• Home
  • /
  • Enterprise
  • /
  • The right people key to realising data analytics potential

The right people key to realising data analytics potential

Everybody is waking up to the power of data, but there’s less understanding of how to create a data analytics team that delivers results. Spoiler: it might not be necessary.
Tanya Long
By Tanya Long, Chief operating officer, Argility Technology Group.
Johannesburg, 05 Jul 2022

Data has emerged as the new gold in the digital world, offering businesses and even governments the opportunity to understand clients/citizens and markets better, and to develop new products and services that will drive profitability. However, it seems too many organisations see technology as the key to the world of data.

As always, technology is important but it’s by no means the complete solution. When it comes to data, it’s the right human skills that make all the difference − human judgement and acumen simply cannot be substituted. So while companies are prepared to spend large sums on acquiring data analytics software, they should actually begin by focusing on how to build the right data analytics team.

The first thing to realise is that these skills are in short supply. Data science has become one of the hottest jobs globally − the third best job in America in 2022, according to Glassdoor.

Locally, there is also high demand for data scientists. The opening of large new data centres by companies such as Amazon and Microsoft have hoovered up those with data skills, as have international projects such as the Square Kilometre Array and MeerKAT.

Another big factor in the local skills shortage is that data scientists can work remotely with great success; local skills are finding work in global markets.

The outsourcing option

Bearing the skills shortage in mind, companies should first conduct a rigorous business case for creating a data analytics capability in-house. The data analytics outsourcing market looks set to grow fast − Grand View Research predicts a 22.8-plus% compound annual growth rate between 2018 and 2025, driven by a combination of the importance of data analytics and the difficulties of creating and, crucially, maintaining an effective data analytics team. This trend is growing in the wake of COVID-19 and the growing acceptability of remote working.

The opening of large new data centres by companies such as Amazon and Microsoft have hoovered up those with data skills.

Here are some factors to consider when looking at the desirability of forming an in-house data analytics team, or opting to outsource it:

Multiple skills needed. Data analytics is complex and one skillset is not sufficient. An ideal team would include a data engineer to oversee the extraction of data, a data scientist who uses hard mathematics to write algorithms to derive insights and predictions from the data, and an analytics engineer, who translates data science into business analytics, thus linking the data scientists and data engineers. Other key roles include a visualisation expert to make the data easily usable by the business, a business analyst who interprets data to provide reports, a business product owner to link the business and analytics teams, and a data scrum master to ensure effective development practice.

Overhead increases. The previous point should clearly show that managing a complex team like that to ensure it is effective and delivers a return on investment is going to be a significant burden. In tandem, costs, including HR and equipment costs, will also be under upward pressure.

Requirement for new domain knowledge grows rapidly. Technology and new data types are constantly emerging, business models and markets are changing more rapidly, and so new tech and business skills will need to be developed and then retained.

In today’s markets, speed is of the essence, so individual growth needs to be continuously developed and accessed rapidly. Businesses of today are increasingly pressured to ensure their team is constantly doing interesting work, growing their skills and engaged.

Being one of the most sought-after skillsets, retention of solid machine learning and data science skills should be a business priority. The same point can be made with respect to the kind of reports and problem-solving needed to support decision-making processes − the reporting and analysis universe within the organisation needs to be carefully curated to ensure it is fit for purpose.

In summary, outsourcing effectively passes the burden of setting up, managing and maintaining an effective data analytics function to a specialist partner, leaving the client free to focus on what the data is saying and then devising strategies accordingly.

Going the in-house route

It is often suggested that outsourcing data analytics is indicated for companies that have great volumes of data, whereas companies with smaller amounts of data might be more suited to developing their own data analytics teams.

There is no hard and fast rule; everything will depend on the company’s strategy, and particularly its ability to manage and sustain the complexities surrounding its data and required capabilities.

Whatever avenue is chosen, a key first step is to develop and support an organisational data culture. The data team, insourced or outsourced, will only reach its full potential if the organisation as a whole understands what data it has and what it can do.

Organisations must also deal with the perception that analytics will be used primarily to automate business processes and so eliminate jobs whereas, in fact, the real value is increased ability to develop new products and services that drive growth. This needs to be made abundantly clear.

As noted above, a good data team requires multiple skills. While the specific recipe may vary, a crucial role that must be filled is the interface between the data specialists and the business.

Expense, both initial and ongoing, is always an issue. Efforts to build an in-house data team should begin small with specific business cases and proceed in an iterative fashion to build momentum and acceptance.

A related point is that no return on investment will be obtained if the basics are neglected − as with all data projects, the problem is not the amount of data but its utility. The first step must always be to consolidate all the data in one place and ensure it is clean and properly modelled.

Data and the ability to use it to generate insights is central to remaining competitive. Is your organisation ready to take on the task of developing its own specialised data team, or does it make more sense to outsource this function to specialists?