app, BI, data visualisation, UX

When I think about the current state of digital art – what with 3D brushes and canvasses, virtual and augmented reality, the advances in real time image processing and, my personal favourite, the incorporation of time to our sculptures, it’s amazing how far we’ve come in the last couple of decades since we first started painting big furry elephants on the walls of our caves.

On the flip side, when we look at the lack of progress in some of the tasks we perform in business visualisation, perhaps we could benefit from taking another really good look at what we currently create. Shifting our perspective, using technology, in order to evolve the paradigm.

To demonstrate this rather “out there” statement, I would like to offer up the humble chart, who’s distinguished history dates all the way back to the 14th century and yet still provides us with a simple and yet elegant method of representing comparative data. Over the years we’ve stacked them, colour coded them, switched their orientation, drawn them in 3D and done endless groupings and data sorts. All of which is wonderful and fun, but it’s not the shift I’m talking about.

What I would like for you to imagine is instead of just a flat two dimensional image representing data, that our simple little chart is evolved into a living breathing interactive vehicle travelling on a series of possible journeys that our customers can take in order to reach the conclusions they need.

What if we allowed those customers to play with and redefine the data inputs into our chart in order to help them intuit brand new possible outcomes? Effectively cause becomes effect and vice-versa.

Whaaaat? That one should have blown your brain a bit.

With the intervention of digital, everything now comes with the potential for interaction. Nothing is static, everything we touch, tap, look at or whisper to, the numbers we feed into our models and the outcomes we get back out, can all be modified and mutated. We now have the capability to look at our history as well as potential futures, all as part of the same journey, just extended with time and demonstrated with animation techniques.

But before we start floating off on hover boards, let’s get back to that real world example I mentioned.

Most of us in the IT world have come across the good old Forrester Wave bubble chart. Or perhaps you use Gartners’ magic quadrant? There are several out there. We point at a product on our chart for the clients and we explain to them how good it is, what the highs and lows are and why we think that it’s the technology of choice for a project.

The story starts as I was chatting with a digital agency CTO recently about their Web Content Management (WCM) offerings. They were experts at using several products on the market, they even went so far as to offer the Forrester Wave report as a download on their web site. As we chatted, I think I saw that chart properly, possibly for the first time ever. And as I really looked at the numbers, all of those little bubbles started to dance and wriggle around the chart.

Just to provide a little context, your standard bubble chart depicts products utilising up to 5 dimensions of data, but primarily 3. Using the X and Y coordinate  axes and then the radial size of the bubble for a third. A seldom used, but possible addition is to include a colour differential, but this can start to get messy with just a little too much information on display. It can work, if applied subtly, as in highlighting a selection of bubbles in the chart above. As for the final dimension, I’ll come back to that later.

Forrester produces dozens of these reports every quarter, for many different bands of software. Their classic bubble chart uses “Strategy” and “Current Offerings” for the X and Y axes and then “Market Presence” for the bubble radius – and if you’re one of the lucky products, you end up somewhere near the top right corner of the chart signifying your superiority over lesser offerings.

Based on their extensive research, Forrester uses a numeric score for each main category, which is itself made up of sub-categories. Each product is then allocated a score of 0-5 for each of these sub-categories. This all sounds pretty good so far but what Forrester have then done is to provide a predefined weighting value (F) for each sub-category score (S) and the sum of all weights and scores per category determines where the product sits on the specified axis. [e.g X = F0xS0 + F1xS1 + F2xS2 + F3xS3]

The diagram above shows our trusty bubble chart and the three set of weights for each of the three data dimensions, or categories.

For our WCM chart, The “Current Offering” category is made up of: ‘Content @ 30%’, ‘Operations @ 25%’, ‘Architecture @ 30%’ and ‘Extensions @ 15%’

That’s fair enough, you might think, it’s their bubble chart, after all. But here’s where we leap off the diving board and head for the deep end. What if we didn’t just accept it?

The method Forrester use is called a weighted sum model, best known from multi-criteria decision analysis (MCDA). But what if some of those sub-category areas aren’t very important to us, whilst others are critical? What if I am very interested in content, but not the architecture? Development but not the cloud? Depending upon my organisational role, I am going to be interested in radically different aspects of a product assessment. As a content controller, as a marketing exec or as a development manager there are powerful reasons for me to want to have different sets of weightings.

Another way of looking at our chart is that Forrester provides an opinionated view of the supplied data. The reason they are doing this is simple, time and space. People reading the reports have little time and so the space given to the visualisation is kept limited. But by weighting and then amalgamating the scores we take away much of their value.

Imagine going to a restaurant for dinner where the reviews suggest that their starters are so bad that you’d rather eat socks, their mains are OK and the deserts are like eating a slice of heaven. Taken as a whole, the averaged score means good cancels bad and we end up with a score indicating at best a mediocre restaurant. And if you did decide to try it you probably wouldn’t even stay for the only section of the meal worth waiting for, desert.

Back to our bubble chart, the first aspect we should consider are the category numbers on their own. Instead of losing their significance in the larger quadrant focused bubble chart, how about letting them have their own voice. And within each of those single category scores are N separate sub category charts. And what about providing detailed product comparisons between our favourites? Hopefully, you can see that I am building up a picture that is pushing our little bubble chart to expand in a whole series of new directions and dimensions.

In my madness I next created some fun MCDA tools that allowed me to play with these scores and weightings and it was astonishing how those bubbles danced on the chart, just like they had during that meeting weeks before. Up, down, left and right. Leaders became losers and vice versa. It was a singularly gratifying moment.

As a next step, I created methods to separate out only the horses I was interested in from the pack, I was then able to really compare these offerings at sub-category level, allowing me to reach some real conclusions.

The diagram above shows the main bubble and category charts with 4 selected products, a product details chart that animates the bars between products and a side by side comparison chart. Additionally, there are an extra 11 (4+5+2) sub-category charts available for the customer to compare products with by drilling down from the category pages.

I mentioned above that there was a possible extra dimension that we could show on a bubble chart, and I think its one that really makes my point for me, as well as being one which hardly ever gets used; Time.

The Forrester charts are updated every quarter. Now imagine how useful it would be to see how a product behaves or animates over time. Is the company investing time and effort into its improvement or is it in decline? Given historical data we could watch this evolution in action.

Then I looked over all I had made, and I saw that it was very good indeed.

To summarise: I haven’t shown anything radically different here, more of a corner of the eye shift in perception which turns the usual cause and effect on its head. Instead of data leading to visualisation, I have journeyed into an altered reality where the interactive visualisation allows us to change the data to help us reach new personalised results.

BI Dashboards cease to be the end result and instead become the interactive tools which we manipulate in order to reach those results. The consequences of this can however be substantial. Charting API libraries need to grow up fast, business analysis will need to move back in time to modelling best practice processes, all mixed with some serious UX work and interaction design.

To finish the exercise, I turned the toys in my sandbox into a real UX driven project, cleaned up the code and popped it into an app container for mobile and tablet. Which, if you know anything at all about app development, you will realise is a ridiculous understatement of the pain, blood, sweat and tears involved in the process.

Some key features of the App:

  • Adjust weightings tool
  • Category detail charts (3)
  • Comparison chart (2-5 products)
  • Detailed product charts
  • Interactive bubble / bar charts (sort, select, touch)
  • Make Notes (colour coded, saved locally)
  • Send emails
  • Store adjusted charts to device
  • Pre-defined persona weightings
  • Tabbed navigation
  • Page transitions (Up / Down / Left / Right / Alpha)
  • Animated interactions on all charts

Here is a quick video highlighting some of the app key features: