Process mining, which can discover, monitor and improve processes by extracting knowledge from information systems, is a phrase that has been banded around a lot. But it is still in a state of infancy in HR, particularly in the UK.
Starting as a pure research topic in the late 1990s, it has proven commercial value in areas such as sales, manufacturing, accounting, procurement and consumer efficiency, where there are direct links between revenues and processes.
With ‘people’ decisions increasingly being recognised for their financial implications, it has also spread to the periphery of HR during the past five years.
Tony Habschmidt, HR consultant at data visualisation company Zegami, says: “Only the more sophisticated organisations currently actually know what process mining is, and even many of these don’t.”
Process mining tool providers like Celonis and Fluxicon can point to no more than the odd example of UK HR usage. There is an acceptance, however, that the waters could be muddied by it happening under a different label, or by central process teams working together with HR and other departments.
Although broadly positive, most comments about its suitability for HR refer to potential rather than usage.
Edward Houghton, head of research at the CIPD, says: “I haven’t come across any process mining being used in HR as yet but it’s ideal for the very repeatable processes HR does.”
What is it?
A specifically holistic subset of data mining, process mining uses ‘event logs’ (events that occur that can be recorded) as units of analysis. Once a dataset has been formed machine learning algorithms are used to look for trends.
The tools will transform the information into a process flow, providing a visual feel underpinned by data, helping organisations understand if they have issues and how to prioritise potential focus areas for continuous process improvement.
Fabian Veit, senior vice-president, people and operations at Celonis, says: “It’s like an X-ray for your organisation to shed light on processes, and it is relevant to the whole HR journey from hiring to retiring.”
While this is certainly much better than having an opinion or relying on a gut feeling for process management, it will only give part of the picture.
Emma Leonis, executive director at HR transformation consultancy LACE Partners, says: “Combining qualitative and quantitative data can provide a rich view, but process mining is only part of the quantitative. Those who use it must be clear on what they are trying to achieve.”
Where can it be used in HR?
Experts feel process mining could be used for everything from expense reimbursement to managing employee requests for training and holiday entitlements.
The recruitment process
The CIPD’s Houghton says: “Internally a recruitment process could certainly be mined for data. Processes such as individuals applying, going to interview, being offered and accepting jobs have data attached to them.”
Celonis is using process mining for its own recruiting system, looking at the speed of processes and why people are dropping out.
LACE Partners’ Leonis says: “It can be used for writing contracts, getting them approved, out to individuals and signed.
“A lot of organisations, particularly large ones, have an electronic signing tool, leaving a physical imprint so data can be analysed. For example, who did it, when it happened, how long it took, and the workloads of the people involved.”
HR query management
Leonis continues: “A lot of this activity will be managed through a system, which will know when the query was made, who made it and what it was about. Then process mining tools could tell you how effective the process is for answering questions for employees.”
Flexible benefit schemes
Zegami’s Habschmidt says: “It could be used to see what benefit choices are made each year, to look at the efficiency of the enrolment process and to determine whether benefits on offer are actually relevant or useful to employees.”
What are the downsides?
At this stage the potential usages seem almost limitless, but with any new technology, users should also watch out for pitfalls.
Possibly more data available than expected
Anne Rozinat, co-founder of Fluxicon, says: “Don’t overlook existing reports or the contents of your data warehouse, and don’t be afraid to request small manual samples as a first step just to show what can be done.”
Don’t analyse data blindly
Rozinat continues: “You must know the domain well or work with an expert who does. Data being analysed by a process-mining tool just shows what IT systems have recorded, and there may be misunderstandings. So you need to close the gap between data and real life.”
Being sufficiently digitalised
Celonis’ Veit says: “You must be very mindful of where you stand in the journey of digitalising the organisation. If you are still very manual then it’s probably not the right time. The system should be useful for the hire-to-retire journey, and if you are recruiting you need an applicant tracking system.”
This piece appears in the March 2020 print issue. Subscribe today to have all our latest articles delivered right to your desk