Part 1 :: Setting the Scene
Given rapidly evolving markets, technology, and regulatory concerns, how should companies be storing, analyzing, evaluating, and monetizing their data? These questions are of great consequence to leaders in tech and data, particularly as the space deepens in importance and complexity.
Data is essential to how firms innovate and deliver for clients, and it’s key to stay attuned to how changes in the landscape are affecting broader goals. With that in mind, these are trends and developments CDOs and CTOs should be aware of, based on recent conversations with Bloomberg senior leaders focused in tech and data.
Definitions
The terminology surrounding technology and data science, including data lakes, data warehouses, golden copies, etc., is often conflated, confused, or used interchangeably. When considering these terms, it’s important to keep the broader applications behind data management in mind.
“This is a very fluid space, especially when speaking about data lakes versus data warehouses. Clients are storing data in different areas within their firms: transactions, fixed income, and equities, among others. They’re trying to get all of it into a single place because there are other things they want to do with this data,” said Naz Quadri, Global Head of Enterprise Data Science.
Since there are more stakeholders than ever influencing decisions around tech and data, it’s also important to note who is using what terms and why. CDOs and CTOs come to their roles from a variety of backgrounds, and someone coming from the world of compliance will have different concerns than someone in the front office. How this information is monetized is still very much in flux, and it’s critical to understand individual motivations.
Underlying factors
As the tech and data landscape changes, so do clients’ concerns. Being more targeted in how data is referenced and analyzed is key in anticipating future challenges and opportunities, such as leveraging alternative data technologies.
“We need to ask ourselves: what is being done differently now, as opposed to what was done in the past? Clients are seeking out predictive analytics, either to generate alpha or manage risk,” said Gerard Francis, Head of Enterprise Data. “To predict the future, you need a lot of the past, and that’s why you need data history.”
In capitalizing on data history, the rise of predictive analytics and how they’re used to generate alpha will be crucial in offering new options to clients.
“Clients need consistent data in all areas, and we think predictive analytics will come to matter greatly in the future. For the core data that people use, they’ll still want one vendor, but they’ll always be looking for something else to give them alpha,” explained Francis. “Which is why people seek out alternative data. It’s important to segment core financial operations and the supplementary data that will give additional value.” Human element New developments in how data is consumed and analyzed necessitate a rethinking of how teams are built. Automation is making it possible to streamline previously manual and time-consuming tasks, freeing data scientists and analysts to put more of their efforts towards revenue-generating decisions. Does the hiring process need to change as a result?
“Clients are at different points along the maturity curve when it comes to how they’d like to operate in the future,” said Quadri. “Do I need to build teams? Are the teams that I have now redeployable? Do I need to hire from different pools than I was previously?” These advancements raise larger questions that speak to how data can be used, and human resources reallocated, to anticipate future needs. “How do they use data to come up with the answers to questions they’re not asking?,” continued Quadri.
Data quality & storage Bloomberg has historically focused not just on measuring data quality, but also how best to store and deliver that data to clients. “Both our historical and ongoing data are in the same format and structure, so clients don’t have to program it twice, which would be an added cost,” explained Matthew Rawlings, CDO and General Manager of Data License. “We don’t tend to change the structure of how we deliver the data over time, and it works. So, for someone using it, it feels very consistent. That’s less work for clients to do.”
While incomplete and fragmented data has incentivized firms to use multiple sources in the past, the quality and completeness of the best-in-class data have rendered this approach unnecessarily scattered and ultimately unwarranted. Bloomberg’s response to ongoing challenges, the One Data One Source solution, ensures storage is streamlined and quality is the best available. “Our most powerful tool is our data history,” said Francis. “That’s our value added.”