Ecosystem Analytics: it's about context
The analytic lifecycle
Starting with Data
Data can be very annoying. There can be too much or too little, there can be gaps and inconsistencies, there can be anomalies and errors. Often it's all of them which makes the pursuit of data quality a constant challenge.
Working through the Visualisation
Why is data quality required? Well, if you put bad data into a visualisation, you get a bad visualisation. And, if you were planning on making any decisions based on that visualisation then your chances of being right are diminished.
Delivering the Narrative
But let's assume you have just the right data quality for your purposes. Now you need to present it in that perfect visualisation. A nice bar chart, a flowing sankey, or a complex radial dendrogram. It's finding the ideal visualisation to... ummmm... do something.
So, what is it you are wanting to do with the data and its nice new visualisation? Often, by the time you get to this point you've spent so much time wrangling the data and twiddling with the visualisation that you're just happy to publish the damn thing and move on.
Relying on the Audience
So you pass the visualisation to your audience and expect them to grasp the meaning and messaging of the visualisation and be totally inspired by your superb analytics. It probably doesn't happen.
Introducing "Ecosystem Analytics"
Take a different perspective. If we move away from thinking about independent visualisations each telling us something, and think about the ecosystem of whatever is being described, then visualisations have context. They relate to the other visualisations in a clear and meaningful way.
Think about a road map. It's a collection of independent visualisations (the roads, the lakes, the hills, the buildings) that come together using a common construct (scale, colours, shape, spatial positioning) to present something that is far more informative in aggregate than in its parts.
Ecosystem Analytics focuses on displaying appropriate data in a way that shows the parts coming together to form a whole. Changes in the underlying data then start to show as changes in the ecosystem. This can be an immensely powerful way of building understanding, identifying anomalies, providing feed-back loops, and enabling insights and inspirations.
Starting with Data
Data can be very annoying. There can be too much or too little, there can be gaps and inconsistencies, there can be anomalies and errors. Often it's all of them which makes the pursuit of data quality a constant challenge.
Working through the Visualisation
Why is data quality required? Well, if you put bad data into a visualisation, you get a bad visualisation. And, if you were planning on making any decisions based on that visualisation then your chances of being right are diminished.
Delivering the Narrative
But let's assume you have just the right data quality for your purposes. Now you need to present it in that perfect visualisation. A nice bar chart, a flowing sankey, or a complex radial dendrogram. It's finding the ideal visualisation to... ummmm... do something.
So, what is it you are wanting to do with the data and its nice new visualisation? Often, by the time you get to this point you've spent so much time wrangling the data and twiddling with the visualisation that you're just happy to publish the damn thing and move on.
Relying on the Audience
So you pass the visualisation to your audience and expect them to grasp the meaning and messaging of the visualisation and be totally inspired by your superb analytics. It probably doesn't happen.
Introducing "Ecosystem Analytics"
Take a different perspective. If we move away from thinking about independent visualisations each telling us something, and think about the ecosystem of whatever is being described, then visualisations have context. They relate to the other visualisations in a clear and meaningful way.
Think about a road map. It's a collection of independent visualisations (the roads, the lakes, the hills, the buildings) that come together using a common construct (scale, colours, shape, spatial positioning) to present something that is far more informative in aggregate than in its parts.
Ecosystem Analytics focuses on displaying appropriate data in a way that shows the parts coming together to form a whole. Changes in the underlying data then start to show as changes in the ecosystem. This can be an immensely powerful way of building understanding, identifying anomalies, providing feed-back loops, and enabling insights and inspirations.
Comments
Post a Comment