Successful automation - a matter of principles

What value are we trying to create?

What does better look like?

How can we make it better for humans?

These are the sorts of questions automation projects don’t always start with. And we find ourselves turning this around, reframing and ensuring we are all clear on ‘the value’ of what we are doing. So what’s the process supposed to look like? Well, I think we’ve got to the alpha stage in the industry, but there are already some key learnings that can be shared.

We tend to look at automation opportunities as the ‘automation of activities’ or the redefinition or jobs and business process (separating the jobs from the tasks). This latter approach typically generating far more value. And when we look at this people typically assume we are talking about the lower skilled, lower wage roles. This myth is understandable, the last industrial revolution did impact the lower end more than any other. The truth is automation has the same potential anything else has, if applied in the pursuit of new value it will in fact create exactly that. And this applies to all levels and role types, with the complexity of a lot of tasks not often being an issue.

As per McKinsey’s recent research - a focus on occupations is misleading. Very few occupations will be automated in their entirety in the near or medium term. Far more likely is that certain activities will be automated, requiring entire business processes to be redefined, and jobs performed by people to be rewritten accordingly (the overused example of the bank teller’s job being redefined with the advent of ATMs). I’m more optimistic, I think we can look at most Job Descriptions and simply agree that there is a difference between them and the reality of the role performed and the tasks undertaken - with the automation potential to take away the noise and low value tasks very early on in the MVP cycle. Take consultancy or agency environments, for example, time-sheets are being removed, and I’ve yet to see a job description with this task called out, and yet it takes up everyone’s time. So removing this task allows these people more time focus on the job they are actually employed to do.

However, it is also important to look at this from two further perspectives. Firstly, if we aren’t entirely automating jobs away then there has to be a relationship between the machine and the human, and that has to be carefully mapped with those humans. As the machine hands back to the human we have already seen ‘trust’ issues, and in some cases the human will simply repeat the tasks if this trust doesn’t exist. And as most automation projects are still written with a negative business case this simply leads to negative or low ROI. Secondly, the tasks that are most complex / have a high cognitive load shouldn’t be overlooked. Often when getting machines or systems to fulfil tasks people see the greatest value where there is the greatest pain, and this is often key to successful adoption.

A final observation - often people only look at what they are automating, not considering what other value this can create beyond task reduction and implied cost saving. Often there are many more moments of value that can be extracted from this; and whilst analysing a particular process and stage we can identify value beyond automation as well - experience moments often being a good example. So, for example, if an automation moment means there is real-time insight available removing the tasks of analysing and reporting there is often the ability to change how this information is reported and consumed.

Summary:

1. Separate jobs and tasks
2. Put the human first and develop the solution for and with them - for the wider benefit of humanity
3. Drive for all the value potential (and create the right backlog)
4. Data rights, privacy and security are as ever sacred and need to be considered at all times

Human centred design (HCD) and Rapid iterative testing and evaluation (RITE) models are well used in building experiences, but these are being applied to automation with ever increasing benefits. Although there are deeper questions afoot, these are big questions that will need to have the involvement of everyone in a position to drive this change, government, senior business leaders and most likely you. Placing the human first approach at the forefront of this prcoess is of course the only answer to achieve true value and help answer the ethical questions on all our lips - so lets start with these principles and work forward from there.