Business leaders understand the value that data analytics can bring to their organisations.
In Australia last year, they collectively spent billions building out capabilities. Capabilities, it turns out, some don’t trust, they then under-utilise it and therefore don’t realise its full value.
KPMG International’s recent Guardians of Trust study found only 35 percent of respondents had “a high level of trust in their own organisation’s use of different types of analytics”. A quarter of respondents “admit that they either have limited trust or active distrust.”
This could explain a Melbourne Business School-A.T. Kearney report last year that found “only 8 percent of companies were extracting the full value of data analytics”.
Unfortunately, data distrust among senior executives is a known problem. We see it all the time.
It’s not that there isn’t enough data or the right data - it’s that the leadership of many companies choose not to use it. They might superficially talk about how valuable it is, but when it comes to making important decisions, it's business as usual.
For whatever reason, they find data a risky business and go back to what they know: “These are the decisions I’ve made before, so I'm going to make the same ones again.” The problem is the safest option isn’t necessarily the best option to drive the organisation toward success.
Just to be clear: no one is asking leaders to suddenly let data drive all of the decisions. But they could trust their data a little bit more and take what it says into account.
Experience and data combined is a “magic sauce” to unlocking value and realising growth opportunities that aren’t obvious based on a leader’s experience alone.
Validation in all the wrong places
There is a view in some circles that trust in data can only be achieved when the analytics reaches the same conclusion as a human or organisation previously did, thus validating the algorithm or model.
Applied in a leadership context, a leader may look at an analytics system as essentially codifying their decades of experience – or some amount of institutional knowledge. But, they may want to check that the data model and its assumptions can replicate what the human decision would have been in an identical situation.
This is a slippery slope
Arguably, it is better to approach data analytics and its output with an open mind; to be able to look at what the data is telling you, not just to find evidence a decision you already have in mind is the right one.
Confirmation bias occurs when people use data to justify what they already wanted to do. Ideally, you want data to help you make the right decision, not justify your own biases in reaching a particular conclusion.
This is a very difficult and challenging area, but it’s important that leaders are open to seeing the difference good data can make, and therefore to using it more in their decisions.
What’s my assurance, then?
Good question. KMPG International says that trust issues often arise because leaders are asked “to make major decisions based on the output of an algorithm that they didn’t create and don’t always fully understand”.
“As a decision maker, you really need to have confidence that the insights you are getting are reliable and accurate, but many of these executives can’t even be sure if their models are of sufficient quality to be trusted. It’s an uncomfortable situation for any decision maker to be in.”
If the data model gives you something completely different to the way your intuition would have structured the problem or response, how – as a leader – do you then validate which one is correct or which path you go down?
The answer is that this should start a conversation – one that perhaps the executive or organisation might not have otherwise had.
You might go through this process and realise the data-driven response won’t work, but just going through it is likely to lead to new ideas and different directions to pursue.
You might also find that many people using the same source data get different answers, in which case it’s an opportunity to ask questions, identify assumptions, and work out whether some assumptions are more valid than others.
With tools like Tableau, leaders have the opportunity to keep asking more questions about the data in front of them. The capability to go in and ask follow-up questions is another way to validate what you are seeing.
Leaders should ask their staff for data to validate ideas and proposals, just as they should be doing the same themselves.
The question should be asked more often than not. If nothing else, it enables the data conversation to continue.
Mac Bryla is an APAC Technology Evangelist at Tableau Software.
