Oct 25, 2023

We’ve recently been conducting research and analyzing data for an upcoming paper on DataOps that will focus on how investment management firms support data management in a world of ever-increasing volumes of data, sizes of datasets, speeds at which data consumers need data onboarded and made accessible, and types of data. This includes structured, semi-structured, and unstructured data in any number of formats for an increasing number of asset types, data consumers, and use cases. Put simply, the paper will examine how firms can continuously better support the data consumer ─ while keeping up with business as usual, of course.

In our research, we differentiate between data operations and DataOps. One distinguishing feature is that data operations tends to have a smaller, more limited group that they would consider data consumers. DataOps teams are more likely to consider everyone a data consumer within an organization, and their job is to support all data users.

In practice, this looks like DataOps teams building out self-service analytics platforms, rolling out data catalogs and business data glossaries, and generally making quality data available to users when they want it and in the form they want it. DataOps works with Data Governance to make sure the structures, policies, and roles are in place to ensure that happens. And while the data governance roles of data owner and data steward are well understood, it’s easy to falsely assume that a data consumer is simply a passive recipient of data. Data consumers also have a critical role, responsibility, and accountability in a data governance framework.

Cutter’s June 2023 Data Management Benchmarking Survey found that while over half of firms generally recognize the data consumer’s role, the level of participation in the data management process varies. Additionally, the survey revealed that only in a small percentage of large firms (13%) had data consumer roles with enterprise-wide participation and that are defined in policy or by clear practices.


Expectations

Nearly everyone at an investment management firm consumes data from and uses it for their own purposes. They may also merge and transform data from multiple sources to create data accessed by others. Data consumers are beholden to the teams that create the data. And, in turn, other teams are beholden to them for the data that they create. For example, performance analysts rely on accounting data for accurate positions, transactions, and data management teams for security reference data to create performance data used by portfolio managers, client reporting teams, and executives. Investment management firms also have external entities that are data consumers ─ regulators, clients, vendors, etc. ─ but we’re concentrating on internal data consumers here.

A data governance framework should include a service-level agreement (SLA) that clearly articulates what data consumers can expect in terms of content, timing, availability of support, data quality, and access. Firms also should establish a robust change management process for both data and any systems that impact the data. If data consumers report an issue, they are entitled to a response that explains the root cause of the issue and how it was rectified. Confidence in data quickly erodes if consumers report issues and never hear back about the root cause and how it was fixed.

Responsibilities

Data consumers may believe that they are simply on the receiving end of the relationship in which they expect good data in a well-functioning data governance framework. But they play a critical role in ensuring high-quality data and some of their formal responsibilities include:

  • Only use data from authorized sources. While this may seem like common sense more than a responsibility, it’s important to reiterate. This is because as data is accessed, transformed, and used in analytics, all downstream uses become suspect once the source is not authorized.
  • Identify data needs not addressed by existing authorized sources. This goes hand in hand with the first bullet. Consumers need to request data they need if it is not available, as opposed to potentially using data that looks like it might address their need without confirming whether they have permission to use that data or if that data even fits the use they believe it will.
  • Report data issues. Data issues cannot be researched and rectified if the data owner does not know about them.
  • Effectively communicate data requirements and SLAs. With clearly stated requirements and SLAs that are well understood, data teams can deliver the request. The result is that there’s less room for misunderstanding and poor delivery on objectives. Consumers should define the intended use for the data, data quality thresholds, and potential data providers/sources.
  • Proactively communicate business changes. Data consumers or business teams need to proactively communicate changes in their business that will result in new data or system requirements with enough time for data teams to plan for and address those needs.
  • Actively participate in ensuring data quality. Participation and input may entail testing requested enhancements or newly onboarded data, defining business terms, participating in training to ensure proper use of and location of authoritative sources, or taking a part in a data working group.
  • Advocate for funding and resources that support new initiatives or issue resolution. While a data team can request funding for projects, the projects are much more likely to be financed if the business team can explain the business need and objective of the project rather than the request coming from the data team.

It’s very much a circular relationship ─ data consumers play a key role in ensuring that data management teams have the resources they need to provide sufficient support and good data to data consumers. It’s important that data consumers feel enabled and recognized for their contributions, and that their issues and concerns get addressed. There is no faster way to disempower data consumers and break down a functioning relationship than when their issues go unaddressed.

If your firm needs help to better support its data consumers, look for Cutter’s upcoming research on DataOps in which we’ll explore how firms currently support data operations and how implementing DataOps could help.