RPA—Robotic Process Automation—is a great concept, but getting the most out of it needs a human touch. The benefit of software robots (and robots in general)—being able to execute tasks repeatedly and reliably—first needs critical analysis to discover which processes are the right ones to automate.
Within “shared service organizations”, where multiple BUs take advantage of resources centralized elsewhere, this question isn’t easy. Because a world-girdling corporation may have hundreds of ways of doing the same thing, all following slightly different processes, without some smart thinking, automating it would mean a separate software robot for each process variant—hardly a model of efficiency.
Let’s look at how the Mavim magic—process mining and BPM, Business Process Management—can make RPA work harder for you.
RPA has Great Potential for Shared Service Centers...
Shared service organizations are a hot topic in business process automation. But the idea behind them isn’t new. It was in 1937 that economist Ronald Coase explained why corporations are, well, corporations and not random mobs, while managers had been putting his principles into practice for centuries before that. It’s all about efficiencies.
Whether it’s called a Competency Centre, Central Buying Unit, or Purchasing & Procurement, any business resource—from raw materials to human capital—can benefit from a consolidated approach that drives down unit costs across the corporation. If Alice, Bob, and Becky each need 100 widgets, and the supplier’s charging $100 for each separate order, combining them into a single PO for 300 might attract a discount. And if the department handling that consolidation has a broader mission to actively seek out similar opportunities across the business, it means reduced costs, standardized methods, greater bang for every buck—enterprise-wide.
The way RPA works—intervening in recognized processes, extracting data for further processing, often using rules-based AI as it roams and skips across your databases—is perfect for straightforward, definable paths. But what happens if the paths are less defined?
… But Has Limitations When Standalone
And there’s the problem. If you look at an operational model of the average large business—all its interacting processes and procedures, happening in real time—the first thing you notice is complexity. Extreme complexity. Hundreds of departments, thousands of processes, often hundreds of different ways of getting the same basic task done.
And RPAs, however “intelligent”, remain robots. They can’t make intuitive leaps of logic and apply them to other processes on the map. They can bring great value to a clearly-laid-out procedure, but won’t achieve much if let loose on the business as a whole.
And of course the resources we share today have less obvious processes than bulk-bought paperclips. SSCs are consolidating cloud services, enterprise applications, intellectual property, brand values. Where a lot of processes evolve by seat-of-the-pants thinking, different in every office.
Automate everything and you automate nothing
The worst outcome of RPA for a multinational is a profusion of software robots, each doing a single task well yet each needing the usual human attention from IT management. (Business processes change and evolve, after all.)
One RPA implementation may be saving a million dollars a year for one office, but if it takes a team of ten to maintain it the savings are gobbled by unexpected costs. (As any Operational Excellence Manager will tell you with a rueful smile.) A million dollars here, a million dollars there, and soon you’re talking real money.
That’s where Mavim comes in. Because the big wins come when you empower RPA even further: not just consolidating multiple instances of the same process, but seeking out processes that should be the same but aren’t in order to bring them into the fold.
How BPM Can Help Scale the Benefits of RPA
Think about it. An MNC may be using an RPA to progress the purchasing cycle. But if every office takes a slightly different approach to raising a PO, the efficiency gains are limited.
What’s needed is to find the “ideal” process—the approach that’d work for everyone—then consolidating all its variants into that optimized one, worldwide. (The economies aren’t marginal: one Mavim client found its simple PO cycle had 127 variants around its offices.) So our software robot friend can work everywhere.
(A note to anyone who thinks this sounds like the Rise of the Machines: it’s rare that RPA takes anyone’s job. The usual outcome is that people become more efficient, with humdrum tasks automated off their desk.)
Mavim calls this discovery phase “process mining”: digging down into your operational model to map out those variants and work out which ones are the role models, relying not on gut feel but hard metrics like cost and time. Presenting its findings in well-understood tools like Visio, it’s not just for technical types, either. In fact, people on the ground often get enthused by seeing how much they could improve their lives by simplifying and consolidating the way they work.
After process mining comes process management: actively putting those improved processes in place across the organization, so an SSC can put RPA to work everywhere. But there’s more than one way to do process management—and more than one company offering it. So why choose Mavim?
Aside from its easy-to-use front-end, the Mavim difference is that it maps out the processes in your operational model in real time, connecting to actual sources of data and building each process model based on their interactions. What happens when a PO comes in? Who’s involved in the process, how many layers of approval are there, how much time elapses before it’s acted on? Process mining builds that big picture, as a digital twin of your organization—and shows you how many variants exist and the metrics for each.
This digital twin is important, because it lets you run What-Ifs and see how each change affects the big picture before rolling it out in real life. After all, every process is connected to other processes; what looks like a good tweak in Department A may cause trouble in Department B. An optimal process plays well with all the processes it touches; a digital twin is your playground for finding it.
That’s how Mavim makes Robotic Process Automation shine at scale. It’s not a single idea, but a love triangle: PM, BPM, and RPA.
Knowing your operational model in detail lets you plan your target operational model: the idealized state where everything’s optimized and scale economies are at their peak. Nobody ever reaches that state of perfection—but with Mavim, you can get close.
To see what Process Mining and Business Process Management—collectively known as the Mavim Intelligent Transformation Platform—can bring to your RPA plans, get in touch!