In the previous post in the Achieving Data Alpha series, we started to explore frequently asked questions about improving data operations in financial services.
Our goal is for financial services firms to make the most of their data assets and optimally enable their business users to make data-driven decisions. We covered five questions we commonly run into and looked at capabilities such as data lineage and data governance, how to reduce data cost, and improve data quality.
In this post, we’ll address the next handful of FAQs, including best practices om improving data operations, streamlining data management, optimizing data assets and the growing role of managed data services. In case of any questions, please do not hesitate to reach out directly to us at [email protected]
1. What does Data Alpha mean?
Data Alpha is a concept by which we mean achieving competitive advantage through superior data management. From our perspective, data management has been seen too much as a control function only for too long. Data management for control, i.e. knowing where your data came from, tracking where it is going, who can touch it and who can use it and having the capabilities to aggregate and quality-check different data sources, is critical. Still, it is also table stakes in the sense that these capabilities have become required for a license to operate.
Suppose you cannot assess the accuracy, correctness and timeliness of your data sets or access it, slice and dice it and cut them up as granular as you need for risk and control purposes. How can you do what matters: make the correct business calls based on that same data?
Data management for insight is what it is really about: how easily can you access your data? How sure are you about the quality of the data going into your risk models? How quickly can you onboard new data sets, integrate and cross-reference them to your current pricing and reference data to make them actionable within your business processes?
To us at Alveo, Data Alpha simply means increasing your data management capabilities to support and drive your business optimally.
2. How can managed data services help?
Managed Services help streamline data management, improve data quality and transparency, and reduce direct and indirect costs. Direct cost includes internal software and data license cost, infrastructure cost and staff cost. Indirect cost is typically the more significant and interesting part but harder to gauge. It includes more nebulous concepts such as future cost of change, operational risk, and the opportunity cost linked to being marooned on hard-to-change or out-of-date technology infrastructure. Ironically, the more complex these indirect cost improvements would be to assess upfront, the stronger the case for managed data services in the first place.
When a firm kicks off a process to investigate or conclude a managed data services contract, this forces the organization to think more formally about KPIs and SLAs they require to support the business. Firms should first be clear about the boundaries of the target service. What kind of change do you include? Do you have just hosting and IT services (run, monitoring) or change and test services? Do you have any business operations? Firms need to think carefully about metrics on uptime, delivery times and turnaround time for small changes deemed to be “business as usual.” Doing your homework will also clarify where different notions about instrument definitions and information to be covered can be harmonized and reflect fundamentally different business requirements. This process usually is enlightening in itself and a necessary precondition to conclude a value-added services contract.
Although a third party provides managed Data Services, it should always bring transparency on the end-to-end sourcing and offer its client a complete window into the processes. On top of the essential sourcing, integrating, quality-vetting and distribution of the data sets in scope and the maintenance services around that, managed service providers can bring extra value. Having outsourced day-to-day data operations and lowering the cost of change through a faster data onboarding process can help firms reach Data Alpha.
3. What does business user enablement actually mean?
“Business user enablement” is often used, and it seems it is something nobody would disagree with. But what does it mean in terms of capabilities required in a data management solution?
First of all, there is a need for user-friendliness. The User Experience of the interface to the data user needs has to match the standards that mobile apps have set since the first iPhone. These standards include well-considered screen layouts, user journey, and product design optimized to support their workflows.
Second, it means presenting overviews of the state of the operations through ad hoc reports that anyone can access, so your users don’t need to go to IT or your Business Intelligence team for every new report, every small change or new request.
Third, it means access to data for those users that require it. This access ranges from browsing for a large subset of users to supporting data operations workflows to quality-vet information, giving sign-off for downstream or external distribution, or deriving new data.
Fourth, it means access to data programmatically via state-of-the-art APIs to integrate it with internal analytical models and the rest of the application stack.
4. How do I streamline my data management?
Streamlining data management can be part of a managed services construct – in which the burden is given to the service provider. Alternatively, it can be done internally as a step taken before outsourcing a data operation.
To paraphrase John Donne: “No database is an island.” Any place of data storage will have many connections of data going in and out, keeping track of who uses what and who changes what is nontrivial. In real-life situations, often, there is an entire archipelago of data islands, meaning that data lineage and audit trails get lost. Storing data multiple times often complicates outsourcing arrangements. To overcome these challenges, it is essential to use service providers with suitable domain expertise in financial data products.
5. What is a data catalogue?
A data catalogue is an inventory of data sets an organization has. It typically also documents who can access the data and where it can be used, what quality checks has it passed, where it comes from, whether it carries a cost, and the related or similar data sets.
A data catalogue provides clarity to all users about what data is in house and where it is used and will prevent double sourcing or searching for information. Also, a catalogue can help harmonize data dictionaries and foster a common understanding of the maintained data fields’ meaning. It will flesh out discrepancies or gaps in the coverage.
Typically, a data catalogue sits one level above a data management system, although the two can also be combined. In that case, the solution provides an overview of the data sets available, their sourcing and preparation status and distribution connections.