I was going to title this post the “Evils of Analytics”, and whilst provocative and attention grabbing it also would have distracted from a real issue that we’ve started to discuss in ways not seen before. Analytics is a major theme in contemporary business, however it’s crucial that it is balanced with an approach to management that recognises the beauty of humanity.

If today’s analytics were available to economist Frederick Taylor when he developed his theories of Scientific Management in the 19th century, I wonder if he’d have found a correlation between employee wellbeing and productivity, and included this in his theories. What would  standard practice be today if it were included?

This is a question proponents of ‘Human Management’ are seeking to answer. Meanwhile, the school of analytics pushes us further into Scientific Management, in ways that would make Taylor the proudest dad ever. Without saying we shouldn’t pursue an analytics program or systematic improvement through measuring and monitoring key metrics (not saying this at all), we need to be mindful of the power we can yield when armed with a bunch of compelling numbers. We can use them for good human outcomes, and not so good human outcomes. We also need to consider all the relevant numbers, not just those that are convenient.

COVID and elephants?

So what has this got to do with COVID, and what have elephants got to do with it? Well,  there is a plethora of data being generated showing the human impact of COVID, and the elephant is a metaphor for mental illness. The full impact of secondary effects of COVID won’t be known for some time (years), however it has got us talking about mental health in a way or to an extent not seen before. Almost immediately after announcing lockdowns, the impact on people’s mental health was considered.

The mental health response to COVID is by no means a surprise, when considered in the backdrop of numbers from the Australian Bureau of Statistics on ‘working age’ (16-85 year’s old) people that show why this consideration is necessary. No fancy analytics here, just some straight up stats:

• In any given year, 3.2 million ‘working age’ Australians experience a mental illness (20%, or 1 in 5)
• Around 43% of females and 48% of males, or 7.3 million Australian’s, will experience one in their lifetime
• Further to that, The Black Dog Institute reports that every day 6 people will take their own lives, and a further 30 will try.

The mental health impacts of COVID, be that directly through health concerns or indirectly through economic consequences, won’t be known for some time to come. What has been known for some time however are numbers published by BeyondBlue and PwC , showing:

• In aggregate, mental health conditions cost Australian employers some $10.9 billion per year
• Investment in health strategies provides an average return on investment of $2.30 for every $1 spent
• And it’s not just in Australia, Deloitte has published similar results from research into the same issue in Canada and the UK.

The RoI number is considerably higher in some sectors. For example, small mining businesses had an average return of $15 for every $1 spent, and small essential service providers received $14.50 (I suspect these are mostly family businesses and not-for-profits) . Logically this makes a lot of sense. Smaller businesses are more dependent on their people, so their mental health will have a more immediate, direct and proportionately larger bottom line impact than in larger businesses. That doesn’t mean larger businesses are off the hook here, as the counter is a large number of mentally unhealthy people could have a significant, negative impact on the bottom line (which shareholders will not be ok with).

So here we have some compelling numbers. Now let’s say, through an analytics program, you identified an opportunity to get these sort of returns from an investment in some equipment. Affordability and opportunity cost aside, would you ignore it? Probably not. You may also consider if you can afford not to invest in it.

Now, the good news. Employee wellbeing doesn’t have to cost much at all. A lot of it starts with just being a decent person, including being respectful, compassionate and understanding towards people’s individual needs. Practicing a ‘servant leadership’ approach is another. What this all boils down to, is that a key element is providing a good culture; that fosters a mentally healthy workplace.

The dark side of analytics

So what has analytics, or it’s offspring – machine learning and artificial intelligence; got to do with culture and mental health? Well, we’ve already seen perverse, albeit unintended, outcomes from leaving machine learning or similar algorithms unsupervised on datasets that are inherently prejudiced. Examples including

• skewing hiring decisions towards white males, due to the history of ‘success’ from hiring such candidates, when it was actually that these candidates are over represented in data sets due to historic prejudices in recruitment
• the Australian Government’s maligned ‘robodebt‘ scheme erroneously issuing debt notices to welfare recipients, due to inadequacies in the data matching process between Centrelink and the Australian Tax Office

So, while analytics aren’t evil; we need to ensure they are used in the context of ‘human management’ in preference to systematic pursual of ‘scientific management’. It’s important to incorporate factors into analytical models, or the decision making that follows them, that ensures outcomes are achieved in full consideration of human factors. That is, for instance, machine learning models should include data that trains them to be to make decisions and/or recommendations consistent with what a “good” kind of human would do.

By all means pursue analytics, just question everything!

I’m all for analytics, machine learning, artificial intelligence, and whatever else you want to call these things. I’m a numbers guys, my whole thing is helping business owners, boards and executives to know their numbers. Numbers, and by extension analytics, are very powerful in enabling success in business. So by all means, I say  get into it.

However, left unchecked, an analytics program could undermine your organisation’s culture, and with it the mental health of employees. It’s not an absolute, however organisational culture is a factor in creating a mentally healthy workplace. You wouldn’t want an analytics program to go against the grain of your culture, undermining it, potentially destroying employer-employee trust; or worse – someone becoming severely unwell from it.

So, just remember, that “the possession of great power necessarily implies great responsibility” (William Lamb, British Member of Parliament, 1817). Give analytics the respect it deserves, and cynicism it demands. It would be pretty drastic if it took another pandemic scale shock to your business, to get you talking about the mental health elephant in your meeting room again.