How to Pick the Right Automation Project
Intelligent automation is the fastest growing area of enterprise tech investment. The potential to boost performance in the typical company with these tools is both broad and deep. But as companies look for places to apply these tools, they often fall into two common traps: chasing quick, easy wins that won’t have much impact, or big, ambitious projects that will offer major strategic advantages. What they should focus on, however, is a third option: projects that build their capacity.
Whenever a new wave of technology splashes onto the scene, managers face the same questions: Where do we start applying it first? Do we go after the “low-hanging fruit” that will produce quick wins and build the case for more ambitious projects? Or should we strategically focus, with no delay, on the applications that will give us a decisive edge over competitors?
Right now, with the arrival of a revolutionary set of technologies for automating knowledge work — artificial intelligence in particular — we see teams grappling with these questions at high levels in organizations. Intelligent automation (the term commonly used for robotic process automation, machine learning, and artificial intelligence in organizations) brings unprecedented speed, accuracy, and pattern-recognition power to business processes that routinely call for deciphering information, from fielding customers’ questions to complying with government regulations to detecting fraud and cyberattacks. Because that describes so much of the activity of modern workplaces, the deliberations about where to start and how to proceed are different than with other technologies. The same old answers don’t apply.
The potential to boost performance in the typical company with these tools is both broad and deep. In one company we know, a team was assembled to survey all of its operations, find areas where people’s time was being consumed by repetitive information-processing work, and come back with candidate tasks for automation. The list stretched to hundreds of things a smart machine could do to leverage workers’ creativity, increase speed to decision, improve accuracy, or enhance service to customers.
There are also strong competitive incentives: Because of this potential, companies are investing in these tools at blistering rates — according to Gartner, intelligent automation is the fastest growing area of enterprise tech investment. The pandemic gave the toolkit a giant shove forward as companies suddenly had to find new ways to perform mission-critical processes.
Whether driven by the opportunities or competitive pressure, your organization will likely soon be using intelligent automation in many, many corners of your operations. So, where should you start?
Instead of framing your goals in terms of quick victories (which won’t really move the needle) or major strategic applications (which require skills and foundations you don’t yet have in place), focus on how your first steps will advance capability-building in your organization. You should sequence the projects you take on — knowing you will ultimately take on hundreds — so that the early ones build the AI talents and put in place the AI tech infrastructure for the projects you will take on next, and next, and next.
Map Where You Want to Go
Capability-building — developing the strength of an organization to solve a class of problems it will keep facing in the future — is a challenge you might have tackled in other realms. In areas from strategy formulation to project management, teams recognize that they can and must get better by learning from experience. And because there are fundamentals that must be mastered before they can advance to higher-order capabilities — they have to walk before they can run — teams often take their guidance from so-called maturity models, outlined by experts who have watched others travel the same path before. Given that your people will need to rise again and again to the challenge of implementing intelligent automation solutions, this is the approach that makes sense, but more of the thinking about the best sequence of steps will be up to you.
Planning this journey requires mapping out how your team or organization will deliberately move from a state of being a novice to being an expert.
The first step is usually an assessment of existing capabilities: the challenges your people already know how to tackle and the sophistication of the tools they have to solve them. Perhaps you already have strong data analytics skills on staff, for example, or people who have been involved in RPA installations elsewhere.
Your next step is a gap analysis. This details the difference between your current capabilities and the demands of the most challenging solution you can envision taking on. This might reveal that your current IT infrastructure is simply not equal to a coming wave of applications that will need to interact with disparate data sources. Or that much more effective collaboration will be needed between software developers and business process owners than has been seen in the past.
Finally, with the beginning and end states clearly articulated, you can then specify a step-by-step journey, with projects sequenced according to which ones can do the most in early days to lay essential foundations for later initiatives.
Here’s an example to illustrate how this approach can lead to better choices. At a construction equipment manufacturer, there are three tempting areas to automate. One is the solution a vendor is offering: a chatbot tool that can be fairly simply implemented in the internal IT help desk with immediate impact on wait times and headcount. A second possibility is in finance, where sales forecasting could be enhanced by predictive modeling boosted by AI pattern recognition. The third idea is a big one: if the company could use intelligent automation to create a “connected equipment” environment on customer job sites, its business model could shift to new revenue streams from digital services such as monitoring and controlling machinery remotely.
If you’re going for a relatively easy implementation and fast ROI, the first option is a no-brainer. If instead you’re looking for big publicity for your organization’s bold new vision, the third one’s the ticket. You can set up a tiger team or separate organization and give it full license to disrupt the existing business. But note that neither of those approaches really prepares the ground for intelligent automation to spread to other applications by the existing organization; they don’t make the people of your organization generally more interested, receptive, or able to apply intelligent technology elsewhere. In other words, as an organization, taking these routes doesn’t take you far up the learning curve, toward greater maturity with the technology.
This is what option two would do — in large part because it would demand that the company get its act together on data. Without a good enterprise data strategy, people in different parts of the organization lack common standards regarding what data needs to be gathered and how it should be organized, cleaned, and prepped for analysis. This is a foundational capability that the company will need to have in place to make headway in using machine learning at scale. From the standpoint of capability building, it is easy to see how progress on enterprise data would unlock, say, 10 other projects — which in turn can be prioritized by the further capabilities they could add. Our manufacturing company could lay out a roadmap showing how, five years later, it will not only be reaping the returns of the specific projects, but also be generally and profoundly more ready to take on truly transformative initiatives.
Fifty years ago, when the legendary Peter Drucker coined the term “knowledge workers,” he also recognized how their rise in the global economy would challenge organizations. “The most important contribution management needs to make in the twenty-first century,” he wrote, is “to increase the productivity of knowledge work.” Finally, in intelligent automation, a powerful toolkit exists for doing that — and the race is on. Avoid the mad dash that has your organization chasing possibilities but with no collective progress. Choose your spots wisely, and your investment in intelligent automation can be a capability-building journey.