Few people in business these days would argue with the value of data. Data means intelligence, data means informed decision making, data means the ability to identify issues early and measure success in incremental steps.
But data is also changing. The volumes of data being produced by digitised businesses are growing at an almost unimaginable rate. Reliance on data is spreading to every corner of the organisation, every operation. And there is growing competitive pressure to work data harder, to glean insights from it in close to real time and maximise value from what it tells you.
So how will this impact on the data skills that really make a difference in your organisation? Here are three examples likely to play an important role in the data-led businesses of the future.
Data science and analytics have become highly prized specialisms. The tasks involved in taking increasingly massive data sets and gleaning meaningful insights from them that make a real difference to business performance are far from easy. A consequence of this is that the ability to generate real value from data tends to get siloed away with a small number of appropriately trained experts within an organisation.
This is inefficient on a number of levels. For one, the growing demand for data scientists has put a severe strain on supply, leaving many organisations struggling to fill roles. It means that the real value to be gained from data – the insights and intelligence it offers – takes time to filter down to where it is put into action on the frontline of business operations. And as every department leans on data more and more, it increases the burden on small, specialised data teams beyond what is practical.
Looking ahead two or three years, it is clear that the knowledge and skills required to work adeptly with data will have to be democratised. If you make the comparison to reading and writing, no organisation could function having a single team digesting and passing on interpretation of all written documents, or drafting all emails on behalf of the entire workforce. A data-led organisation needs all of its people to be comfortable and adept using data.
Gartner expects 80% of enterprises to have implemented data competency programmes by the end of 2020. Soon it will no longer be enough to say you have data analytics capabilities and expect that to deliver value. It will be about how fast and agile you are in making data work in the right ways, and that will require having data literacy – the ability to interrogate, interpret and act on data-led insights in the appropriate way – embedded at every level of the organisation.
From a recruitment perspective, this will trigger a shift from looking for Chief Data Officers who can run the technical side of data management and analysis, to seeking out skilled leaders and communicators capable of developing and running data skills programmes.
We’ve already touched on the fact that the growing reliance on data, not to mention the exponential growth in data sets being used, is pushing the burden placed on data science specialists to the limits. The problem is that the ‘last mile’ stages of data analysis – the phase where most of the value gets added by turning information into action – remains a largely manual, human task. Computers can crunch the numbers, but we still rely on people to interpret what the numbers are telling us to do.
That is changing thanks to advances in AI. Amongst numerous crossovers between data science and AI, we’re now seeing software emerging that can reliably automate the all-important interpretive aspects of data analytics, effectively generating ready-to-run business intelligence reports from raw data sets.
Along with AI-automated data management processes, such ‘augmented analytics’ offers the speed, efficiency and accuracy required to see the data-driven economy evolve to its next phase of development. This will also create demand for new skills amongst data specialists. Professionals who also possess the software development skills to programme their own automated data pathways will be worth their weight in gold.
Data fabric architecture
As data systems inevitably become bigger, more complex and more embedded across every part of the organisation, another question presents itself – how do you ensure that you are consistently getting maximum value from your data at every point where it is being used?
We’ve already explained how and why data is outgrowing a single centralised management approach in the form of a specialist team who take care of it all. But in lieu of a centralised data team, what you can do is create a data fabric – a unified overlay of data services, pipelines, semantic tiers and APIs which can be used to orchestrate the management and use of data resources across a distributed environment.
The use of the term ‘fabric’ captures the idea that this structure must be tailor-made to wrap around the existing structures of any organisation. It therefore must be custom built from scratch for each business, requiring considerable expertise in systems architecture and design. For the enterprise seeking consistency as well as agility in its data services, these will also become highly desirable skills within data management roles.