Apr 18, 2020
Every action follows a certain path which is base for further action. This path, from the beginning to the end, represents an overall process that contains various sub-processes. This concerns private life as well as business. In particular value-added chains of operational organizations are based on processes. These processes - or procedures – have been established for a long time and often follow the same pattern ever since. But: outliers are not uncommon and sometimes it’s not even clear if the established procedures really are ideal or if they are just there because they have been there forever and a day: “We have always done it this way”.
Was „always“ always good though? Or can „always“ be improved? And what if someone decided himself to leave the common path and changes procedures just because it seems better to him or even only more convenient? Which target procedures deviate from actual procedures and is our assessment of right and wrong correct at all? Or might it all just be based on our believe that our procedures are as good as our perception makes us think? How do we find out which processes need optimization, how do we recognize if somebody establishes individual procedures that do not necessarily help our organization?
In the past external consultants were hired to analyze processes. They made interviews with people in charge, evaluated subjective observations and when found necessary, changed the process. This necessity almost always exists and still many times is not recognized because subjective evaluations come with a catch: They're tedious, not objective, inaccurate, unreliable and expensive since this method only pictures a snapshot based on subjective statements, opinions and individual observations. Also, not just rarely, process responsible managers will try to make their procedures look ideal as they have been in charge of them for a long time after all. Nobody wants to be taught a lesson.
So how can we reach neutral process data? Devoid of individual manipulation und self-serving reasons? By Process Mining – digging for precious metal in a gold mine. Every organization already has its own gold mine, full of precious data: a treasure chest many companies never opened since there was no tool, no key.
Process Mining changes this: The process wisdom already contained in available data, can now be made visible and usable. Digital, objective, honest. No more impact of single-sided thoughts, no more keeping actual procedures secret. Authentic findings extracted from incorruptible data in your ERP system. Process data don’t lie.
Now we know the key to the treasure. But how does it work?
The best way to explain is the following, slightly uncommon image: Almost every ERP system records each event in operations with an allocated time stamp and transaction number. So that data is already there. So far, so good. But: All this process data - or operational sequences - are served on a big plate looking like Spaghetti, unclear where one starts and ends. Prior to collecting the nuggets, someone has to sort the noodles in a clear arrangement. And that’s when Process Mining comes into play.
Data analysts use Business Intelligence systems (BI), such as Qlik Sense, Power BI or Tableau, to collect business data from your system. This data being the foundation for process analytics still is the pasta plate though. So how do Process Mining experts sort this out? By using of a software that adds on to the particular BI system, sorting all time stamps and transactions into a useful order. The Process Mining expert only ensures the software learns all operational procedures correctly and enables it to identify target processes and therefore outliers as well.
From now on the Process Mining software will record comprehensive data for each event at any time and visualize the ways procedures go: The process flow. The program images a process chart that immediately shows which transactions follow on certain events. It also shows if a path matches the target process or deviates from it. The program can also automatically identify and extract all outliers from the target process: The target process filter. The noodles are sorted now. Each Spaghetto is separated from the rest, beginning, course and end of all the noodles are clear and visible.
Now we have an overview of what happens in which order in our organization. However, it’s not only important to know what happens but also how often it does, how long it takes and how much it costs. That’s the part leading us from sorted Spaghetti to the nuggets we’ve been searching for. Thanks to the time stamps recorded in our system we now know when an event kicks in. But all steps and paths sorted in correct order also show us the time span between them, resulting in durations and averages. The user chooses which event to start the analys where to end. Now about discovering the gold:
Each ERP system records all costs, prices, invoices, turnover, payments as well as in- and outgoing goods. All this data including values is digitalized and available in your organization, practically already allocated to certain events (time stamp and transaction number). Based on this each single event, as well as selected groups of transactions or specific time spans can be evaluated at any time. Regardless at which point of any Spaghetto the analysis starts and where it ends: The overall process can be analyzed everywhere during its course. Single organizational units can be extracted and compared with each other, i.e. according to their performance in terms of value and time or with a look at their reliability comparing the target process. That’s the data treasure Process Mining provides the key to.
Long story short: the organization becomes entirely transparent. If something leaves the correct path, takes longer than average, is overdue or moves in a loop: One click on the target process filter will show such cases in an instant. Deviations get already identified when they begin, before they can establish, causing trouble and expenses without anyone questioning them because “we’ve always done it this way”. Process data don’t lie.
Nov 9, 2020
Nov 2, 2020
Sep 5, 2020
Aug 10, 2020
Mar 5, 2020