This E-3 Special is available for download at the bottom of the page.
To understand and optimize business processes, a company’s data wealth needs to be raised from the depths. To accomplish this, Celonis Process Mining uses the digital traces that almost every transaction leaves behind in the company’s global IT systems – and understanding the actual business processes creates clarity, transparency and agility, resolving conflicts and bottlenecks in the organizational structures and workflows.
What data mining is for the data analyst and BI expert, Process Mining is for company management.
Process Mining analyzes and visualizes business processes, improving SAP customers’efficiency and agility. Unlike the purely quantitative improvement that comes with a faster Hana system, Celonis focuses on quality improvement.
A flawed system that merely runs faster after a Hana and S/4 upgrade is still flawed, but Celonis offers the SAP customers a wholly different approach. The software analyzes the processes to create a visual map, making the optimization potential more obvious, ensuring transparency, harmonization, efficiency and agility.
Initially, Process Mining sounds like magic; how can you visualize something and find weak points in a totally dark environment? Every ERP system is a black box, isn’t it?
„That may have been true for the time before Process Mining,” Bastian Nominacher points out in discussion with the E3 Chief Editor.
Specifically, he notes, „Gaining access to systems and data was often difficult and time-consuming. Long waiting times and laborious access paths were frustrating to those working on them. That also made fast and flexible decisions impossible, the kind that agile companies must reach today. Our technology has made this data treasure accessible. And it has done so in a simple way that everyone can understand.”
Securing The Digital Traces
SAP customers now have the chance to visualize processes exactly as they are. „It’s about gathering up the digital traces that each digital transaction leaves behind, and transferring them into a picture, in other words visualizing them,” Bastian Nominacher explains.
The weak-points in companys’processes emerge from precisely this analysis. „At a glance, you see what is really happening and from there you can go straight into the analysis of root-causes,” he explains, based on many very successful implementations of Celonis software in different companies.
Classic error-analysis methods reach their limitations quickly. Obviously, they can examine whether a process is running optimally, but they often fail to provide insights on why these bottlenecks emerge, why deviations occur, and where starting points for optimizations can be found.
„By contrast, Process Mining by Celonis faces no such limits. Our technology puts us in a position to use, assess, and analyze enormous datasets, involving millions of process instances,” Nominacher explains, pinpointing the difference between Celonis and traditional analysis tools.
„One of the really wonderful aspects is the ’aha’effect, the sense of revelation we see so often. The insights gained from our process analyses are often very surprising for customers,” Celonis’s-Co-CEO notes, explaining what can be new for an SAP customer, „Many people presume that the snag lies in a single process or a certain business unit. Yet for a long time, backing this up with proof was often very laborious. The root causes of process deviations or delays can also in fact be found somewhere quite different to what users initially think.”
With Celonis you can see at a glance where the current process deviates from the requirement, and what effect that deviation will have on the whole company’s efficiency.
For instance, if orders and invoices are given manual follow-up processing, or orders are submitted too early or too late, you can kiss efficiency goodbye.
Without Deep Interventions
Questions emerge that are probably on the minds of everyone responsible for SAP infrastructures: which preliminary settings are necessary and how deeply does the Celonis software reach into the ERP system? Which authorizations are necessary?
Bastian Nominacher puts the fears to rest. „There’s no need to reach deep into the ERP system.” This is because Celonis uses many standard connectors for a company’s core processes, so the linkup to SAP works quickly and smoothly.
„If you process data directly at the productive Suite-on-Hana instance or S/4 instance, you also save the replication, additional hardware etc., and can apply Process Mining directly in real-time in all SAP processes,” Nominacher points out.
In principle, the approach of positioning specialized connectors and software agents at specific places in the system is not new. Agents are an established reality both for system monitoring and particularly for the network supervision of wide area networks and local area networks.
Up to now, however, nobody had the idea of applying this technique to business processes, or to operational and organizational structures. Celonis is an absolute pioneer in this sector, so it took some time and hundreds of e-mails even to convince SAP of this.
There’s another important question for the CIO and Head of CCC: if Celonis software embarks on the search for weak points, does the productive system’s performance suffer as a result?
Here too, Bastian Nominacher allays fears: „Celonis Process Mining runs in the background. It is optimized so as to have no effect on the systems’productive operation.“ And perhaps the CEO and CFO ask themselves: can certain processes and functions in an ERP system „hide away“ from the Celonis software?
Does it have any ’blind spots’? „There are only blind spots when there is a media break, i.e. if the process is continued manually,”
Nominacher observes. As soon as a process operates digitally, it leaves behind traces. Process Mining only needs three components in the event log: a datestamp, a characterization of the event (e.g. ’pay invoice’) and a key to that operation, such as the invoice number.
„Our process-mining technology works independently of the process logic’s actual implementation. The only important thing is which traces emerge, i.e. database entries. The process-mining algorithms can reconstruct a process flow from these traces. If a customer has set up tables in the Z-name area, these first need to be registered, i.e. made known to the system as an event source – and can then be further processed without a problem.“
Abap’s And Java’s Weak Points
Noteworthy, of course, is the question: can Celonis find weak points in Hana database functions that are programmed to completion in Abap or Java? And in NetWeaver Java-Stack?
„Celonis identifies the process variants based on the digital traces, i.e. entries in the tables, at database level. This procedure’s great advantage is that it is entirely independent of the stack positioned beneath it. Whether it is Abap stack or Java stack, the process-mining technology can be applied unchanged, so as to obtain process transparency.” This is how the Celonis co-founder describes the system’s fundamental functioning.
Does this mean that for Celonis, an SAP system is like a ’transparent patient’? „Patient is perhaps not the right word,” as Bastian Nominacher puts it, „but yes, we can extract all data relevant for the process flow from SAP.”
Neither is there a difference between on-premise and cloud systems. „No, for us the sole important thing is to have access to the system. Where the data is located does not matter to us. In fact, it is often possible to set up data access more quickly from cloud systems, because the providers already offer corresponding interfaces,” Nominacher points out.
„Of course, Celonis contains a vast amount of human brainpower. A big team of developers and architects has put its very best into it and is still doing so.
The latest insights gained from artificial intelligence were also used,” Celonis’co-CEO observes. Celonis Pi (Proactive Insights), the latest generation of Process Mining, aims to set standards in the combination of artificial intelligence (AI), machine learning and Process Mining.
„In an extremely short time-span, Celonis Pi correlates millions of process-related data units – doing so with the most modern machine-learning methods,” Nominacher says. „The result is thought through from the user’s perspective: The fault lines in a process are identified automatically. This saves the user a lot of analysis work. Better still: Celonis Pi also learns in this process and suggests what measures could be successful.“
Up to now, when an SAP system was inefficient and error-prone, the root cause analysis was carried out by a user’s own IT team, with the Head of CCC or an SAP partner’s external consultant.
Does Celonis software render this competence obsolete? „Not at all,” Bastian Nominacher replies, „using Celonis Process Mining, weaknesses and inefficiencies in the processes are brought to light. The insights mean that teams and IT admin units can focus on their core competences and on optimizing the processes analyzed. So the software helps users to work more efficiently, resulting in huge time-savings. In doing this, Celonis primarily assists companies in working in a more agile way and in improving customer satisfaction.“
On this basis, who does Celonis see as its point of contact in a company? Who takes up this search for the root cause? The SAP base, the Head of CCC, the CIO, the
„That fully depends on the company,” Nominacher remarks. „Our solution is open to users on all levels, and that’s exactly what makes our work so exciting! As we see it, no purpose is served by offering analytics solutions for a handful of specialists. Everyone must be enabled to reflect on their individual processes and to optimize them. At Celonis what we are about is an intuitive solution that cuts across departmental borders and can be used for all processes.
“You don’t need special IT knowledge for the application – but a bit of process knowledge certainly helps.“
And now the result: what is the outcome when an SAP customer researches root causes? Is it the adaptation of the customizing? The optimization of the Abap code and/or the Z-functions? Or is it the adjustment of the operational and organizational structure?
„This is a tough question,” Celonis’s co-inventor tells the E-3 Chief Editor. „It totally depends on what problem the customer has diagnosed specifically and what the customer’s starting position is. All the measures that you have just named could be the outcome of a process-mining analysis.“
Practice often shows that it is a combination of technical and organizational measures. Most of all, what is really helpful is that you can immediately recognize the measures’success straightaway, thanks to the live link-up of the data.
„If a particular measure does not bring the desired effect, you can steer away from it. If the approach succeeds, it is worth broadly rolling it out within the organization,” Nominacher points out, based on everyday practice with customers.
„Moving away from a Big Bang optimization towards fact-driven, continuous process optimization, is what makes our customers so successful and is the basis for sustained process improvements.”
It is always left to those directly involved to decide which measures they take; however, recognizing weak spots can and should bring about changes. These changes can take place in all areas, in technology or in the organization.
„We merely offer companies the starting precondition,” is how Nominacher defines it, adding: „This means comprehensive insight into processes, indications of what the root causes are, and even recommendations for action. Here’s another advantage: with Celonis you see, almost in real time, whether a measure is taking effect and, if need be, you can adapt the direction being taken.“
The added value from these efforts takes very different forms for various SAP customers. „The great thing about our technology is that, in all areas, processes are made transparent and are visualized.
Again and again, this means that the benefit lies elsewhere. For instance, in Purchasing departments, the matter at hand is organizational advantages; in Production there is a greater focus on technology. Our software offers the customers real-time insights into processes and full transparency for their process flows.
Almost always, this leads to extremely fast ROI.“ Celonis customers confirm that, in many cases, the investment proves to be worthwhile within half a year.
For many SAP customers, SolMan has established itself as central platform for documenting business processes, especially in the most recent version, 7.2.
To what degree can the results of root cause analyses be integrated into SolMan’s ERP documentation and BPM?
„The process samples identified by Celonis Process Mining can be exported as BPMN models and thus directly be taken over into a process documentation. This can be either done to start the documentation from scratch or to add to and update existing files,” Bastian Nominacher points out.
Currently there is not yet an automatic, bidirectional interface available between Celonis and SolMan. However, extracted process models can be exported from Celonis as a BPMN model and then be processed directly. Via Pi Conformance it is possible to load existing Business Blueprints into Celonis and identify deviations between as-is and to-be processes. „Machine-learning algorithms help to determine root causes and possible approaches to solutions for deviations like this,” Nominacher adds.
The concluding questions are these: where does Celonis still see a need to act in the future? What is coming up next? What significance does machine learning/ deep learning have for Celonis? What can the SAP customer still expect from Celonis in the future?
„We are working continually on our solution and further developing it. My firm conviction is that the process data still holds very many more elements of hidden potential. Celonis Pi was a quantum leap for us – above all, because it uses the latest algorithms from AI and machine learning. We are very proud of what we have achieved. But of course we are still constantly developing – SAP customers can still expect some innovations!” Bastian Nominacher remarks, concluding the E-3 discussion.