Designing Dashboards that Deliver: Translating Tableau Insights for Business Users

With Tableau dashboards, we need to understand our business users. What are they doing with this information? How long are they really going to spend looking at this? And, once they have what they need from this, what questions do they ask next?

Designing Dashboards that Deliver: Translating Tableau Insights for Business Users

I never did artistic interpreting in ASL. People are often surprised, as I come across as quite artsy. Here’s the rub: I loved the English form too much and couldn’t make a proper match in ASL. Take poetry, for example:

Some say the world will end in fire,
Some say in ice.
From what I’ve tasted of desire
I hold with those who favor fire.
But if it had to perish twice,
I think I know enough of hate
To say that for destruction ice
Is also great
And would suffice.
Robert Frost

For as simple as this poem is, there’s a lot going on. Fire and ice, both used literally and figuratively for hate, starts the poem. The rhyme creates a beat and each corresponds to their related idea. For me, this is where I put my focus, because it’s the part I most relate to. The point of the poem requires some heavy thinking. Compare this to Crom Saunders, an expert in artistic ASL translation. What Crom does here is genius.

For non-signing people, make a 1 and then make a 5 with your fingers. We call each of these a handshape. Now watch the video, and you’ll see his hands do these shapes a lot.

To translate the rhyme into an ASL-focused context, Crom favors signs like ‘this group’ rather than a more literal ‘some (people)’ to keep with 1- and 5-type handshapes. This is how you do poetry in ASL. Fire favors more open 5 handshapes while ice favors a more clawed handshape. Simple, subtle, but genius.

He also adds navigation for his audience, being more explicit about transitions. A very literal translation with the cues might look like this.

This group, over here, they say the world with burn and fade away
This group, over there, they disagree and say no, it’ll freeze (to death)
(Brief pause with question coming)
My experience so far with (heated) desire, I support this (fire) group
But, if the world perished and destroyed, but resurrected again, I imagine understanding hate very clearly
That equal to that –
The world freezing and crumbling –
(This would do)

Translations like these do no justice to their original form. Notice how much more information is included in the opening. This is pretty standard in ASL and is called scaffolding. The ending, with an allusion to a crumbled world in his hand, is softer, more inferred, but an extremely potent device in ASL. It’s something you can only do in a signed language, to literally hold a figurative point.

So why does this matter in dashboard design?

When we think of translation, we often lock in on form. Too often, we expect people to provide a fairly literal translation. This is why many of those machine-translation texts do decent at rote things and terrible at ideas that are highly abstract or require clarifying. As analysts, the things that we understand are different than the things normal business people want to understand. Take the above ASL translation – how much more are you going to look into this?

Dashboards also rely on visual means to make a point. While we can bog them down with linear language (lots of text, tables, etc), charts allow people to look and understand outside of language. I can look at dashboards in other languages and get a rough idea of what the charts are telling me (some metric over time is going up) even if I have no clue what the text says. Add in more support for visual understanding (such as icons or pictures) and I may even get a rough idea of topic.

Understand the Goal

Crom’s goal isn’t to emphasize the beauty of the English language, the rhyme, or the beat. If he were teaching an English class, it might be which would change what he does. Instead he wants you to understand the point, which is made in a very abstract way. He, too, wants to keep this abstraction. Translating abstraction is hard, because we have to really understand it ourselves.

With Tableau dashboards, we need to understand our business users. What are they doing with this information? How long are they really going to spend looking at this? And, once they have what they need from this, what questions do they ask next? (In interpreting, these questions would be to find context, setting, and intent.)

For example, this dashboard attempts to find loss leaders and identify how much lost profit is hidden under the iceberg, if you will.

It has an absurd amount of information, which we’ll delve into. This is a dashboard that can only be loved or appreciated in Tableau. Printing it out misses the point. Even the images probably annoy you, because without interactivity, its use is pretty limited.

The goal is to find what items are loss leaders. I did this dashboard to help a client see how we could take a concept they used previously and display it in Tableau. The goal had very little to do with sales and everything to do with identifying potential products that were not effective loss leaders. In many markets, some things are intentionally sold at a loss to get people in. The goal is balancing this.

Set the Tone

The first thing people always notice about this is that its not red and green. So often, when we deal with profit, the answer is it MUST look like it belongs in a Christmas store. I tossed out the tree with this and let it try a different scheme for a few reasons.

  1. To prove you can still highlight the bad (in any color you want really) without it being harsh. Too often, we want to hit home that RED=BAD and we become obsessed. Red is a strong color and contrasts exactly to green. If you consider berries on a bush, it stands out well, though that’s also a function of tone (the red is usually darker). When we do this on a dashboard, all the colors (usually combined with other colors) just add to sensory overload.
  2. Colorblind people get exhausted telling you they can’t see it.
  3. People who can see all the colors get exhausted with you using red to punish them. There is a point of over-saturation and too much red will eventually make people immune to it.
  4. Blue works well with the metaphor. Seriously, icebergs are white and blue, not red.

There’s also circles galore. This is in part to clarify what’s what and to start building towards more complexity. That, and while analysts hate them, business users tend to like circular things. Just look at the Nest – there’s no reason, other than gratuitous flair, for it to be circular.

This dashboard starts out high level (here’s the total, kids!) and starts exploring what that means by various breakouts. At a glance, we’re losing money. We can see an outsize amount is lost in furniture with a fairly close amount lost in tech, but I make a whole lot more over there. Each item I add is a facet off the one before. The interactivity supports this – it’s only actionable in a few places, rather than everywhere.

Effort = Reward

Often in data visualization circles, we debate the complexity of a chart and whether it’s worth it. This is where I diverge: just like with writing, it also needs to balance in the entire piece. You can have complex sentences, but you also need some simple ones. Same thing with charts.

If we look at interpreting, Crom makes an ASL user work more and more. He starts us off gently, giving us what we need to know first. Fire group is over here. All references to fire will be here. Ice is over here – hey, notice handshapes are more rigid! By the end, we’re getting less and less direction, because we need it less.

We do this with dashboards, too.

The circles set the tone on the colors and position. The profit iceberg is a fairly complex chart to read from an end user perspective. There’s positive and negative (that match the bubbles) and then how they end up (reference line). The last chart reuses a bullet chart, but remixes it with bubbles and uses a darker blue to indicate its different. Just like with writing or any other good story, we have a mix of complex and simple charts to support the intent.

The reward is that the end user doesn’t have to do the math. They get the positive and negative without having to sort out the net.

This is the other advantage of using a very limited color scheme. It helps people know where to focus. Good interpreters also do this. For most of the poem, Crom’s signs are in a small bubble. When he hits the key points (the 2 groups, the world) his signs get bigger. Wherever we have complexity, it’s good to balance with simplicity.

The end result of translating in this way is that our audience benefits. They get the information in a way that matters to them when we translate function, rather than form.

Happy artistic vizzing!