Get the message: how financial services can find data that matters

• Understand the importance of bringing insights to the forefront to minimise decision latency
• Gain a unique perspective by supplementing propriety data sources
• Financial firms continue to spend on technology, but budget is still needed to verify reliability and usability the content

Financial services professionals are overloaded with data. In today’s climate, how can they work out which data will help them make optimal decisions?

By 2025, 463 exabytes of data, each one of which is equal to one billion gigabytes, will be created every single day.

According to market estimates, as much as 90 per cent of that data will be unstructured, which means it is not stored in an organised manner. It can be anything – text, video, audio, images – and by nature it is more difficult to search, organise, analyse and act on, and it is growing faster than structured data.

In this ever-expanding universe of data, where quantity is overwhelming quality, how can financial services professionals reach the valuable insights they need to make targeted and robust decisions?

Decision latency – the time it takes someone to act in response to information – has a meaningful impact on business performance. Effective data analysis can both speed up and improve an organisation’s decision-making – from idea generation via risk management to compliance and regulation, settlement and client interaction.

To prevent the ongoing explosion of data volumes from sabotaging decision latency, a good place to start is data monitoring. This is where data insights teams syphon off irrelevant data and only put forward meaningful signals and usable datasets to the right people at the right time, giving decision makers the confidence to act with precision.

“For a period, we expanded our data sources as much as possible, but there was so much noise in our datasets it made it difficult to focus on what was relevant,” says Chris Ainscough, Portfolio Manager and Director of Asset Management at Charles Stanley. “So we decided to refine our focus down to the decision-relevant information.”

Creating a data gatekeeper function also helps to reduce confirmation bias, which is where we are naturally drawn to sources that confirm our preconceptions. More of the same data does not add value, so data teams should instead use alternative sources and insights that challenge existing assumptions and uncover unforeseen risks and opportunities – and lead to better decision-making.

By deploying techniques that bring the valuable, trustworthy data points to the surface, organisations can improve their employees’ data fluency. Other techniques in this category include the likes of data visualisation, dashboards and intuitive user interface design.

“The best way to empower employees to get the insights they need is to serve the insights to them within their existing rituals or processes,” says Jeremy Hunt, Global Head of Data, Analytics, AI and CRM at Schroders. “And doing so in a way that makes them easily consumable.”

Helping employees to reach the insights they need independently means they may not have to wait for support from data specialists – a crucial potential time saving that could reduce an organisation’s overall decision latency.

Increased automation makes a lot of data available to everyone at the same time, which creates ‘noise’ – excessive amounts of potentially worthless data. In response, some financial services organisations are searching for more exclusive data sources, which have specialised content under restricted access and are even building their own proprietary datasets to access insights that competitors are unlikely to be privy to.

“There have been scenarios where we’ve procured a large sample of data internally or externally and then analytics have revealed unexpected trends or inflection points to explore,” says Helen Zhu, Managing Director and Chief Investment Officer at NF Trinity. “To have a differentiated viewpoint, organisations need to collect proprietary data themselves.”

According to Zhu, it can take years to build representative samples and it needs significant investment in time and data storage. “But if you run a number of these projects, and even if only 10–20 per cent turn out to be useful, predictive and can help you to make good returns, that's still worthwhile,” says Zhu.

The caveat to all of this is: decisions can only be as good as the underlying data. Undoubtedly, the rise of data as an asset has created opportunities for financial services, but if the data has no quality control it can create errors in risk calculations, financial reporting and even regulatory compliance.

Data is only valuable if users know they can rely on it, which also depends on what the data is and what it is going to be used for. “We use the term ‘fit for purpose’ to describe a data source that is raw or not confirmed as 100 per cent accurate due to the need to produce data quickly,” says Karen Hiers, Chief Data and Analytics Officer – Enterprise at Northern Trust Corporation. “So for example, a newswire is fit for an analytical purpose – it's directionally correct. But we wouldn't want to put our money on it.”

Financial services will have to assess whether certain data types and sources are authoritative enough to be used for transactional and operational decisions in real time. Signposting the credibility of data points or sources through pop-up notifications on a platform or user interface design features can go some way towards reassuring decision makers.

As organisations spend big on the technology to process the vast range of data available, an equal budget needs to be spent on the safety checks to verify and grade the content on its reliability and usability, including investment in the staff that control the information flow. After all, there are no prizes for being the first to make the wrong decision.