With news, social media and conversation completely dominated by the COVID-19 pandemic, I was hesitant to write a blog that even mentions it. At the same time because I am so passionate about Business Intelligence and can see how it is being both used and abused in this situation, I felt compelled to share my viewpoint.
Firstly, let’s talk about the data and the metrics being used to track the pandemic.
The three main metrics being tracked in this pandemic are:
- Confirmed Cases
- Total Deaths
- Total recovered
This first metric requires people to be tested and, as we all know, that is only possible in places where testing is available (and confirmation takes a few days) and only a fraction of people have been tested. Health experts have explained that many more people have already had the virus without any or very mild symptoms. As more testing becomes available this first metric will increase significantly. Unfortunately, this has not stopped some people from dividing Confirmed Cases by Total Deaths to get completely wrong information about the death rate of this virus.
A few years ago, when the BI world was obsessed with Big Data, I started many of my BI presentations with the following formula:
BIG DATA + BAD DATA = BIG BAD DATA
The COVID-19 situation needs a new formula:
MISSING DATA + GOOD DATA = MISSING GOOD DATA
We are really missing data and so the metrics are incomplete and, as such, we should ask ourselves the question: “Are we measuring the right things?”
I have always believed that Business Intelligence is only 50% about analyzing the data and that the other 50% is the human action taken as a result of that analysis. If that data analysis is based on missing or bad data, then actions taken on it are likely to be flawed.
Some universities and institutions have built out predictive models based on this data which are even more likely to be erroneous.
Business Intelligence can be and is being put to good use in this crisis by organizations analyzing trusted data sets which are usually their own. Industries which have been hardest hit like airlines, entertainment, hospitality and restaurants can use it to intelligently reduce their costs as well as better understanding the impacts of options for their customers, employees and overall business.
Other organizations that are currently being overwhelmed like Healthcare, Supermarkets, Distribution and Delivery can also benefit from good Business Intelligence right now and every business can use it to analyze their supply chain and be better prepared for the inevitable global economic slowdown.
Some of the best BI solutions I have seen have been in the area of spend analytics. Controlling costs will be important for a lot of organizations in the coming months and I plan to write about this in an upcoming blog
On the healthcare side, back in 2016 I blogged about how BI Dashboards were helping drive healthcare data from analysis to action in the UAE. Over time, those dashboards evolved to creating a command center to monitor metrics real time from multiple hospitals . Among other life-saving benefits, they were able to direct ambulances to the closest hospital that had beds available for the particular patient condition. They were also looking at creating a mobile application for the public to see wait times at Emergency rooms of the different hospitals.
There is no doubt that Business Intelligence will play a key role in helping organizations navigate this current crisis both from a healthcare and business perspective as long as we use good data and measure the right things.
We can all start by asking the right questions given the circumstances we find ourselves in. What worked in the past, may not be what is needed right now. After the 2008 Financial crisis, many organizations changed their key performance metrics as they learned they were not measuring the right things and I am sure many will change their metrics again after this crisis. I urge us all to take a step and analyze our data not in a reactionary mode but to consider what we are measuring and if we have all the data we need to measure it.