You have an important presentation. You can present:

· An exciting, cool, document with lots of different chart types. They will never see the same type of chart twice.

· A professional, serious, document with a limited number of types of IBCS-sanctioned plots, that re-appear over and over.

What do you do?

I would obviously fall for the first, and I hope you would too.

Assuming your objective is to convince rather than entertain, this would be the wrong move.

Edyta Szarska from IBCS gives this example.

You have to drive from A to B. There are two ways to get there.

One way is marked with the boring, professional, serious, standard traffic signs we are all sadly familiar with.

The other is marked with flashy, exciting, cool, original traffic signs just designed by the local art school.

Which road do you pick?

I would pick the first and I hope you would too.

Like a road from A to B, a presentation is a way to build consensus on a set of evidences, and lead your audience from A to B. Has nothing to do with cool and little to do with entertaining (except if you are Hans Rosling, but you probably are not).

Given the choice, most members of your audience would have more entertaining things to do than listen to your however entertaining presentation.

The IBCS standard explicitly limits (no radars plots, not funnel plots,…) the choice of plots to a smallish set that, according to IBCS, are the easiest to understand and pack the highest volume of information.

Edyta proved this to me when I showed her my very own “IBCS flavored” dot plot (plot at right).

Dot plots are cool and are favored by many data practitioners. But they do not stand a chance against their IBCS equivalent, the “multitier bar chart” (plot at left) that Edyta suggested I use instead of my very cool dot plot.

Both plots can be plotted with this tool in seconds, because I could not let go of my little creation.

Should you renounce to the plethora of information present in the chart at left but miserably absent from the chart at right (absolute variance value, percent variance value, integrated legends,…) for the higher cool factor of the chart at right?

Not cool is the new cool 😎.

I rest my case.