Businesses around the world have long been recognizing the urgent need to digitize processes in an accelerated manner. However, after the first waves of digital transformation projects, it became certain that one should never digitize for the sake of digitizing. Digitalization has to be done in a strategized manner based on findings from different tools and methods that help with the analysis of tasks and processes in organizations and give a clear picture of where the company is in its digital development path. One of the faster-developing tools is task mining.
What is Task Mining
Task mining is a technology that helps companies automatically analyze how work is done on users’ desktops or workstations. It is a relatively young field within bigger categories of data science and business process management.
Task mining includes a set of techniques and algorithms that together help to identify and extract information about the tasks being performed by individuals within an organization. That information can then be used to identify the bottlenecks and improve overall efficiency.
In simple terms, task mining helps companies get a better understanding of how the work is being done by employees. This gives them visibility over where the time is usually spent and how it could be improved. The larger organizational goals behind engaging with task mining tools could be improving productivity, increasing employee well-being, or decreasing costs.
Task mining was made with the goal of improving employee satisfaction and not as a surveillance tool as it may seem at first glance. While it does get information from users’ digital traces, the insights it gets, as a result, can actually help to see where the inefficient work is done. If that gets eliminated, employees can have better chances of working more effectively, not having to do repetitive and tedious tasks that most people hate doing. It can also help managers to make more informed decisions about how to allocate resources and improve overall productivity.
There are several technologies that are very similar to each other in terms of the goals they achieve and thus get confused often with task mining. The closest ones are process mining and process discovery. Although they might be aiming for the same improvements, the underlying technologies and analysis methods are slightly different and will be discussed further in the article.
How does the technology work?
Task mining monitors what is being done on people’s computers, extracts the needed information, and then analyzes that to bring meaningful insights. It uses various algorithms and techniques to identify patterns and trends.
Task mining is focused on analyzing individual tasks rather than end-to-end processes. The task is a smaller step in the process. An example of user interactions done on computers could be mouse clicks, keyboard shortcuts, etc.
That being said, task mining technology is constantly evolving and the algorithms and analysis techniques it’s using are also getting more advanced.
Steps in task mining
1. Monitoring and recording users’ activities
The first step in task mining is to record all the activities that employees are doing while working. For that, special tools are implemented on computers. They record mouse clicks, copy-pastes, and all the scrolls from whitelisted applications.
2. Understanding the background with Optical Character Recognition (OCR)
From the recordings that have lots of information, OCR helps to find bits of information that could be useful for further analysis. Those pieces of information could be words, numbers, links, etc.
3. Leveraging Natural Language Processing (NLP)
After all the necessary information is collected, NLP helps to group similar tasks and provides the data on them in a digestible way.
4. Identifying business tasks
Once activities are grouped together, they will be matched with the corresponding business tasks that need to be improved.
Examples of typical tasks in task mining:
- Copy-pasting the information
- Uploading/downloading the documents
- Scrolling through the systems
- Mouse clicks
Task mining tools
Task mining is a relatively new technology, but companies have already been picking up the trend. There are several bigger task mining vendors, and process mining companies have increasingly been acquiring task mining companies as well. Some of the more popular vendors are Nintex, Pega, UiPath, Kofax, Abbyy.
The ecosystem is growing further, bringing to life many innovative startups that are bringing more depth and better ways of discovering business data.
Task Mining vs Process Mining
Process mining and task mining are two technologies that get mentioned very often when it comes to improving business processes and tasks. In fact, many companies use them together in order to achieve complete discovery and analysis of knowledge work processes. This trend can also be observed from large process mining vendors as they’re actively broadening their portfolio of services with task mining capabilities. With the similarities being mentioned, the techniques and underlying methods they use are different and are worth being mentioned.
One of the main differences between process and task mining is where they get the data from. In the case of process mining, it uses the log data generated by systems and technologies, e.g., enterprise resource planning systems. On the other hand, task mining gets the data from people’s actual use of all the whitelisted systems and applications.
Another key difference is, of course, the level of granularity. Process mining helps to record every major step in the process that event logs produce the data for. For example, in the Accounts Payable process the steps could be:
- Invoice is received,
- Invoice is reviewed,
- Invoice is approved,
- The amount is paid to suppliers.
Task mining, in contrast, helps to see what happens between these bigger steps through snapshots and recordings. In the above case of Accounts Payable, the tasks could be opening the invoice document, copy-pasting the invoice number between systems, and sending the invoice via email for approval.
Both technologies are very effective and are able to help managers with valuable insights into the workflows. Depending on the use cases at hand, either of them or both could be a good place to get started with digital improvements.
Benefits of task mining:
- Improved understanding of work. Task mining is able to provide a data-driven view of how work is being done, which is extremely useful in large companies with hundreds and thousands of different processes and tasks.
- Better efficiency & productivity. With task mining, people can find tasks that are not being carried out in the most effective way and focus on improving those. Moreover, it can help to eliminate boring tasks and increase productivity.
- Increased customer satisfaction. As a result of more efficient workflows, teams can have more time for more value-adding activities and focus on improving customer-facing tasks. Task mining can help in making the quality and speed of services better.
- Enhanced resource allocation. Task mining provides data-backed information on different workflows within an organization, which in turn eases the decision-making process, allowing managers to allocate resources in an effective manner.
Challenges with task mining.
- Adapting to new technology. Task mining has been around for quite a short time, and compared to process mining, is just emerging.
- Accuracy. As it is a young tool and its techniques are developing every day, the accuracy of algorithms might not be the highest compared to more established tools.
- Expensive to implement. Task mining licenses might not be very expensive, but together with deployment costs they might add up to bigger sums.
- Security issues. For companies that have sensitive information, such as insurance or public organizations, it’s risky to trust task mining tools.
- Usability. Some task mining solutions require users to initiate, pause and end task tracking manually, leading to inefficient work.
Privacy issues with task mining
One of the main concerns companies have when it comes to implementing tools that have at least some elements of tracking is privacy. Monitoring what employees do sounds very alarming and sensitive, and managers don’t want to compromise employee well-being for efficiency gains.
However, one should understand that these kinds of tools are not for spying on someone, but rather to make their job easier. Without knowing the issues in current processes, it’s impossible to know what should be done to eliminate those issues. A lot depends on the buying companies’ intentions as well. Algorithms that different vendors use vary, however, there are many advanced techniques that help to find bottlenecks in workflows in an anonymous way without exposing any individual bits of information.
Check out this advanced process+task mining tool that processes the data in a completely anonymous manner.
Task mining use cases
There are many different areas and business cases where task mining could be used. Many of them are unsurprisingly similar to process mining and process discovery use cases.
- Discovering bottlenecks in the work. Task mining brings insights that help to find inefficiencies in the workflows and thus assist in decision-making processes.
- Automating the activities. Task mining helps to reduce and eliminate manual and monotonous tasks, freeing up more space for creative work.
- Improving collaboration. Task mining brings visibility into interactions and dependencies between tasks and people, making it easier to communicate and work together as a team.