Why Explainability Is the Foundation of Trust in Data

When you were a child, how many times did you use the argument with your parents “but everyone else is doing it” to rip holes in the knees of your jeans or dye your hair blonde or steal beer out of the fridge to take to a party? Only to be countered with “well, would you jump off a cliff just because all your friends did?” Wow, the frustration I felt at such a logical but basic [insert eyeroll] argument. It didn’t help that my own mum had seen it all before as the matron of a boarding school.

While it may have been frustrating to a gangly teenager with a truly terrible haircut trying to keep up with his friends, it is an important message and lesson. That we should always question rather than blindly follow or obey. This is particularly true when it comes to the growing role of Active Intelligence in business; the triggering of immediate actions from real-time, trusted data to accelerate business value across the entire analytics supply chain.

And there is one word that I want to focus on in that sentence: trust. Active Intelligence relies on a symbiotic relationship between data, machine and human. As Sally Eaves put it in her article ‘A matter of trust’ in Qlik’s new magazine, this relationship is more akin to “teammates or partners bringing together complementary strengths.”

At the center of this relationship is trust. But not blind trust. The human element is there to not only complete the action, but ensure that the KPI or metric the action is based on can be explained. The good news is that the ability to provide evidence and visibility of the data as it has transformed has become much easier with the move towards Explainable BI.

Why is this important?

Without trust, we are not comfortable taking the risks or making the changes needed to be responsive, innovative, and forward-looking. To make the most of data in driving better business outcomes, we need to be able to trust it and the process of the intelligent machines that compel us to act. We need explainability.

The reason we need it now – and why I’m writing a blog on the topic – is three-fold.

  • Data has been democratized across the organization with the rise of self-service analytics. Every business department is now an active participant in collecting, consuming and analyzing data and there are more users of analytics than ever before. However, in the opening up of data, our understanding of where the data originated and how it has been edited and manipulated – the data lineage - has got lost along the way. Many now struggle to answer the simple question of “where did that number come from?”. And trust in the insight is lost.
  • That in turn has created an analytics adoption problem within businesses. People won’t act if they don’t have confidence in the insights. And few today are willing to bet their fortune or reputation on a data point they can’t back up.
  • Explainable BI is critical to collaborating with data. Better decisions are made when data informs the discussion; where we can question it, challenge its output, and tell the story of its potential impact. Understanding the lineage of that data is the foundation for these discussions and when considering what other data could be introduced to provide even richer insights.

To trust in the potential for data transformation, we need widespread adoption of Explainable BI to provide visibility into the data’s lineage – the source, governance and business logic applied to the data that has informed the insight. We need to understand what transformation that data has gone through to get to that decision and how trustworthy is it. That visibility empowers individuals to take actions in the business moment with confidence and seize new opportunities.

This is only going to grow in importance as we move closer towards more widespread adoption of the Active Intelligence model, where we’re not only using the data to inform our own decisions, but it is proactively served to compel us to take actions and trigger automated responses. That is only possible through trust.

The organizations that have stood the test of time are those that have adapted, innovated and taken risks. Consider the continued success of General Electric. The conglomerate embraces “Finding solutions… Not just imagining. Doing”; it was the first business in the world to own a computer and continues to make surprise market expansions into new areas, like renewable energy.

Businesses have to take risks. But that is very different to making risky decisions. If I had to choose a route based on insights backed up by explainable, trusted data and a decision based on ‘a number on a page’ – I know which one I’d go for.

Trust in #data is central to building a data foundation. Read more on this in our latest blog post by @jamesafisher

 

In this article:

Get ready to transform your entire business with data.

Follow Qlik