A recent report by SSON, the 2022 State of Industry IA Market Report, found that 53% of survey respondents in APAC are considering investing in process mining, 45% are considering process discovery, and 41% are unsure which option is better.
I was joined by Ziv Ilan, Head of Kryon Professional Services, Nintex, at the Intelligent Automation Asia Pacific Summit recently, where we discussed the merits of process mining and process discovery. We also talked through five key questions to ask if you are reviewing which technology is the best fit for your organization.
First, let’s take a look at the difference between process mining and process discovery.
Process mining versus process discovery
Process mining has been with us for the past 30 years. The technology uses system logs and databases to map processes and identify bottlenecks, and a strong level of expertise is required to interpret the output data. A good example of process mining is a financial services company using transactional data stored in their system logs to identify repeatable processes.
Process discovery is a more recent technology. Robots are deployed to individual user machines to identify processes from the user or front-end perspective, by taking screenshots and gathering metadata. Process discovery tracks what users are doing as they complete tasks. Each action performed by a user can be tracked visually, which makes it easier to identify standardized processes, as well as variations across people, teams, and geographies. For example, with process discovery, you can see how someone in Sydney executes a process versus someone in Singapore.
One of the easiest ways to explain the difference between the two technologies is in this way – process mining is system-focused and process discovery is user-focused. With that in mind, during the webinar, Ziv and I discussed five key questions you should be asking when considering process mining, process discovery, or both for your organization.
1. What are the trends that are driving the uptake of process mining or process discovery?
Remote and hybrid work has changed the way organizations learn how workers complete tasks. When most workers were on-site, understanding processes would have been done in workshops or sitting with team members and watching them work in an office environment. A lot of that learning is now done virtually.
We are seeing more mature organizations consider a move towards process discovery, often in addition to—or to replace—process mining. And the pioneers of this move to process discovery are often financial services and insurance organizations, as well as large organizations with global operations. Mid-sized organizations exploring process discovery are generally more focused at team level automation, for example, finance and operations teams across various industries that have multiple repeatable processes.
The key driver for the move towards process discovery is that both groups are being asked to do more with less, and they are looking for ways to remove inefficiencies and automate operations. One of the biggest challenges teams face is finding time away from daily tasks to document processes – which ultimately will help to deliver more with less.
There are multiple advantages to documenting processes. Organizations take control of their in-house knowledge, which can be shared across employees and teams no matter where they are located. This means that work can be completed faster and more cost-effectively. It also removes the risk of knowledge walking out the door if a key employee leaves. New employees can also be onboarded faster, and all employees will be trained to complete work the same way, following the same processes.
2. What are the key challenges?
As we talk to some of the larger organizations about where they are in their process journey, they often hit a point where their continuous improvement program stalls. Why? Because these types of organizations usually have some type of process mining already in place, and the most obvious candidates for automation have already been identified. In these cases, process discovery can help by providing better visualization of what is happening in the field and how people are completing tasks and processes. This allows organizations to uncover new areas that would benefit from automation. Skills shortages are another area that is currently proving to be a challenge across all industries. Automating onboarding and training can help reduce the time it takes to onboard new employees or upskill and reskill current employees. Process discovery uses guided or assisted automated bots to build training based on a discovered process learned from current employees. This allows organizations to provide hands-on learning of real-world work situations for new employees, which reduces time to productivity.
3. Do you have to choose between process mining and process discovery?
For many global organizations, the answer is probably no, especially for those who already have data mining in place. We have seen a lot of organizations add data discovery to their data mining implementations, and it often complements the technology that is already implemented.
4. Where is the best place to start?
Understanding what problems you are trying to solve is the best place to start. Once you are clear on that you can assess the strengths and weaknesses of each solution and choose the option most suited to your requirements. An important reminder here is that data discovery is more suited to user-led processes, while data mining is often more suited to system-led processes.
5. What does the future look like?
There has been a move towards data discovery in the market. It provides an ideal platform if you are looking to build a Center of Excellence, and we are currently seeing this being prioritized by many organizations. A strong advantage of data discovery is that it can be retrofitted natively to the way an organization is already working by automatically discovering processes.
The next big step change we are expecting to see is data discovery used in conjunction with process mapping. This will provide a powerful solution that will allow you to take your discovered processes, and document, map, standardize, and ultimately, automate them at scale.