Data analytics – what NEDs need to know

In the June event of our NED Network 22/23 series, we looked at data – specifically what NEDs should consider and be aware of. Sarah Belsham, partner, Shakie Kawuyu, director and Kaitlyn Reth, manager, all from RSM’s Digital and Data Advisory team, highlighted the strategic significance of leveraging data effectively. We’ve covered some of the key points below, you can watch in full here.

The challenges

Skills

Despite the economic climate, investment in technology is increasing, , particularly in data analytics and automation. However, there is little value in collecting data and using advanced technologies , such as predictive analytics, machine learning, and artificial intelligence, if organisations do not have the skills to interpret and make the best use of it. As we previously explored in our April event, the current war for talent and increasing difficulty in finding the right people for the right role remains a challenge for businesses.

Picking the right data

Finding, selecting and using the right data to help you increase business intelligence and ultimately achieve your goals is not always straightforward. For NEDs, who may be serving on several boards, or who are not looking at an organisation’s entire data ecosystem, this is even more so the case.

Looking at the right data in a timely manner creates an additional challenge for NEDs. It can often take several weeks for data and board packs to be collated, particularly if there is a reliance on manual processes. This means that data is out of date by the time it is reviewed and any actions taken as a result of the findings could therefore have limited impact.

When we polled our audience and invited comments on this topic – they agreed, explaining that they experience a tough balancing act when they need to act quickly but the measures are not in place to gather the data. A key question for NEDs to be asking is how comfortable the board is with the quality and reliability of the data they have access to.

The solution

Automation is both the simple and complex answer. There are many measures boards can implement to gain accurate real time reporting. But they will need to be clear about exactly what they want to see. Dashboards are one way of doing this, but boards must take care to design these properly, focusing on which fields will help them make practical decisions.

How mature is your data?

There are four levels of data maturity.

• The descriptive (what happened?) – the analysis of historical data, for example sales data. This is normally presented in basic reports and tables in Excel spreadsheets or similar.
• The exploratory (why did it happen?) – using more advanced tools, like Alteryx, to discover trends and insights, including customer behaviour.
• The predictive (what will happen next?) – using historical data to forecast future outcomes. For example, forecasting cash flow can allow organisations to plan effectively. This would require advanced analytics tools such as Azure Machine Learning.
• The prescriptive (how can we achieve this?) – this analysis goes beyond the predictive to provide an actionable recommendation. Using algorithmic tools, such as Python, to offer online shoppers suggestions to add to their shopping baskets.

Managing data risk

As with most things, there is a level of risk when using technology and leveraging data. If we take automation and machine learning as an example, we must consider bias. The basis of machine learning is that it will ‘learn’ from the data it is modelled on. If the data is poor or biased, the results will reflect this.

So, how do we manage the risk that comes from using data? It’s worth thinking about three types of risk management.

• Digital risks – examples include third-party risks caused by introducing additional vendors, or data security breaches, as well as the resiliency of a business in the wake of a breach.
• Data security or GDPR risks – in considering everyone’s ‘right to be forgotten’ we must place emphasis both on data security and retention.
• Cyber risks – including ransomware, phishing or identity theft.

What questions should you, as a NED, be asking?

Our panellists shared the top five types of questions that any NED should be asking.

• Where has the data come from and how much has it been manipulated before it reaches a report that the board sees?
• Who owns the data, and have they approved its use in the way that it is being presented to you?
• What time period does the data cover and when was it last refreshed?
• What potential bias is present?
• Has the quality of the data been independently reviewed?

To learn more, watch the whole event here, or contact Emma Kennedy for further information.