Top Reasons for Process Mining Failure

How to Minimize Risk of Failure

As with any business program, there are many challenges that can derail even the best-laid projects. To increase your odds of success, ensure you have the following items checked before kicking off any process mining initiatives.

1. Lack of Stakeholder Buy-In

In general, process improvement is not typically associated with back-office activities, but rather something is done on a manufacturing line. Therefore, when trying to persuade teams on the value of process improvement and more specifically process mining, it can be very difficult for people to grasp the opportunities and insights available.  This can lead to stakeholders who do not adequately understand potential benefits and therefore view process mining as a useful activity but not critical for success.

In practice, this can mean not receiving adequate resourcing, low SME involvement, and worst of all a timeline that is not achievable to perform analysis and find value. All this can doom an individual project and more importantly can damage the reputation of the overall improvement programs a company is looking to undertake.

Tip: Develop a project charter with key timelines, goals, hypotheses, and SMEs before kicking off any project. It is better to have a smaller well-defined project than a large project which might not be successful to build enterprise capability and buy-in.

2. Limited Data

Process mining requires data to be successful, and in most cases a lot of data. In enterprise environments, there is always the challenge of finding a readily accessible data source and then interpreting tables and customizations in a timely manner. If a direct source connection is required, there are times when this data is just not easily attainable which means ingestion and analysis must be put on hold. When determining project plans, it’s often an overlooked hurdle in unlocking this data and as a result, a process mining project can quickly fall behind before any process mining work is started.

While these “data delays” are easy enough to overcome, the real risk is with disappointed stakeholders and missed opportunities for readouts. As this is a somewhat newer technology, it’s critical to focus on the value and insights, not why projects are behind schedule.

Tip: Include this connection as part of a charter pre-requisite. Once data is populated, build some simple analysis and present it to SMEs as a “first-look” to validate the process and quickly identify areas of customization or further required data.

3. Goldilocks Scope

When launching a new process mining engagement, it’s important to ensure that there is an adequate scope. I like to start all our projects with a few hypotheses or questions where SMEs are looking for insights and opportunities. Then it’s critical to ensure that you have a scope that gives enough flexibility to investigate and find other opportunities. Too small and you risk not finding any major insights and too large and you risk finding insights that are too large to become easily actionable. The key is to find a system and process that spans multiple sub-processes, sites, or product lines but is not too large where a few SMEs can provide insight.  One of my key principles is that process mining is a program, not a project. There should be enough scope to find some quick wins and get insights, but also enough opportunities left that can be actioned in subsequent projects once the team finds some initial success.

Tip: Once data is loaded into a process mining tool, the process mining team should be able to return some initial insights and start on-boarding team members within 4-8 weeks. Any longer and you risk moving into a situation where you’re building analysis that cannot be quickly actioned. Remember, you can always add more systems/sites/insights later, but project success typically hinges on how quickly and enthusiastically the SMEs can drive change.

Every process mining project is different, and I’m sure many of you will have more items to add to this list in terms of watch-outs and failure reasons.  Drop me a note or submit a comment with your key learnings.