questions to ask are, can I do something different to change the process? Instead of going “downstream” and applying RPA, can I go “upstream” and address the source of the problem? (Check one of my previous blog posts for more on RPA)

For example, a health insurance company that was automating its claims handling processes focused first on identifying the key data it needed for straight-through processing and changed its upstream intake processes to obtain this data at the source.

We often use Design Thinking techniques to help our customers re-imagine their processes and customer journeys—focusing on the end-user and customer experience, then applying automation as an enabler for a loveable experience.

3. Build the right business case

Automation is about more than just removing costs. It also can help reduce errors, and improve security, compliance, customer satisfaction and productivity.

For example, we’ve been working with an automotive original equipment manufacturer (OEM) to use automation to help improve their relationships with their dealers and reduce dealer administration so the dealers can spend more time servicing customer needs. The number of people employed by the dealer does not change much, but the productivity and time spent on value-added activities increases significantly. Look at the entire back-office side of sales: The more you can automate administrative work, the more you can invest in people and assets to better sell, and the more time your people can devote to selling and servicing customers.

   4. Make sure you’re business-led and technology-enabled

We are often asked, who should own automation within a company? The business side, or IT?

The business side, our experience shows, must own the outcome. Senior leaders—such as the vice presidents or directors of finance, procurement, data management, and so on—should create and own the business case, identify potential automation opportunities and design the automation functionality and process changes needed for success.

Yet, business leaders should not make the mistake of moving forward with an automation implementation without bringing IT on board. That’s a recipe for trouble: There may be critical missing elements, such as proper cybersecurity or bot failure recovery.

IT should own the technology strategy and the technical implementation (including security, audit, environments and exception handling), and also partner with the business on opportunity identification. Ultimately, automation success requires a collaborative “handshake” between business and IT.

In the next installment of this blog series, I’ll address additional practices that delve more into the technology nuts-and-bolts of implementing automation, operational considerations and the role of artificial intelligence to yield and scale benefits. Stay tuned.



Popular Tags