In July, the MIT Sloan Management Review (published by MIT’s business school) that companies’ progress regarding data science in finance gets hindered by conflicts between business operations executives and data science groups. CDOs are introducing enterprise data policies and standards, such that higher quality, well governed and secure data is available to generate more insightful analytics, allowing the business to make more informed decisions and improve its processes. Yet, the benefits and value of this work may be lost, because some data science insights and innovations are seen as disruptive to the stability of business operations. The article suggests that companies need an intermediary group to better integrate these varying interests.
Is such an added layer in a corporate structure the best way to address this problem? There is an alternative that can be implemented by corporate leadership, or through better education about the common data characteristics that matter to all parties, and how data science in finance produces value.
The central concept
The most important, most central and shared concept to grasp is that raising the quality of data has many business benefits and prevents many problems from happening. Analytics applications will function with greater accuracy if they are fed with good quality data, as Searchdatamanagement.com’s guide for data managers and business intelligence leaders has noted. Better data quality makes it possible to expand how much a firm can use business intelligence dashboards, as well as analytics tools. It also frees up data teams’ time when they need not clean up as much data. They can then help their colleagues better leverage the data they have. So the business side should be concerned about collecting the highest quality data possible to feed into data science teams.
Data science-influenced insights, improvements and innovations for the processes and products owned by the business can be adopted smoothly and with less disruption when data quality is already consistently high. This is because there are fewer step-change remedial efforts involved during the change or adoption process. On the foundation of a history of high data quality data, change can be achieved without multi-step reconciliation of new outputs with previously incorrect outputs, having to keep maintaining an old system alongside a new one, duplicate data entry, expansion of reference manuals for surrogate keys and workarounds across more systems that are having to adapt because of poor quality data.
As GoldenSource’s own Tom Stock pointed out in March, making improvements in financial data quality management “reduces errors and their consequences.” This means reducing operational risk and costs. Even simply getting the handling of critical data operations correct ensures that business processes will run efficiently, which should be valued by the business operations units of companies.
The value of good data quality, and the resulting ability for financial firms to produce more credible and accurate analytics, seems to be catching on among some companies in the past couple years, even without forming committees and focus groups to justify such efforts or put them into action. According to an April 2020 McKinsey research article (“Designing a data transformation that delivers value right from the start”), a U.S. bank projected over $400 million in savings from rationalization of its IT data assets, and another financial firm projected a 25 percent growth in its bottom line as a result of data-driven business initiatives.
The rise of data science in finance
Also in the U.S., firms are seeking out more advanced analytics tools and seeking out contributions from data scientists at a rate that outstrips the number of these professionals who are even available, according to a January 2021 research report by the Information Services Group (ISG) consultancy on data services providers’ experiences. ISG suggests that Covid-19 pandemic challenges fueled this recognition of the importance of data analysis for business growth.
These examples show that many financial firms and corporations have already recognized the value of data science and good data analytics, ensured the necessary communication and collaboration, and are reaping the benefits, without remedial restructuring of leadership. They know that getting data scientists on staff and ensuring consistently high data quality are necessities, even without the input of task forces and special committees. Business operations departments already seem to be grasping this—and even if they don’t, data and analytics leaders can have direct discussions with business stakeholders to understand what they expect and need. Data and analytics professionals, as the Gartner consultancy recommends, can themselves take action and lead their firms by setting data quality standards and implementing necessary changes.