Sonar Book Club: Factfulness and the concept of ‘lonely numbers’

Pretty rare when a book can change your world view – but that is what this did for me! A few months ago I read Factfulness by Hans Rosling. It is pretty rare when a book makes you take a step back and question the way you see the world, but that is what this book did for me. It starts by asking you 10 multiple choice questions, and then gives you the answers. I got most of the 10 questions wrong – revealing that the way that I saw the world wasn’t accurate, and also that things were actually a lot better than what I thought. The book goes through the 10 questions, in detail, with data to back up each answer.

I loved this because I find typically we hear so much negative about the world, that it was nice to read something, backed by data, that had a positive message.

The other thing I liked was that the tool provides you with tools to apply when questioning things that influence our world view. There are some great rules of thumb here: lines might bend, look for the majority, question your categories, but most notably for me on this was to ‘beware lonely numbers’. The recognition that single numbers on their own can be misleading, and to always look for comparisons, more numbers for context, or ideally be able to divide numbers.

I find the concept of ‘beware lonely numbers’ to be a well articulated rationale for digital marketing analytics. Often at Sonar, we are asked to comment or report on a single number or metric. For example, ‘Is the bounce rate good or bad?’ Or a revenue amount for a given period. What we typically suggest to clients is to use a few key metrics to better understand health of a website or trends in revenue vs other metrics over a period of time. In very much the same way, we are saying ‘beware lonely numbers’, understand the context of the user experience by understanding the same metric over time (eg. bounce rate over time), and relative to other key metrics (overall website traffic, revenue, and key user journeys) to a collective view.