How Process Mining Technology Creates a more Agile and Informed Enterprise.

Quantifying Root Causes behind KPI movements.

Traditional Business Intelligence (BI) tools and capabilities have long had an Achilles’ heel; they only allow us to understand a process through a series of predefined snapshots. Colloquially put, this can be described by the common business phrase “what gets measured gets managed”. Conscious decisions are made on what Key Performance Indicators (KPIs) and data visualisations are presented for operational or management reporting. Human intuition and experience link the different KPI snapshots and their trends to build a story of what is happening in the process. However, this human intervention increases the risk of ‘confirmation bias’; where data that confirms a hypothesis subconsciously increases in importance in contrast to evidence that might contradict that hypothesis. This could lead to misinterpretation and ineffective management. The commercialisation of Process Mining technology now allows organisations to create a quantifiable link between process KPIs and the activities in a process. Companies can now react faster in a more informed and agile manner to process inefficiencies or detrimental activities. This delivers tangible increases to working capital, productivity and profit margins.

Better Understanding of Process Execution

Process mining technology extracts data from the same transactional ‘system of record’ applications that business intelligence tools connect to. Yet in stark contrast to business intelligence tools, process mining technology also extracts change log data to build up an event log of the as-executed process, linked together with a single process ID common to all the process activities. Therefore, organisations can report process execution and activities alongside more traditional analytical components. As a result, process mining can correlate the contribution of different process activities with the improvement and degradation in business KPIs, a blind spot of business intelligence tools.

Using a procurement process as an example, upon extracting the data we can understand the supplier on-time delivery achievement for each purchase order. However, using process mining technology we also know the different process activities that each order underwent, we can also average the supplier on-time delivery based on all the orders that contained a certain activity or process sequence. As an example, we can answer the following question very quickly in a data driven way; “what is the impact on supplier on-time delivery on process activities occurring after purchase order issuance?”. From activity data, process mining technology allows us to also see the impact of timing. Say we understand that amending purchase order quantity after issuance causes a 10% reduction in supplier on-time delivery. We can equally understand as an example that changes made within five days of the requested delivery date, cause a reduction of 30% in supplier on-time delivery, compared with say 5% reduction if made at least five days before the requested delivery date. Using extracted metadata, we can increase our specific understanding of this effect on certain master data, such as suppliers, items or contractual agreements.

A screenshot showing process mining output from Celonis. Price changes after issuance lead to a 93% late delivery rate, compared to an average of 33% of all purchase orders sent. On average the price change is made 7 days after PO issuance.

Process mining technology allows each process KPI to be viewed at activity level and therefore quickly identify root causes behind process inefficiencies. Equally as significant, it allows us to quantify the impact more accurately in a way that traditional business reporting capabilities previously hasn’t allowed us to. Companies that have already adopted process mining talk candidly about their reaction when they saw their process execution laid out in a way that previously was unattainable through other process mapping and process improvement methods.

Driving Agility by Highlighting Opportunity Cost

The insights of process mining allow organisations to respond more quickly to process inefficiencies and increase the agility of an organisation. A common observation of processes is that the upstream perception of issues differs greatly to the downstream impact experienced. This results in an underappreciation of cause and impact in an organisation’s business processes. As highlighted earlier in this article, process mining allows organisations can quickly identify root causes of process inefficiencies and understand their impact and contribution to process KPIs. Used in an operational manner, this increases agility through improved prioritisation of work and process understanding.

Process mining data better allows organisations to drive the impetus to act and to suggest next best activities in an operational setting. Analysis of a historical dataset using gives us the insight and understanding of process inefficiency. However, by using process mining in near real-time, we can identify similar situations evolving and allow an organisation to take better action sooner than would have otherwise been the case. As an example, we can understand the contributing factors to late delivery of goods to our customers. From the historical data we know the importance of those different factors in affecting late delivery by observing the degree of change they have on the KPI and the frequency they occur. Thus, in real-time when an order is created and the process starts to contain certain activities, predictive data analytics are now available. Overtly simplified; “given steps X and Y have occurred in the process, at this juncture given prior process execution knowledge we would expect a late delivery probability to increase to Z%. As a result, organisations can act sooner with more confidence and awareness of the risk of inaction. This ability to quantify the adverse impact of ‘doing nothing’ is critical to improve the agility of an organisation as it drives the impetus to act.

We are now moving into a phase of execution management and hyper automation

It is now time for organisations to push forward from insights and understanding into the realm of execution management. This ability, using process mining technology, to cross the data analytics divide from ‘descriptive’ and ‘diagnostic’ to ‘predictive’ and ‘prescriptive’ is what drives improved agility. It supports employees in making much more informed decisions, especially those new to a company, role or those with less experience. It helps to strengthen interdepartmental ties as now all process stakeholders understand the potential impact of certain activities and can better collaborate to mitigate the risk. It can improve the negotiation and performance with third parties, knowing how your process variants affect your supplier’s activities can allow you to better adapt or demand improved performance from them. In short, the increased understanding of the precisely as-executed process creates a more informed enterprise and drives the need to act, through showing clearly the outcome of not doing so.

Process Mining is Becoming More Accessible

The commercialisation of process mining has grown over the last decade to provide organisations the opportunity to reap the benefits. It represents a seismic shift from the business intelligence capabilities that have been common place over the last two decades. Many extraction routines from well-known and widely used enterprise software have been defined and templated to improve the time-to-value of process mining. That said these extraction and transformation routines remain flexible to allow organisations to connect to home-grown systems or those without pre-defined connectors widening the use of the software within an organisation.

Process mining has provided a significant tactical improvement opportunity to processes without the risks of software upgrades, resulting in many cases to significant income statement or working capital improvements. That said, in equal measure it has helped companies streamline processes ahead of upgrades, deliver financial and productivity improvements in doing so, and increase process standardisation de-risking the subsequent migrations.

Many organisations have adopted and continue to widen their use of the technology. It has proven itself as an enterprise ready software capability that is rapidly moving from the periphery to become a core capability to support processes. It will be interesting to watch the continued uptake of the technology over the coming years.

Share this post!