Data Visualization Personas

If you want to sell better data visualization practices you can’t use the same approach with everyone. Marketers use archetypes and like to create stories around them like if they were real people. Their marketing messages are then tailored for Jane (archetype #1) or Theresa (archetype #2).

Let’s try this. Allow me to introduce you to three of my co-workers.

Co-worker #1: Anna, a Newbie

Anna was asked to create a chart, something that she rarely needs to. After playing with the wrong Excel options, she comes up with a really ugly and inefficient chart (misconceptions about charts don’t help). I show her how a simpler one could solve several perceptual issues. She changes her chart but keeps some of the chart junk (she finds my chart too minimalistic and laments her lack of of graphic design skills).

Anna is now aware of what “data visualization” means – clearly more than creating a few Excel charts. She should work on her data analysis and communication skills and stop worrying about graphic design skills. Corporate culture and peer pressure will push her to the dark side chart junk side of data visualization. I hope this seed is strong enough to withstand it, but only time can tell.

Co-Worker #2: Peter, a Middle Manager

Peter agrees that some of our processes are terribly inefficient and wants to change them. We could try to improve data visualization, but that doesn’t make much sense if everything else remains the same.

I’ve been sharing some tips with Peter on how to create better charts and he starts to recognize a bad chart when he sees one. Problem is, his previous model (Excel defaults, PowerPoint templates) is shattered, but he is unable to create a new one. He feels lost.

My tips seem to make sense at the lunch table but when he tries to apply them something is missing. Tips have this effect on people: they create an illusion of knowledge but the lack of context renders them almost useless. He needs a crash course on information visualization.

At this level, we are not discussing how to improve a chart. Instead, we must discuss how to add best practices in information visualization to the data management model. Selling this to top managers isn’t always easy. But Peter likes to bang his head on a wall…

Co-Worker #3: Frank, A Professional Chart Maker

Frank creates presentation charts every single day. Ten years ago he was creating exactly the same charts, 3D effects and primary colors. He doesn’t recognize the problem and the audience seems to be ok with this routine. If he needs something a little more complex than 3D pies and bars his manager asks me to help. This could spark his curiosity, but it doesn’t.

I believe Frank will never try to improve his data analysis, management and visualization skills, unless he’s formally ordered to do so. He’s not dumb, he’s just a little too comfortable in his comfort zone.

It’s Your Turn

I want to add more details about Anna, Peter and Frank and I’d like you to help me. How do you see them? How are you going to sell them our product (better data visualization practices)? Would you like to add your own characters? An imaginary boss, perhaps? The IT guy? Tell us about them!

Photo Credit: Smaku

7 thoughts on “Data Visualization Personas

  1. Great post. I am struggling with this set of people, too. I am trying to get my workplace to improve its data visualisation standards. We’ve just purchased a bunch of Tableau Server licenses, and are beginning to roll them out – this is very exciting. However, I despair when someone I had hoped would be a champion in their area just doesn’t get the whole data viz thing, and creates some lousy, messy tables. That person is your example #1. I’ve sat down and I think I have planted the seeds.

    I’d add this person:
    #4, Nial, the report consumer
    This person needs reports in order to analyse what’s happening in his area. He has been around a long time, and is accustomed to receiving huge tables, with so many categorical breakdowns, that it is impossible to discern any trends or meaning from the data. New techniques are presented to him, but he is too tied to his regular way of doing things to see that aggregated data presented in a visual way, could help him make better decisions. What to do? Deliver the reports exactly as he wants, but also slip in a few more interesting and effective visualisations. Overtime, we hope he may come to recognise that the visualisations give him more info in less time.

  2. One quick note: the average boss that you’ll encounter will agree with Frank till a overpayed consultant company tells him that the data analysis, management and visualization skills used are from the past!! !:D:D

    Meantime, Jorge character will become undermotivated and turn himself into a “Frank” :D:D:D

    (at least I’ve seen it happen before!!!)

  3. I have lots of #4’s Nial in my circle. They want the same and more of it. Suggesting a new technique freaks them out. I stick with the same as Andy suggested – deliver “the usual”, but try to introduce improvements in parallel hoping that they’ll agree with the value of a better visualization. I’ve had some successes and some that just don’t want something new. All we can do is keep our own expectations high and hope others come to the light.

  4. Andy and 9.2.5
    Slipping in a little extra is great and all, but how will you get the client to pay for it. In my experience, you’ve got to sell this up front and sell it good or else the client will be on the defensive about how much it costs, and ‘why couldn’t you just give me the same as before.’

    What do you do to get paid for your efforts?

  5. I think you might be a little hard on Anna – is ir ‘worth the effort’ to convert her, when all she needs to do is add a little chart to something. It may not be the most elegant, by-the-principles visualisation, however it’s sufficient for her purposes.

    The Annas I come into contact with are front-line operations people, presenting some of their own work / analysis / proposals to their managers – neither are interested in the purist approach, nor need they be.

  6. I would like to add a fifth character: The consultant from the communications department. Everything the company does, goes to a certain extent through the communications people. They are extremely professional and highly motivated people. The believe in the importance of communicating the right message to the right people. However – they are mostly trained in language and design. They look at a chart like this:
    1) It’s got to be familiar to the audience (thus pies are good).
    2) There should be some variation in visualizations, so as to avoid a monotonic ‘look’ – thus making the publication “interesting” and “inviting” (thus no consistent practice).
    3) Each chart should bear the distinct trademark of the organisation – at the very least in the choice of colours, but preferably even more. The effect should be that any chart reproduced anywhere immediately reveals it-self as a company “this-that” chart (much as excel reveals it self) (thus chart junk).
    4) (this is very big with my company) We dont want to risk being accused of fiddling with the data- therefore; just use tables. Everybody understands a table… And then the stakeholder can make their own charts, defining their own take on the data-story (after all, there are lies, damn lies and statistics).

    In all fairness I think the communication people are – to a certain degree right. Some customers really don’t want nothing but tables.

Comments are closed.