Data is critically important to almost any business – it’s the lifeblood that makes an organisation function. But managing, leveraging and realising the commercial advantage of data is also hard.
If fully leveraging data is difficult enough, the bar is only getting higher – with AI changing the game. Traditionally, approaches to data have been through a systems-focused lens, ensuring that the systems a business operates on have the requisite data flowing through them. But now, the task has moved to ensuring that data is AI-ready. This means that it needs to be properly tagged so that AI applications know what the information is, where it came from and how it has been used before. This enables the AI to understand what the data signifies without having to interpret it for itself. Without this, the AI might guess incorrectly, leading to mistakes, unintended consequences and even hallucinations.
In the AI age, robust data governance has bever been more critical. Organisations need clear policies, rigorous standards and well-defined processes to ensure data quality. Equally important is the systematic management of data throughout its lifecycle, including consistent curation, secure storage and management so that data remains accurate, reliable and fit for advanced analytical and AI applications.
Building the data team
All of this means that having the right data team in place is imperative. However, it has also led to increased competition in the market for professionals with advanced data skills and experience. Clarity over the team and roles you are trying to build is therefore essential. So, what does a ‘data dream team’ look like?
In our experience, both from the recruitment side servicing the market and as a data practitioner building and running a team, you have to get the data team right, along with the necessary data processes and structures, before you can even think about pushing far into AI.
Several roles are critical. Firstly, data engineers are fundamental because they set the processes to collect, manage and store the data for the business to use; they lay the foundations. Data architects ensure that the data flows and connections between systems align to business needs, and can be properly scaled and supported. Then, skilled data scientists and data analysts use the data to draw out actionable insights including applying AI techniques and potentially starting the evolution towards machine learning and automation stages. BI (business intelligence) analysts also play an important role by bringing a business/sectoral lens to what the data is showing. As a business becomes more mature with both data and AI, the need develops for AI/ML engineers to design intelligent systems using the data flows created.
Another key emerging role is what one might describe as a ‘data translator’ or perhaps a ‘data solutions engineer’. These individuals form the link between the data team and the business, acting as a conduit to help translate the insights from the data into business actions that can be taken. This demands both technical skills and knowledge, and business acumen and understanding. It’s a role that often gets overlooked, although more businesses seem to be realising that it’s a vital part of the puzzle.
It is worth noting that these roles are all in high demand, and can be hard to fill. As a result, data-related salaries have jumped significantly in the last 18 months or so. Whereas many tech role salaries have only risen at or around the rate of inflation, some data roles have put on perhaps 15-20%. A good analyst may command somewhere in the region of £70-90k, engineers and scientists perhaps £80-110k, while an accomplished data translator/solutions engineer could attract £120k or more.
As can be seen from the above, a good data approach is about having a multi-discipline team comprised of different roles, working closely together. Therefore, it’s by no means simply about finding people with the right technical skills – cultural fit within the team and the business should also be key considerations. As so often the case, it’s as much about the people as it is about the technology. Businesses shouldn’t expect to assemble the right data team overnight. It’s an organic and incremental process that can take perhaps 12-18 months to fully reach fruition.
The leadership question
Then there is the issue of leadership: who should take executive responsibility for data and the data team? Most businesses have a Chief Data Officer or equivalent (Director of Data or Head of Data etc) – the key question is where this individual sits in the management hierarchy. In an ideal world, the head of data would be on a par with the heads of technology and product, and have a seat in the boardroom. In practice, this is often still not the case – but is something that we expect will change in the coming years, especially with the continuing march of AI. There are other variations. For example, in companies undertaking large-scale implementations of AI, we sometimes see a Chief Data Officer working alongside a Chief AI Officer, but in some businesses the two roles are merged into one. There are no ‘right’ answers here – it really is down to the individual dynamics within an organisation.
Finally, it is also crucial to realise that data is not only a matter for the data team: everyone is a data user. Therefore, data literacy across the business must be on the agenda, with training and resources available to help everyone increase their data competency and confidence. Only in this way will you fully realise the benefits of all the work you have put in to make high quality, granular and relevant data flow around the organisation, feeding business decision-making and unlocking commercial returns.
Jack Capel is UK south director of Harvey Nash. Adam Asprey is director of Data at Maximus UK.
