Critical insight is submerging in vast data ‘lakes’ or hidden from view in business siloes. Organizations must re-think their data architecture to create effective data supply chains that will aid their recovery, increase their resilience and make them stand out in the market.
Zhiwei Jiang, CEO of Insights and Data at Capgemini believes that organizations must re-think their data architecture to create effective information supply chains to aid their recovery in a post-COVID-19 era.
What has led us to the point of almost being drowned in data?
There are so many sources that capture data, quite a lot of willingness and even enthusiasm to capture it. Also, the varying means of technology (e.g. cheap cloud storage) available to store and access data – but this often means less control, less of an overview and not much of a strategy to actually ‘activate’ that data.
The impact of this ‘grey’ data on business is a significant gap between a business’s expectations of data being a core asset versus the inability to access it, navigate it, explore it, analyse it, share it with other businesses and – more than anything else – make it in the right, activated way available throughout the organization.
What kind of data ends up submerged?
Realtime data (also from IoT), external data sources, local business unit data, and unstructured data.
Do we know the hidden value of the data that lies submerged in these data lakes?
It’s already difficult enough to value the data that we do have available and visible, let alone what is hidden or difficult to access.
What role does the structure of an organisation – or the risk of siloes – have?
A data landscape should – just like applications – be an effective mirror (even ‘digital twin’ of the actual organization. It could also reflect the aspired organizational structure, as a way for the business to transform towards it. So data silos are not necessarily bad, if that properly reflects an effectively operating organization. But in any case, too much of a divide between the data architecture (e.g. very centralized) and the actual organization (e.g. very distributed) is bound to create unmanageable problems.
What impact does this have on business performance, agility and decision-making?
We depend so much on data that it needs to be activated as close as possible to the actual business operations (or literally right in the middle of it) – in order to improve decision-making and create data-powered performance improvement. The closer data is activated near the business, the more agility this typically will bring.
How should business better structure their data architecture?
Again, one must first find a proper balance between the current organizational dynamics, the aspired ‘to be‘ situation and the architecture of the data landscape. They need to be aligned, albeit likely as art of an ongoing transformation. In general, data should be activated closer – or within – the business and the data architecture should reflect and enable this ambition. Also, a lot of ‘grey’ unexplored data is not only at the ‘edges’ of business (e.g. in the field, or in devices), but is also outside the organization. Every modern data architecture hence should enable the open collaboration with other businesses – as in an agile, federative data supply chain.
Are any cultural changes required within the business?
The closer data is activated near the business, the more data-savvy businesspeople need to become: everybody needs to be a bit of a data scientist, a bit of a data engineer, a bit of a data storyteller. It’s furthermore fair to say that strong centralization of data and its management is not the typical way to transform right now, which also puts central IT / Data units on the spot in terms of the necessity to change its culture. Also, the ability to truly treat data as an asset – just like any other key corporate asset – needs to be built up, e.g. in terms of valuating data and its ownership. If only companies would treat data with the same passion (or scrutiny) as they treat finance, people or their physical assets, they would be well on their way not to drown in data.
