Improving processes starts with understanding processes. That is why every automation project starts with a thorough process analysis. But even the most detailed analysis of a business process only tells part of the story. With task mining, the analyze of work that happens locally on the employee's computer, we find the missing link between business process and daily work.
When you start working to improve and automate processes within your organization , you probably think of business processes. The workflows that ensure that the business comes in, the customers are served and the invoices are sent. And by streamlining these and automating them with RPA, you can make a lot of profit.
But there are other processes within your company, at a completely different level, that you cannot see as well.
Think of a legal assistant who always starts every new document by copying the table of contents from a model document. After all these years, she no longer even realises that she does this three times a day. And all her colleagues seem to do it too. And sometimes they take the wrong version, or things go wrong with cutting and pasting. Things that have to be corrected later on.
Downloading forms every day, only to modify them and upload them into another app. Starting your day by looking up the schedule in your inbox and opening the five apps you'll use for the rest of the day. Manually classifying and assigning incoming messages to teams. There are countless hidden tasks that people do every day without even thinking about it.
These are the things that do not surface in workshops, interviews or observation sessions. Or they do surface, but are not given priority because you are not going to start a whole iT project because it remains unclear how much time they actually take.
With task mining, an observation tool watches an employee's computer. The programme contains smart technology, such as machine learning and text and language recognition. By registering which apps are used and how data travels through those apps, the programme can find recurring patterns that have an impact on your business processes. You get a clear picture of which tasks take up the most time and which tasks recur most often. You then use these 'missing puzzle pieces' of your process data to find candidates for further automation.
Process mining, the technique whereby we use data flows and systems analyze to discover undocumented business processes, is similar to task mining, but focuses on the data that is visible in your back office. To make those business processes work, and to get the data into your systems, people do all sorts of things locally on their computers all day long that you don't see in process mining. Opening e-mails, using spreadsheets, modifying documents in Word... All things that remain hidden from process mining algorithms. But they are often essential, recurring parts of a business process. By combining data from task mining and process mining, you get a very detailed picture of your business processes, and for the first time you can really see where your people's time and energy is spent, where mistakes are made, and where you can achieve the most with automation.
As soon as you, as an employer, start putting software on your people's computers that tracks throughout the day exactly what they do, which applications they open and what they click on, you will meet with a lot of resistance. Task mining can be experienced as 'spy software' that keeps an eye on people and checks whether they are working hard or fast enough. This is not how task mining is meant to work. It is important to be open about this and to regard the employee not only as a data source, but also as a partner in the process. Ultimately, you also need the human intelligence of the employee to find out why certain things are done.
Ultimately, of course, it is the employee who benefits most from the automation of recurring tasks. With the current developments in RPA and no-code development, the time is also not far off when the employee will start working with the data himself and create his own 'digital colleague'. In a somewhat more distant, hyper-automated future analyze we will no longer be manually processing task mining data, but software will do that for us. With the tools from UIPath it is already possible to generate a list of automation candidates, including flowcharts, or even a prototype on the basis of task mining data. This functionality will only mature in the coming period.
Thus, every day we come a step closer to saying goodbye to boring, repetitive and error-prone work and, thanks to our digital colleagues, we can focus on the tasks that people are meant for.
I understand.
