Follow the money! That’s a foundational mantra for journalism. Want to know what the story is? Pay attention to who got paid, and for what.
Similarly, in the business world, revenue streams drive everything. Understanding where the money comes from and why? That’s the first step in thriving. But it’s only the first step.
The next series of steps, which typically span multiple production systems, were described by Chinmay Trivedi as the “value chain.” Trivedi is the Treasurer for GE Healthcare. He recently spoke at Celosphere 2021, where he was interviewed by Celonis co-CEO and co-founder Alexander Rinke.
“You have a complex set of systems and processes; it’s not a super simple business. What you do is very unique,” noted Rinke. He then shared a whopping data point with the virtual audience of thousands: GE Healthcare achieved “$1.3 Billion of increased free cash flow,” with Celonis.
For a company with as many moving parts, divisions, partners and customers as GE Healthcare, freeing up $1.3 billion of cash flow is an absolutely stunning accomplishment. To put that number in perspective, you could pay $80,000 to 16,250 people with that much money.
“We’ve gone through a lot of environment challenges, and we’ve come out pretty well,” Trivedisaid. He noted that they focus on “continuous improvement. It’s a collective team effort to see how we can get better every day. We had a good year last year, and it’s not a one-off.”
Those words are music to the ears of process enthusiasts everywhere! And this period in the evolution of enterprise computing could easily be described as the golden age of process automation. That’s partially due to the maturing of process mining and task mining.
Continued Trivedi, “If you understand the value stream, when you map out the entire order-to-cash process, the distribution of partners, to get that order in front of the customers; installing that order; and then doing all of the work… you realize you are connecting with many different parts. There are multiple transaction streams that touch this process.
“So, how do you connect the data across those transaction systems, and have a view across this transaction stream? Action won’t be taken by one entity. That becomes a challenge for a global company like ours.”
Trivedi said they took basic principles from lean manufacturing. “It’s all about finding waste and eliminating waste,” he said. “One of the biggest wastes is waiting: order to manufacturing, installation, billing to collections. You can almost track where your value train stops.”
And that’s exactly where the combination of process mining, task mining and action flows comes into play. The first step leverages data primarily from log files of production systems, such as SAP’s Enterprise Resource Planning (ERP) technologies.
The second step fills a critical gap: task mining captures the desktop activity of users who touch some part of the process. This helps complete the picture of the end-to-end business process. Rinke noted that Celonis currently has connectors to more than 100 different systems.
Once all the data is ingested, the process manager can see in a visual flow exactly where the bottlenecks occur, or, as Trivedi noted, where the value chain stops. And that’s historically where most process management tools left off.
If you wanted to improve the process, you would need to figure out some way of doing that within the production systems themselves. Sometimes, the tools enable such change management. More often, custom coding has been used to amend enterprise workflows.
But as far too many an executive has learned, custom coding only works for so long. It tends to be a brittle fix, that later requires significant maintenance. What’s more, if there is no clear documentation (which is common), then troubleshooting or improving gets that much harder.
To paint the picture clearly, consider this old Russian proverb, which conjures up imagery of cold, winter battles against the Germans or the French, depending upon which century you’re considering: “There is nothing more permanent than a temporary solution.”
The Last Mile
What makes Celonis special is their commitment to enabling real change. They’ve even coined a new term to describe what their technology really does: an Execution Management System or EMS. As the name suggests, this system literally manages the execution of business logic.
What Celonis has effectively done is to create a center of gravity around the critical change agent of business: process management, including the “do” part, not just the monitoring and reporting component; but actually getting things done.
They accomplish this with their writeback capabilities. When process managers figure out how to solve a particular problem, they use the EMS to optimize the workflow with specific logic that gets published into the production system itself, thus achieving digital transformation.
Granted, there are other technologies which for many years now have served the purpose of automating business processes. Automation is frankly the core building block of every piece of software ever written: applications automate; that’s what they do!
What Celonis did that was so compelling is to create this vast marshaling area for both functionality and process optimization. In doing so, they’re creating a workbench for process experts (read: digital transformers) to ply their trade, and incrementally optimize the business.
Machines That Learn
Because Celonis acts as both design studio and execution engine, the platform can not only ingest tremendous amounts of data from all manner of business processes (invoicing, procurement, payment, supply chain, human resources); it can also manage those processes.
Add in the power of machine learning, and you’ve got a very interesting development. When you consider the scale of scenarios that Celonis has already captured (Rinke noted they have 1,000 deployments), you quickly realize the richness of the process tapestry that’s already in motion.
Take something as simple, and painful, as duplicate invoices. Most ERP systems only check for duplicates when invoices are posted, and only catch them when they are perfect. What you really want is a sort of fuzzy matching, that enables the system to catch similar documents.
The machine learning component of the Celonis EMS uses the aggregated data of enterprise processes to optimize the predictability of fixes. For example, the system can catch many duplicates that are not exactly the same, and even fix the mistakes automatically.
At first blush, a fix like this might seem small. Upon investigation, most companies will realize that the extended workflow around a duplicate can triple or even quadruple the amount of time spent on a rather rudimentary part of business. This not only wastes time, it lowers morale.
The net result of using a platform like Celonis is what one presenter called: “cross-system, real-time intelligence.” That’s an excellent moniker to describe the fuel that will drive the next generation of cloud-native, scale-out enterprise systems.
Rinke offered, “Data is not something to look at, and stare at, and admire; it has to drive operational actions in the business.” That’s exactly what the EMS facilitates. And the canvas upon which these artisans can work is as broad and wide as the business world itself.
“We will blow your mind with the speed and quality of our EMS,” exclaimed Rinke. That will certainly be a welcome experience throughout the C-Suite. But the happiest people? Probably those Chief Financial Officers. They’ll be smiling all the way to the bank.