Why Task Mining? - Its place in the Future of Work

I have been thinking of automated Process Mining and why it isn’t ubiquitous yet. It seems like such a good idea – you can’t really innovate and optimise until you understand exactly what the business is doing now – the current physical model. And people are more likely to buy into future innovation if they have confidence that you understand the reality of what they are doing now. Process mining can help you do this – and gives you a firm basis on which to build an evolving work environment – a “new logical model”, which doesn’t exist physically, so far.

So, exactly what is Process or Task Mining? Well, both Process and Task Mining primarily automate the capture of operational data from user terminals and possibly forms and other process deliverables; store them as structured data; and then analyse them, so as to build up a task or process model. This is very much a high-level, even over-simplistic, view – but I don’t want to go into implementation detail here. I will say, however, that mining, if mismanaged, could be construed as an invasion of privacy, so it is important that you explain to your workforce what is going on in advance and that you get some kind of “informed consent” (and never abuse mining tools to identify, reward or punish individual workers).

Process Mining is large in scope, usually complex and involves the whole organisation. Task mining is smaller in scope, optimises task execution and operates at the departmental level. In a Mutable world where tasks and the technology are no longer tied to the office, task mining could be a useful way to support an evolving (Mutable) organisation as it adapts to the Future of Work – or, better perhaps, the Future of Business. I am not saying that Process Mining isn’t a good idea, just that it involves more organisational commitment, cultural change, and significant availability of resources. This, along with an increased risk (associated with its organisation-wide impact) to the organisation if you can’t get it to work properly, possibly because of some flaw in your organisation, means that Task Mining might be a good first step, even if you intend to proceed to full Process Mining.

I have been talking to UltimateSuite about Task Mining. Ultimate.Suite is the product of UltimateSuite s.r.o, a Czech company with a history with Robotic Process Automation (RPA) etc. going back over a decade, although it was only founded, as a spin-off of the automation and Process Optimization team from its parent services company, StringData s.r.o., in 2020.

You would, apparently, use Task Mining to increase the understanding and knowledge of your workforce (and, I presume, capture good practice); optimise process at the task level, largely by identifying and removing bottlenecks; and increase efficiency and effectiveness with RPA where (and only where) it is appropriate. Ultimate.Suite claims some impressive ROIs from all this; but (although I am sure it can justify these) I am always unsure of how much of such figures is due directly to a new product and how much to a newly revitalised workforce talking an interest in what it is doing for a change. In any case I’d want to see the small print around the calculation of ROI figures, and I am more interested in holistic business outcomes (possibly including social factors) anyway. It is becoming accepted that use of ROI goals often “oversimplifies a very complex decision-making process” – see the Harvard Business Review – although my annoyance with large ROI figures doesn’t mean that I think Ultimate.Suite doesn’t work!

Task Mining is supposed to help you address common pain points in the organisation around Workforce Management, such as low productivity; limited use (or knowledge of) good practices; remote people management; and (I like this one) poor employee experience. It also facilitates better process and task management, in particular identification of the root causes of process workarounds and deviations (that management may not even be aware of); and identification of non-compliant process. It achieves this through intelligent automation, where it is appropriate, with impact analysis of the effect of automation and innovation on the business and even assessment of the IT (automation) environment from the user’s point of view.

Of course, all this would apply to process mining too, but remember that task mining is at a departmental level, so it’s possible scope of impact is less and there should be fewer barriers to its introduction. If you want to become more Mutable (evolution-focused) as an organisation and want to continually improve working practice, I think that Task Mining could be a useful facilitator. Possibly the main risk from using it might be that you ossify the existing process you have discovered, rather than seeing it as an opportunity for improvement – but no tool can survive being used by the wrong people. However, Task Mining, even in that case, might well bring bad practice and aversion to innovation out into the open, where it will be hard to ignore.