I was delighted to see that my recent blog on “why I hate LinkedIn” was popular. As a middle-aged English man, there are few things I take greater pleasure in than telling the world what I hate.
So in this post, I want to elaborate on another thing that irritates me in content marketing: that is, the lazy use of statistics.
Someone (exactly who is disputed) said there are “lies, damned lies and statistics”. That doesn’t mean statistics are inherently mendacious. It means that numbers are very often used to give a bogus plausibility to weak arguments.
Too many people invoke “stats” like drunks use lampposts – not for illumination, but for support.
Statistics show that of those who contract the habit of eating, very few survive.
George Bernard Shaw said that. He’s right, isn’t he? You can’t say it’s false. There is an unavoidable correlation between many of the habits of the living and eventual death.
But so what? Eating is not usually the cause of death. There are any number of correlations between facts, but correlation alone says nothing at all about causation or anything else.
So, it may be true in a correlative sense that “companies that tweet more get more sales” or that “video content gets 150% more clickthrough than non-video”, but these are equally “so what?” statements. It’s pub quiz material at best.
We know that some companies that tweet a lot really suck at it – so the impact it has is negligible or even negative. JUST TWEETING MORE will not lead to your sales going up.
Come on – we’re beyond having to justify what we do with tabloid headlines, aren’t we?
99% of all statistics make no sense out of context, commentator claims
For businesses and other organisations, I think this bears on the question of what it means to be “data-driven”. Being driven by data means using it towards achieving a goal; it involves being driven as much as it involves data.
The DIKW model is often used in information theory. I think it can be very helpful to anyone who wants to be meaningfully data-driven, by giving a framework for evaluating the material at hand.
The model describes the functional and structural relationship between data, information, knowledge and wisdom. It often gets called a pyramid, because you need more of whatever’s lower down to support what goes on top.
Data is a mass of signals, symbols and facts without context or any place in a wider schema. As Ronald Reagan famously (but accidentally) said “facts are stupid things”.
Only when facts are processed into Information and given a description can they have meaning or relevance to a context.
By synthesising more Information from a variety of sources, one can generate Knowledge – which is most simply characterised as “know how” from practical experience. It’s not just more Information, but enough Information to base decisions on.
Finally, synthesised Knowledge becomes Wisdom – the ability to increase the effectiveness of a process and to evaluate reasons for doing something at all.
For example: that rumbling sound you can hear is Data. That it is a truck coming towards you is Information. Your Knowledge tells you to get out of the way of the truck and Wisdom tells you that you will have to dodge fewer trucks if you don’t stand in the middle of the road.
When you’ve only got a hammer, every problem looks like a nail
“Companies that tweet more get more sales” masquerades as Knowledge. It lets you think it’s a causal statement, implying a course of action. But it’s isn’t – it’s Information. Alone, it’s just a statement of a correlation. To try to act or plan on the basis of Information alone is to fall into the trap of seeing every problem as a nail because you happen to have a hammer. You must always be aware of whether you’re dealing with Information or Knowledge.
This parallels with a model that we often use at Axonn Media – the Golden Circle outlined by Simon Sinek.
Information concerns questions of “What”
Knowledge concerns questions of “How”
Wisdom concerns questions of “Why”
Sinek urges us to “start with why” and that’s an approach which the DIKW model can help with when assessing what any data could be telling us – because “why” is usually a matter of defining into actionability any given goal.
But goals are never givens. A goal is inherently a “why” question – it always concerns a purpose (“We want to sell more to increase our profits”) - and a comparative assessment of related “how” questions. One way to sell more would be to drop prices by 90% - we reject that as a course of action because selling more at 10% of the price defeats the profit object of selling more (usually).
To be data-driven, you have to construct and test your goals. You have to decide what information is relevant to answering the question you are asking and how you will evaluate it (and how you will collect it, store it, reassess it over time etc). It might be glib, but being data-driven is not an action; it’s a way of life. Or a way of living.
That’s why I can’t agree with Hilaire Belloc – I know: you wait weeks for an early 20th-century poet to appear on a content marketing blog and then two come along at once. I don’t agree that:
Statistics are the triumph of the quantitative method, and the quantitative method is the victory of sterility and death.
Statistics are tools, means to ends. Being data-driven means having a purpose to drive towards. That purpose is never just a given – it has to be worked at, and it involves decisions on values. And making decisions about values is what being alive is all about.