What Hans Rosling Taught Me About Numbers on Dashboards

by

Akshay Roongta

During a recent test match on television, I kept seeing the same insurance ad which was making a case for itself, by saying over Rs. 2000 Cr (225M USD) in claims had been disbursed to its customers. 

Every time the ad  came on (which was a lot, especially given how the match went), something nagged at me about the claim they were making. And I kept wondering what it was, until I remembered what I’d read in Hans Rosling’s famous book on data biases - Factfulness.

An Ames room, invented by American scientist Adelbert Ames Jr. in 1946

In his chapter on ‘The Size Instinct’ he cautions the reader to always ‘Compare the numbers’. He says,

The most important thing you can do to avoid misjudging something’s importance is to avoid lonely numbers. Never, ever leave a number all by itself. Never believe that one number on its own can be meaningful. If you are offered one number, always ask for at least one more. Something to compare with.

Be especially careful about big numbers. It is a strange thing, but numbers over a certain size, when they are not compared with anything else, always look big. And how can something big not be important?


There it was. That’s what was troubling me about the ad.

Admittedly, 2000 Cr is a huge number. And in the insurance industry, that number in disbursements is meant to make you feel confident. Signal trustworthiness. More claims paid = more likely they'll pay mine. Right? Well.. Rosling would push back..

To really make sense of the number, 2000 Cr, wouldn’t it be more useful to compare it to other major insurance providers, maybe see it as a percentage of actual amounts claimed, or see those ratios across a number of companies.. Without context, numbers can be misleading, and frankly not that useful.   


In this case, it’s not useful to you as a customer, and maybe you could dismiss it as clever advertising. But if I think about it in the context of MEL in the impact sector, it makes me think about various dashboards, where lonely numbers around output metrics, or statistics from surveys stand out. 15,000 people trained or 85% response rate on a chatbot and so on.. Either too big, becoming less believable, or too small, making one feel despondent.

And unlike insurance advertising, these numbers shape decisions about where resources might go, which programs get scaled, what interventions get funding.

When a standalone metric is all you have, you can't tell if you're succeeding or failing, if you should double down or course-correct. The number might make you feel confident or concerned, but it won't help you decide wisely.


So next time you come across a lonely number, take a moment before reacting, and ask a few questions. Try and tease out other numbers so you can understand that original number in context.

Akshay Roongta

Co-Founder Dots by Ooloi Labs

With over a decade of experience working across WASH, public health, financial inclusion, agriculture, and education, Akshay brings a deep understanding of how complex, ground-level realities can inform better decisions and systems.
Before co-founding Dots, Akshay worked with a range of nonprofits, networks, and mission-driven businesses to enable collaboration, learning, and long-term change. His work blends systems thinking, participatory research, and product design to create tools that help teams work meaningfully with qualitative data. He believes that the most valuable insights often lie in lived experience, and that good technology should help uncover and act on those insights without losing their nuance.