Home Bots & Business Lessons learned using process mining – part 2

Lessons learned using process mining – part 2

by Guest

In this article I continue sharing learnings from process mining implementation. This is the second part of the series; you can read Part 1 here.

Balance between targeted analysis and exploration

Process mining technology has an explorative nature and offers a lot of possibilities to analyze bottlenecks in a process. However, to bring value with the findings it is essential to define scope of the business challenges into questions and iteratively relate the findings to them.

Strike a balance between untargeted and targeted exploration

In the initial stage of any process mining project we define questions in collaboration with business process owners and/or the problem owners. These concrete research questions guide a process mining project and scope the analysis phase of the project.

On the other hand, we leave room for explorative untargeted analysis – this is when the most surprising findings are revealed, so unexpected that they would not be considered upfront. Similar to any data-driven exploration, process mining can generate unexpected findings, where concrete research questions are needed to deep dive with further analysis.

It is important to remember that explorative analysis does not guarantee that you will find something impactful to solve the business challenge but potentially can be time-consuming and not best use of resources. As such, the recommendation is to focus most of the efforts on the specific research questions of business instead of getting excited about all the different analysis possibilities.

Start small and show the value, then scale up

There are typically two ways to approach process mining project – waterfall and iterative. Waterfall approach starts with scoping the project with the set of research questions, followed by configuring the tool to allow rich functionality, succeeded by an analysis phase to answer all research questions, after which engagement with problem owners takes place and multiple improvement initiatives are initiated.

Each phase can be lengthy – scoping, development and analysis phases can take up to several months. As such it takes too long before an impact is realized and visible. We have learnt that first insights already give enough potential to start an improvement project; and not to forget the absorption capacity of the business to cope up with the task to implement such improvements.

Work in short cycles and deliver value incrementally to keep your stakeholders engaged

Often it is more effective to apply an iterative or agile methodology, where phases are executed in parallel. As such we would be able to ensure rapid delivery of value and an ability to make scoping decisions and changes throughout the project. This approach is way less chaotic in practice simply because we are dealing with smaller improvement steps and not heading for a big bang. Another benefit is that project sponsors would see value being delivered continuously and sooner, and time-to-improvements would be shorter. In such way you will get continuous support and engagement of senior stakeholders.

You will gain faith of stakeholders in process mining by showing visible improvements in a short period of time. This is much more valuable than diving into long analysis and development cycles, then starting a large number of improvements simultaneously and not being able to demonstrate successful results for a long time.

Equip your team with the right expertise

Implementation of process mining in a large organization on scale requires people with different skills, expertise and experience to be involved. Most of the articles mention four key roles: project manager, process expert, business analyst, and data specialist. We had all these roles in the core team from the beginning.

We quickly realized that composing team is more about skills required during different stages, rather than involved roles. For instance, the person analyzing the process using a process mining tool might have different skills and aspirations than the person who is driving improvement projects with problem owners in the market. The first one would likely have an analytical and process-driven mindset and be tech-savvy; while the skill set of the second one would likely be in the area of project management, change and stakeholder management, and strong communication.

Different skills need to come together to make process mining a success

We have also learned that the business or problem owner is an important stakeholder as well, in addition to process expert. The larger a company, the more complex the stakeholder field gets. In large companies, where the process is executed globally by many people and process owner role is centralized, this teaming up for success becomes an important element.

It is important to put together a multi-disciplinary team of people complementing each other to get the most out of a process mining.

Ekaterina Sabelnikova is Innovation Consultant at Philips Innovation Services


Misschien vind je deze berichten ook interessant