What statistics and language have in common and what Marketers could learn from it.

Edward Taaffe
5 min readJul 1, 2019

Delusion about absolutes.

In order to be able to discuss things and achieve things we need the ability to give things names and these names need to have a roughly shared meaning to everyone in the conversation. You are probably thinking, “If only” and you’d be right on the money, because in reality; very few things if any, obey strict rules that make them identical to others in their class.

Statistics are a mathematical manifestation of the same issue in that they predominantly define averages which are then used to define loosely almost every member term in the group.
A “Large business” has a turnover between x and y, yet you may hear heated discussions about big business that seems to assume they are exactly the same because they are described as “big”.

People will talk about Americans, Israelis, Russians in precisely the same way, yet how could you possibly describe an American other than in terms of where he lives.

More than half of a tree is totally integrated with the earth but we refer to a tree as though it were a unique standalone object.

Even a person contains a very high proportion of non-human matter in the form of organisms that live within us, sometimes carrying out critical functions for our health.

In all of these conversations people are very content to live with generalisations as though they were facts and indeed they are confident that on the law of averages they will be right more often than wrong and even to the point of forgetting that the subject matter is not, in fact, as absolute as it sounds at all.

Every time a Mathematician, that master of absolutes, sits down to think great thoughts he will begin with an assumption such as, “all cars are 105 times the weight of their owner within an error factor of 5%” then he will write something like “Let’s assume that all cars are cubes” and there goes the theory. From here on in we are in the land of make believe.

Calculating the time it will take to complete a project, or indeed the cost, is always based on the average time or cost for that type of project. Every living PM knows that on average his estimates will be wrong on average.

As Sam Savage[i] puts it, apologies Sam if I have reinterpreted your wise words, Knowing that the average depth of the Colorado river is 2.5 feet wouldn’t save you drowning in a 30 foot pool, but quite possibly the opposite. In fact, by and large, averages and everything built with them or relying upon them is either useless, fatally flawed or seriously dangerous.

Don’t be average

Business decisions based on averages are usually also heavily flawed and especially in Marketing. The fact that the average customer arriving at this page with x=y leaves without purchasing does not mean that this customer would not spend a lot of cash left to her own devices. The fact that I searched for women’s stuff yesterday does not mean I am gay, it means someone’s birthday is approaching. If people were nearly so predictable, there would be no Marriage Counsellors, no shrinks and probably even no jails, certainly no marketers.
Margaret Thatcher PM of the UK has often been misunderstand for her insightful comment;

“They are casting their problems on society and who is society? There is no such thing! There are individual men and women and there are families and no government can do anything except through people”.


While we all know that she was indeed against the state doing a lot to help, what she meant was that a government can only address individuals and deal with their individual needs. it can’t deal with society with a brush stroke. Another way to put it is this; while helping the individuals and families with their problems you can help society, but it doesn’t work the other way around.
This is such an important lesson that I will break with tradition and add yet another example. I once inherited a sales team and had to learn to keep them performing. What my mentor taught me still serves me daily. I have two conversations to maintain, one with the sales people and another with their managers. To the sales people, whenever I addressed them; by researching the market and knowing the product, if they do their best all day every day, they can achieve this fantastic target and earn a big bonus. Notice, I never mentioned averages.
To their managers, if you keep them motivated and remove barriers within your capability; then by doing their best, they will achieve the average conversion rate on the average daily activity and you will earn your very large target bonus.
Think about that for a moment. The average is real and it helps the manager in the right context, but in the hands of the salesperson it would be disastrous. Averages are always historical, not current and not future.
Let me put it another way, if you set out to perform to the average, your average performance will be very poor indeed on average.


Whether we are using words to hold a conversation, or using averages to calculate how much we hope to earn next quarter, its vital to be aware of the weaknesses of words and of numeric assumptions based on averages.
You are not Mr or Mrs average from hometown, you are an individual.

The average temperature during the last ice age was 10 degrees Celsius, surprised? Words may not have very strong absolute meanings, but with effort you can ensure that your precise meaning comes across to your audience and you understand theirs, not a rough estimation. As an afterthought, when you are next dragged into a discussion about AI and you definitely will be, then remember just how weak our mastery of language and numbers really is even when alive with our eyes wide open and take the cautious approach.

[i] Sam Savage A Stanford Professor and author of “The Flaw Of Averages “ a book from John Wiley & Sons, published June 2009 https://web.stanford.edu/~savage/faculty/savage/



Edward Taaffe

Ed is a technical consultant and writer in the areas of Digital and Products. He writes here on random subjects that catch the eye.