Back in 2020, Celonis’ Chief Evangelist, Lars Reinkemeyer, released a book called ‘Process Mining in Action’. The focus, as the title suggests, was looking at how some of the world’s leading companies were using process mining to better understand and improve how their organization runs. However, technology doesn’t stand still and since then Celonis has moved beyond the idea of using its platform to understand processes in isolation – and has moved towards what it is now calling ‘Process Intelligence’.
This was underpinned by the launch of the vendor’s Process Intelligence Graph, which allows companies to develop a knowledge graph that maps an organization’s entire ‘process knowledge’, and creates a ‘digital twin’ of how a business is running. This is made possible thanks to developments in Object Centric Process Mining, which allows organizations to map how different ‘objects’ (e.g. order, production, shipments, procurement) all interact with each other.
Given these developments, Reinkemeyer is publishing a new booked called ‘Process Intelligence in Action’, which is aims to explain how the Process Intelligence Graph is taking process mining to the next level, underpinned by an examination of 12 operational use cases from some of the world’s leading companies – including BMW, Bosch, IKEA, Merck Group, PepsiCo and Siemens.
diginomica has had a first look at the book and has also spoken to Reinkemeyer to understand more about how process mining technologies are advancing, as well as what can be learned from the customers that support the book’s key arguments. They key principle for Reinkemeyer is that this latest publication helps move customers beyond insight towards action:
With process mining we’ve been getting insights, understanding processes – mining, getting transparency, seeing where the issues are, and getting that X-Ray. Now, with process intelligence, it’s more about saying ‘it’s great to have insight, but it should be more about the action to value’.
What does an organization do with that insight? How can Celonis support our customers to drive impact action with action flows, intelligent flows, automation, and also realize value with the roach approach of an operating model with the value methodology. So the evolution for me is moving from the pure data insights towards advising customers on what to do with that and how to drive impact with that.
On the one side that is the methodology and the organizational approach. On the other side, it’s also the technology – we have the Process Intelligence Graph – which is the data backbone.
Celonis is working on a number of fronts to expand how it delivers process automation to its customers (that’s the action part, beyond process mining insights). For instance, last week it announced the launch of its Orchestration Engine with Emporix, which goes beyond simple robotic process automation (RPA) of tasks, such as calculating a price or creating a quote, and allows buyers to fully automate the entire sales or purchasing processes across an enterprise.
Reinkemeyer’s view is that across all the customers he has spoken to, the common theme for organizations seeing success is that they understand the purpose of what they’re pursuing:
It’s not about the bits and bytes, it’s not just about the insights or the data, but it’s about the purpose. Talk to somebody in charge of a process and say, what’s your priority? Do you want to drive automation? On time delivery? Reduce working capital? What exactly is top priority? If I provide you with the right capability and insight, are you going to use this and ask your organization to work with that to drive action and value?
And understanding the desired outcome is key:
In respect of purpose, what do you want to measure? Is it dollars? Is it productivity? Is it employee satisfaction? Is it carbon emissions? What exactly is on your agenda? Then working towards making sure that the person in charge can use the capability to have measurable, tangible impact and improvement.
It changes the discussion from a cost center to a profit enabler.
People are key
From speaking with customers for the book, Reinkemeyer also was able to identify common pitfalls the adoption of process intelligence in the enterprise. Firstly, he said that not having a sharply defined purpose, as explained above, can result in organizations getting distracted – leading to poorer outcomes. But equally, it’s important to utilize the right people:
Not everybody in the organization is keen to change. You have to find those people in the organization that are keen on data, who are keen on transforming, and who are keen on making things different. Not many people are sincerely behind driving change.
You need a good skill set in the sense of that change management understanding there. Seeing who in the organization are the people who are open to change. How do I build this cascading structure across the organization with the right people, with the right skill set? Every organization has people who are reluctant to change and people who recognize that something can help them.
An approach that has long been advocated by Celonis to ensure getting the best out of people and driving change is the adoption of a Center of Excellence. The vendor has found that establishing an effective Center of Excellence, with strong executive sponsorship, can result in greater transparency, can lead to more impactful use cases being prioritized, and can deliver higher cost savings.
Future ambitions
In addition to this, Reinkemeyer believes that the introduction of generative AI into the Celonis Process Intelligence Graph will further drive adoption and change. But key to this is recognizing that the Process Intelligence Graph can act as a new engagement layer for the enterprise:
One of our core approaches is talking about this common user layer, where Celonis acts as a user interface on top of SAP, Oracle, Salesforce, etc. In respect of IT architecture, take a company like Siemens, it has established one common data lake in Snowflake. That is its one common data lake, where they have all the event logs and process information in there, independent of whether an order has been in Oracle or SAP or Salesforce.
Siemens can visualize end-to-end via that common platform, which is available in Snowflake. Everybody is dreaming of AI, but if in two or three years a common user asks the Celonis co-pilot ‘tell me my on time delivery’, the LLM needs to have all the data – that’s why you need to have one common database, where you can connect that information to provide a meaningful answer. That’s where we very much believe in the Process Intelligence Graph as one single data source.
Companies that move towards this common data layer, however, will benefit from the use of LLMs within Celonis in the future, Reinkemeyer believes:
I expect that in the next couple of years that with a co-pilot it will be even easier. That everybody in the organization will be able to use natural language to just type in ‘tell me my working capital improvement’ – this will make it even more attractive and easier.
That’s why we are so keen to launch this co-pilot – people without building an exclusive skill set can use it. In my view, I think software should be interactive, without needing to train people to work with the software.
In terms of where process intelligence in the enterprise is heading, Reinkemeyer suggests that generative AI will allow organizations to become more proactive, with a co-pilot predicting the needs of a user:
Let’s just imagine there is a smart co-pilot that actively alerts somebody from procurement saying ‘these are your top 10 priorities’. A co-pilot that is capable of learning which blocks can be auto-removed in order to cash. A co-pilot that is able to understand a user that’s interested in inventories, one that asks: what can i propose to this user to reuse working capital? I think the way it’s shifting is that we are going to see those smart copilots.
My take
The thing that makes Reinkemeyer’s new book compelling is that it’s underpinned by a dozen case studies from some of the world’s leading enterprises. Whilst highlighting the successes from these organizations, it doesn’t shy away from the learnings about how things could be done better. As companies shift their focus from process mining insights from individual processes towards actioning process intelligence across multiple processes, this will require organizations to think more broadly about how to motivate people to think beyond silos. That culture change is hard, but there’s plenty of opportunity for those that understand how to really implement change.
