I’ve just had the pleasure of reviewing an “improved report” for the executive of a board of a UK company worth many billions of pound. Now I’m not about to start sharing their confidential data, but most of the graphs looked an awful lot like this:
Does it ring any bells? It should for anyone who has been in spitting distance of a corporate PC in the last 7 years. It’s a good ol’ default Excel 2003 graph. It’s the one that you get when choose a default Column Chart.
Why does this rattle my cage so much? Am I obsessed by presentation over content? No, there are two reasons why this kind of sloppy presentation offends me.
My main objection is that using default graphs shows that zero thought has been given to the way in which the data will be read and interpreted. In this example, why would you show the target as a separate bar (apart from the template graph dictating how you show your data)? A small target bar (in a different colour) overlaid on the actual would make far more sense. Where’s the context for the data – specifically a trend?
I also think that the MS default template breaks some basic rules of effective data presentation:
Spare use of colour – why have a grey background? What extra information does it convey? It uses extra toner when you print it and adds nothing to understanding.
Gridlines. If the values are important why not use a data table as well? The are extra clutter and add little to accuracy.
Too many digits on the axis. These are big numbers. People are generally poor at coping with big numbers. If you need to deal with something big, deal in 1000s or millions as fits – don’t dazzle with 10 digit numbers.
Border – more clutter around the graph area this time. The box adds nothing, so why have it.
Use of colour. Why blue bars with a black border line? Colour can be fantastically powerful at steering the eye to important highlights, here it’s doing nothing. Why not make the bars a mid-gray with no border and save colour to pick out specific columns, targets or exceptions?
None of these points sounds significant on their own, but roll them all into one graph and you end up with a graph that shows some thought for the user and some design.
Next time you need to create a graph, look at the details on the graph and ask the simple questions “What does that do for the reader, does it inform or obscure?”.