The data ownership challenge: managing the cost, driving the value

Optimizing the total cost of data ownership has three components that yield increases in return on investment for data management. Reduced costs, increased productivity, and proactive data vendor management. Organizations can achieve short-term benefits by reducing data management costs by ~10% and increasing productivity by ~40% for quantitative analysts, data scientists, and data integration into workflow tools and processes.

In a detailed study by McKinsey Technology in 2020, the short-term benefits of addressing the ‘Cost Levers’ in data management established a range of short-term savings possible for each ‘Lever.’

The research analyzed these ‘Cost Levers’ from a range of organizations, with financial markets enterprises having a much higher than average savings. Using this table, the short-term savings, within the initial 6 to 12 months, from adopting a new approach to data management would be between 9 to 13% of the cost for these levers. The research study also discovered that in addition to these levers, sophisticated data users could spend between 30 – 40% of their time searching for data if a precise inventory of data is unavailable and then devote 20% – 30% of their time to manipulating, normalizing and cleansing data to make it fit for purpose.

~10%

reduction in data management costs possible as a short-term benefit

~40%

increase in productivity for quantitative analysts, data scientists, and data integration into workflow tools and processes

30 – 40%

of their time that sophisticated data users could spend searching for data if a precise inventory of data is unavailable

20% – 30%

of their time then devoted to manipulating, normalizing and cleansing data to make it fit for purpose

Time spent locating, accessing, testing, and applying data for supporting workflows or updating and creating models constitutes data wrangling. An advanced data management platform supported by experienced data engineers eliminates a significant proportion of basic data wrangling.

Platform effectiveness & efficiency

Organizations must modernize their data management platforms and architecture to accommodate new data sources in various formats and input types. So, how does an organization achieve this without increasing the total cost of ownership of all this data?

The answer is using a specialized data management platform, socializing data enhancements and quality improvements, and systematically managing data vendors.

An effective data management platform becomes a repository for derived data sets supporting users’ discovery, access, and utilization.

The development and scheduling of data extracts directly into the workflow. Scheduled extracts in the correct field order and format might be for updating a trading model, meeting compliance requirements, or assessing risk. Enabling system generation of repetitive tasks will ensure consistent quality.

The current state is that data requirements are growing, and persisting with the existing operating model will drive up costs, not curtail them.

Data vendor management

Proactive management of third-party data vendors. Numerous case studies show that passive management of data vendors leads to duplicated data licensing, further compounded by repeatedly fixing core data issues by many users. These data fixes are typically not socialized across a common platform or fed back to data vendors to remediate at the source.

Reduce unnecessary third-party licensing spend by reviewing vendor contracts, identifying duplicate data licensing, and using a platform that manages permissions, sets quotas, and tracks each user’s consumption.

Building a better business is always the focus

Reducing the total cost of data ownership is not an end in itself. Any program reviewing data management costs and utility should focus on sustainable productivity gains by both technical and analytics resources.

Implementing a platform that addresses the critical issues of vendor management, data quality, and providing workflow-ready data will deliver a better outcome for the organization and its customers.

Reducing the total cost of data ownership is not an end in itself.

Organizations increasingly rely on data analytics and creative insights to establish and maintain a competitive advantage. The key challenge is managing the increased need for broader and deeper data while stabilizing the total cost of ownership.

This challenge has risen steadily in the past decade. Data management and advanced analytics are now a point of differentiation for organizations in data-intensive market sectors.

Compounding this challenge is a real talent shortage of the skills required for the disciplined management of data and the creative insights that drive revenue and margin growth.

Organizations will acquire more data from more sources, then integrate these new sources to convert data into information for critical decisions. All this must occur while ensuring the data is the best quality, discoverable and accessible. These conditions add value to operational and trading decisions, the basis for realizing a competitive advantage from data.

RoZetta Technology’s approach

RoZetta’s DataHex is a SaaS product that enhances value at each data lifecycle stage. The absolute focus is the journey from receiving the data from the source and delivering productivity and data enhancement gains at every step.

The RoZetta data management ethos is the concept of an Enterprise Centric data management platform.

An Enterprise Centric platform, as opposed to User Centric, optimizes productivity and speed to decisions by presenting the best possible data to every end user in a format that is consumption ready, whether for developing trading models, discovery analytics, or reporting.

Because remediation of all data enhancements and data quality issues are in the platform, the organization develops a unique data asset tuned to their specific needs, creating a competitive advantage through data.

The data flow in a User Centric Model is like a bowl of spaghetti, making it hard to manage compliance and licensing costs. The source data acquisition is undertaken by varying organizational levels and roles, often multiple times, with end users doing much of the data clean-up and integration. It is a very task-driven model.

An Enterprise Centric model is characterized by clear lines of accountability for acquisition, storage, normalization, completeness, quality, and access. DataHex has an intuitive interface and a suite of APIs to access data, and as a platform partner, RoZetta manages vendor issues.

The data flow in a User Centric Model is like a bowl of spaghetti, making it hard to manage compliance and licensing costs.

The alignment of responsibilities with expertise is a real productivity gain. Data Scientists and Quantitative Analysts get on with identifying and executing value creation free from repeatedly wrestling the input data into a ready-to-consume state.