making better presentations using a modelling approach

I wondered for a long time about the title for this post, as it sounds like click bait. You know the sort of thing “This simple hack means you give better presentations than anyone else in your company“. But it’s just a new way of exploring the mindset for better presentations that come to me in the shower (don’t ask!). Any expert who you ask about how to make better presentations will say a variation of the same thing – but I’m hoping that this approach makes it easier to apply. (If you don’t want a five minute read, there’s a two minute video at the bottom!)

Let’s go back to basics. My PhD is well in the past but one of the mantras of my co-workers sticks with me. “All models are wrong. Some models are useful.” (George Box – attributed.)

better presentations scattergraph example 1

Suppose you have a two dimensional graph of how fast your car is vs how long it takes to get somewhere. All other things being equal, the faster you can drive, the sooner your journey is over. In a graph that might look something like this.

Now suppose you have to make presentation where you explain this concept. Okay, in this instance it’s not hard, but it could be something like “machine downtime vs profits” or more complicated relationships such as marketing spend and turnover.

A better presentation is one your audience can remember

No one can remember the data for the dots, but they can remember the pattern. That’s pretty straight-forward and obvious. On the other hand, for many presentations there’s a need to pass on the details. But if no one can remember the data in your presentation, what’s the solution?

It’s obvious in principle – though how you do it is variable in reality of course!

The aim of a good presentation is to:

  • provide the pattern to your audience – for some of them that’s all they’ll need
  • get your audience excited enough in that pattern that they want the details – they can’t get them from your presentation of course, ‘cos it’s too hard to remember (and boring!)
  • provide the data to those who want it after the presentation – it might be a handout or another physical takeaway such as a report or even a download of the raw data!

However you do that last bit, don’t try and put it in the presentation

Here’s where it gets to related to better presentations

better presentations scattergraph example 2

Obviously the way to best describe this relationship is by adding a straight line through the middle of the dots to capture the pattern, right? Sure. And yet that’s not what presenters do. Why? Because that’s only a very crude approximation.

In our anxiety and our desire to be seen as experts, we make presentations which are the equivalent of a graph which joins all the dots!

better presentations scattergraph example 3

Yes it’s accurate, but exactly how helpful is it?

Presumably the rests of the presentation goes on to carry on joining all the dots to the right of the graph but who cares because the audience has fallen asleep by then.

In short, better presentations go for the straight line graph option!

Better presentations are just about drawing the straight lines?

Let’s set aside the fact that in the real world, depressingly few relationships are straight lines the same idea applies for your presentation no matter what shape the actual pattern is… so yes, better presentations are about drawing the lines, straight or not 馃槈

However, while that’s going to give you a better presentation, it’s not going to give you a great one. To give an even better presentation you could try looking at what a modeller would call outliers. That is, points on the graph that are a long way from the red line (we call it the trend line).

Take a look at the very first dot in the graph – top left. Then ask “what’s different about it that means it’s not like the others?”. This is an idea opportunity to make a better presentation by any (or both of!):

  • a story which looks at outliers like this one
  • a side note exploring why the general pattern isn’t perfect

Think of it this way… better presentations talk about the pattern and even better presentations talk about the patterns and the exceptions.

The story of the data

What we’ve just been talking about is the story in the data – that is, things that the information we have tell us. May I take a moment to talk about a common mistake presenters often make? Imagine that you’ve been told to make your presentation more interesting. “Using stories makes for better presentations” goes the (mindless) mantra.

But what stories? If you come from a scientific and/or technical background your first thought might be to scrabble around and alight on the the first and obvious story – the story “of” the data. That is, you start to talk about how the data were collected. I’ve got news for you… no one cares about how you:

  • struggled to find the dirty bits of your data
  • were up until two o’clock analysing your content
  • suddenly had your ah-ha moment in the shower

Actually that’s not literally true of course – but think of it this way people are only interested in how you collected the data if that’s the point of your presentation itself – or if they’re already interested in the story held in your content and they’re checking out it’s validity.

Once again, in video

Leave a Comment

Your email address will not be published. Required fields are marked *