Process Mining projects

Insights

Process Mining projects

April 17, 2020

Authors

  • Andrea Taglioni

    Director

In the previous article we talked about what Process Mining is, what are the main tools and features, and what are the typical Use Cases and benefits.

In a nutshell, Process Mining is the set of IT tools and professional services capable of collecting the (digital) traces left by the processes on the information systems and, from these traces, reconstructing, graphing, measuring, and simulating the process in all its run-through variants.

We also highlighted that Process Mining tools are useful in one-shot analysis (BPR, Assessment RPA), but their value is even greater for measuring changes (also periodically) and above all as continuous monitoring tools (e.g. anomaly detection, continuous performance measurement).

Listen to the Podcast

What are the typical Process Mining project phases, what are the typical activities and timeline, which actors are involved?

We will try to give Bip Consulting's point of view in this article.

Before a project, in the Scoping phase, a first definition of the process area is made (number of processes, macro-volumes) in which we try to understand whether the minimum requirements for tracing on the systems exist (does the data exist? Are they easily extractable? Is there the minimum enabling data set?) and if there are already evident pains or suspicions of inefficiencies presently plaguing the business process.

From this Scoping it is clear whether Process Mining can be applied, and an implementation roadmap, timeline and cost is defined (more on this later).

We then move on to the Implementation Phase; the phases of a typical process mining implementation project are the following:

Phases of a Process Mining project (Source: Bip Consulting)

  1. Process discovery and data assessment: a first phase of refinement of the process knowledge, identifying in detail the stakeholders, the operators, the information systems on which the process rests and reports/logs to be used, and obtaining information on volumes (also useful for choosing and sizing the tool and related licensing). In this phase, the Client has objective evidence of the percentage of the process actually traced by the system and of the "non-measurable" portion of the process. The process mining tools are gradually being enriched with Task Mining features that can also capture events NOT tracked by the systems but evidenced by individual productivity tools (mail, office suite) in compliance with the Privacy and Security laws and regulations (e.g. GDPR). In this phase the first considerations on the quality of the traced data are also made.

  2. Data preparation: a phase of data collection and engineering, which also includes the identification of the structured mechanisms of consolidation and reconciliation (if the data come from multiple information systems). In this phase detailed quality controls are also performed on the data collected and the processed feeding files are produced, which are then loaded into the tool.

  3. Process analysis and improvement hypotheses: this is the phase in which process analysts, leveraging the appropriate instruments, expose evidence, document, and measure process KPIs, in order to export all the insights on the process analysed. Typically, this takes the form of analysis documents and measurement dashboards created with the Process Mining tool.

    In this phase, improvement actions and updated policies and regulations are defined together with the client-side Stakeholders.

    These actions are carried out by the Customer/client (with or without the consultant's contribution) in parallel with point 4 (see below); the actions could last for hours, days or months depending on the type of change required.

  4. Operational monitoring: this is a crucial phase whose importance is often underestimated. The action of Process Mining up to point 3 is important for defining the optimization plan, but then what does actually happen? Process Mining technology was created to be used also as a discreet (periodic) but above all continuous (daily, hourly or even near real time) monitoring tool.

    In this phase of the project, therefore, we work to define KPI dashboards, and prepare governance to monitor the process and understand if the actions lead to the desired effect.

    In the most virtuous cases, the alerting functions are activated, and a real governance of the anomalies is carried out as they occur in the logic of anomaly detection or anomaly prevention.

The interaction with the Client's contacts takes place through a series of workshops, which primarily involve the business representatives (Process Owner / Process Experts and optionally also Process Operators) throughout the project, representatives of IT / System Owners especially in the definition of flows and creation of extractors, but also during the installation of the tool (if on premises).

The Process mining tools are mainly Cloud based, but some allow the adoption of on-premise solutions or a private cloud solution (on the cloud of the Customer/client), should the necessity arise.

A typical process mining project can last for two / three months per process (depending on the complexity of the process and the sources given), with possible synergies and parallels for multiple processes, or if the same process is to be analyzed in several countries but the data source is homogeneous.

Sometimes Clients want to proceed with a Proof of Concept, in which they can grasp the potential of the instrument, a tangible evidence they can ‘touch’ by hand, "turning it on for the duration of the pilot only", feeding it with impromptu extractions and not integrating it into the actual business process/IT infrastructure. In this case the typical duration is around 5 weeks.

As regards the instrument, on the other hand, the licenses are typically listed as an annual fee, by process or by number of lines / events, with the possibility of varied bundles to satisfy all needs.

The return on the investment (ROI) depends on the process, on the basis of the as-is efficiency, and obviously on the method of use (one shot assessment, periodic monitoring, or constant monitoring / alerting), but the Payback Period is normally very quick (the ROI can be obtained in a few months).

The success of the first pilot or of the first complete realization, often leads to the start of scale-up programs of both variety (extension to other processes) and geography (extension to other countries for Multi-National Corporations), which can be devised to maximize the parallelism and synergies of man-days (reducing the unitary impact of project costs per process), and licenses (going towards multi-process bundles and multi-year contracts, typically more discounted).

We invite you to contact us to get more detail about Process Mining HERE.

Get in touch

Are you ready to make sense and make things?