How to implement Process mining proof-of-value initiatives

In this fourth part of our process mining series we outline the approach and key insights in the act phase of our process mining framework. As shown in the illustration below, the act phase is the intermediate step in our approach to process mining, where proof-of-value initiatives are prioritised and executed, building on the process diagnostic insights gained in the preceding focus phase.

Our structured approach for the act phase

A structured approach is required to drive ahead with and act upon the process mining insights generated in the focus phase. Key services and processes need to be introduced to industrialise and further embed process mining within your organisation. Demand management processes capture opportunities and prototype developments before pilot projects are executed, encompassing proof-of-concept and proof-of-value initiatives. Roll-out and operational processes, such as solution maintenance and continuous improvement, also need to be introduced. The end-to-end process mining life cycle processes and services, from demand management through to operate, are supported by key capabilities and enablers (see illustration).

Generation of testing hypotheses through lean tools and value driver trees has proven to be a key success factor for prioritisation and execution of improvement initiatives during the act phase. A high-level value driver tree for the procure-to-pay process – one of the more transactional and straightforward process mining use cases – is shown in the following illustration. Value driver trees for each process mining focus area can be a starting point to rapidly identify potential improvements and capture the associated benefits.

The structured approach outlined above lays the foundation for translating process mining insights and initial process diagnostics into a set of prioritised improvement initiatives across several end-to-end processes. The aim of these improvement initiatives is to act on the detected improvement potential and capture value within the context of continuous improvement, in line with Lean Six Sigma (LSS) values and with the overall ambition to employ process mining at scale. This structured problem-solving and “hypothesis-based” approach does not preclude a more “exploratory” approach where data scientists and process experts jointly analyse the data in the process mining tool to assess and identify potential insights and areas for improvement. We have found that the best way is to combine both approaches during the act phase.

Our experience is that during the act phase the most successful companies broaden the scope of process mining so that it is embedded into their strategic technology transformation programmes, ranging from robotic process automation (RPA) to workflow optimisation and enterprise resourcing platform (ERP) upgrades and deployments.

Process mining has proven to be an accelerator for such transformations as it facilitates an as-is assessment and process discovery across key end-to-end processes by complementing process workshops. From an RPA perspective, process mining enables proper understanding of automation potential and monitoring of automation progress during roll-out and deployment.

The same holds true for process mining as an accelerator for ERP upgrades such as S/4 HANA transformations, creating transparency for process design across all project phases (pre-migration, template development, implementation and post migration). In addition, process mining can be used as a further accelerator for all implementation scenarios such as a new implementation, system conversion or landscape migration. The main benefits relate to as-is process mapping, identification of key process variants and archetypes and generating data insights to remove unnecessary process modifications for a future system set-up.

What was the context and client background?

Deloitte was asked to support a significant E2E transformation programme across a range of core process functions to enable the client’s strategic vision. For Order to Cash (O2C) the project focus was on optimising global business processes, standardising master data management and embedding a customer-centric approach, all with the goal of enhancing ways of working and the resulting customer experience.

Why did the client want process mining?

Process mining was expected to be useful to both the wider transformation team and business stakeholders. Within the overall transformation context, process mining supported the decision-making process, generating additional insights to make the transformation project faster, reliable, and more accurate. It created transparency about existing business processes, drove deeper understanding of the influencing factors on the business and made comparison of the performance of different business units possible.

How was the process mining project approached?

A comprehensive data model was created for the O2C process along with tailored dashboards in the dedicated process mining software. This incorporated collaborative feedback from customer service and finance teams to develop new transparency and insights into the end-to-end process. A tailored value driver focusing on key process bottlenecks and major process steps was employed to drive project execution.

Results of the process mining

Based on validated and customer-tailored O2C dashboards and KPIs implemented according to individual O2C business requirements, several areas for improvement were identified across the main 10 business markets. One particular finding was that after comparing various markets one was found to have a particularly low throughput rate, owing largely to the order intake process. A further deep dive with the process mining tool employed found that 30% of orders were being blocked and in 90% of cases released on the same day. These blocks were discovered to be due to low credit thresholds and a “one size fits all” approach for credit limits applied equally across all customer groups. One of the key improvement levers was to undertake a structured customer segmentation approach with associated appropriate credit thresholds being applied, resulting in a significantly increased throughput rate for the end-to-end process. This reduced the overall workload for customer service reps and increased customer satisfaction.