Over recent years, data visualisation has certainly made a positive impact, which is no surprise given that it can help organisations achieve real insight from data. But beyond this, it’s also a smart way for us to understand complicated information in a more user friendly way that furthers cognitive understanding. To my mind, if organisations can implement and use data visualisation correctly, it can deliver immediate and actionable insight – and therein lies the advantage to make more informed decisions. But the caveat is that has to be implemented correctly.
However, as with all concepts there are always challenges. While undertaking some research on this topic, I came across a really interesting article entitled Rise of the Data Visualisation Competency Center, that asked the question: ‘what do organisations do with all the data visualisations floating around the organisation and how do we know that they are accurate and effectively representing data appropriately?’
This got me thinking. Especially with many data visualisation tools being sold and adapted directly by business users without central buy-in, approval or governance.
If the visualisations these users produce aren’t represented or interpreted correctly, it could completely alter the message contained therein and in turn lead the audience to the wrong conclusion or even worse, to make incorrect decisions. Especially with a powerful data visualisation tool, it is so easy to get the table calculations, summaries, labels or statistics wrong, if they are not carefully checked and validated. This can create huge challenges for an organisation – as they run the risk of losing out on potentially important opportunities, cause the business to go in completely the wrong direction or in worst case, lose the attention of the organisation’s stakeholders altogether.
So it then becomes essential for organisations to figure out a way to manage these risks associated with data visualisation – there needs to be careful governance of the outputs provided, and the messages conveyed, but without hampering the opportunities it can bring. Part of visualisation’s power lies in data exploration, where new insights are discovered through fairly unstructured trawling and investigating of the data through new powerful visualisations that expose the trends and gems of insight contained in that data.
So what resonated well with me in this article, is that the writer indicates that for data visualisation to be successful, organisations need to ensure that the right combination of design principles are followed to create a meaningful story. I have covered the importance of storytelling though visualisation in a previous blog. Of course, the story must not only be meaningful, it must also be relevant, correct and accurate.
Enter the DVCC…
The what…? In the cited article, the writer makes reference to the Data Visualisation Competency Centre (DVCC). This concept is fairly new to the industry and quite frankly surprised me a bit at first! The DVCC is essentially a cross-functional organisational team with defined tasks, roles, responsibilities and processes that support the whole “system” of data visualisation. It’s objective is to ensure that the visualisations produced actually deliver benefits to the organisation.
I do agree somewhat with the writer’s sentiments, as in some organisations it makes sense to group the data visualisers together because of their technical common ground, as this article suggests. However, what happened to the approach of having subject matter experts (SMEs) and data visualisers located in the various business functions, where they understand the business, the data, the culture and where they provide a direct “service” to their immediate colleagues? Much like the super-users or power-users we employ for reporting and dashboard development and deployment. But of course, the difference with reporting power-users is that they typically work off a semantic layer (a la BusinessObjects Universe or a Cognos Framework Manager) that ensures some of the governance around what data is used together and exactly what it represents to the business. The visualisation tools I have worked with don’t support that semantic layer concept that well, although of course, it can be implemented – even if in a bit more complex way – through views and permissions applied to the data warehouse schemas.
Simply put, I think it would be more beneficial for an organisation to have subject matter experts in their respective fields, such as strategy, marketing, operations, sales, IT and finance experts, who understand the data as well as the tools to make data visualisation successful and then filter their expertise and experience into the rest of the organisation. The various groups of visualisers throughout the organisation can then be pulled together through a special interest group, where they can be guided by a steering committee and they can run regular workshops to discuss and apply governance, as well as share approaches, intellectual property, techniques, success stories and so on.
In my opinion this approach would make data visualisation easier to implement while still keeping it governed to be relevant and understandable for the entire business.
