Bridging the Gap Between Data Science and Business Functions


July 6th, 2017

Once a novelty, data science has now become an integral part of any corporation that is serious about maintaining its competitive edge. As companies strive to incorporate new data management technologies and IT fluencies into their business development processes, they must ensure that:

a. They have the requisite expertise and skillsets at their end, to successfully utilise these technologies. This includes the ability to operate big data platforms and to run various analytics techniques like Machine Learning, using tools like R and Python.

b. They have a way to bridge the growing communication gap between their highly-specialised data science and engineering teams on the one hand, and the line functions that rely on them for data-driven insights.

While much has been written about point ‘a’, it is point ‘b’ that that is presenting companies with a whole host of unforeseen challenges. Data analytics is only useful if it can generate operationalizable insights, insights that can be articulated in terms of the problems at hand. Data scientists are often ill-equipped to translate the highly technical results of their work into the actionable language of reports.

What companies need is individuals with hybrid skillsets that combine familiarity with data science, domain-specific understanding of big data platforms, key analytical tools and techniques, and an ability to express highly technical insights in terms that are easy to understand and implement. One observer has described this hybrid, cross-functional skillset as Big Data Business Analytics.

 

How to avoid getting lost in translation

 

Company operations have historically been distributed across distinct functional units, such as marketing, R & D, finances and customer services. Each unit has its own management structure, goals and priorities. Any data that is generated pertains to these objectives and is domain-specific, resulting in the build-up of ‘data silos’. Data silos “are stand-alone containers of data. They exist separately; without sharing, cross-referencing, interpreting, adding to each other’s self-contained data sets.” Any insights from such data are localised rather than holistic or enterprise-wide, meaning they have very limited utility.

 

Meaningful alternatives that address these limitations at both the structural and the individual levels, in a way that combines new technologies and employee skillsets, have emerged. To combat data silos, organisations are instituting Enterprise Data Hubs. Hubs are where enterprise-wide data from each of the line functions are concentrated, and made available for anyone to discover, index and analyse from anywhere in the organisation.

 

To operate such systems requires both data engineers and analysts who can understand, synchronize–and help facilitate communication between–an organisation’s IT, big data, and business imperatives. There are two roles that enable this synchronicity:

Enterprise architects: EAs design responsive, enterprise-wide data architecture that aligns business as well as IT strategies and processes. They “perform…large-scale program oversight, monitor technology life cycles and determine how individual technologies will evolve over time in regards to the company’s demands.” Their jobs require them to liaise with individuals across the company in order to set-up customised data storage and processing solutions. Thus, they must demonstrate “in-depth knowledge of technology, business, the ability to think in terms of processes, and the ability to interact with people operating at all levels of the organization.”

 

Enterprise architects create the enabling environment or framework within which ‘business translators’ or ‘big data business analysts’ operate and broker data-based insights.

 

Big data business analysts: Big data has fundamentally restructured the role of the business analyst. Today’s BDBAs can no longer base strategic decision-making on rudimentary market-research, hunches or perceived trends. They must rely on insights derived from data analytics to make concrete business gains. This entails communicating analytics-derived insights to a diverse non-technical audience, such as managers as well as clients, to facilitate their implementation.

 

BDBAs are responsible for producing assets like reports that convey meaningful, actionable information to executives and data discovery tools like executive dashboards. Executive dashboards are interfaces that display real-time, organisation-wide insights into a company’s KPIs. These interfaces are also interactive, allowing the user to search for/enter queries requesting specific information.

 

Whereas EAs have “both feet firmly planted on both sides of the business-IT divide”, the role of the BDBA is “weighted towards the business [side].” Some experts recommend a merger of these two roles, which could result in a job profile that’s fifty percent applied data analytics and engineering, and fifty percent data-driven business intelligence. Any individual capable of functioning in both registers, would make the ideal ‘business translator’.

 

Over the past five years, organisations have begun to experiment with an approach known as DevOps, which integrates IT and operations across multiple functions. Analytics are now being introduced into the mix, with an eye to optimising DevOps workflows just as DevOps is being applied to big data projects. These developments demonstrate how business processes drive, and are driven by, increasingly agile and collaborative technologies that depend on experts who are equipped with interdisciplinary skillsets like EA and BBDA.


To learn more about how to build a data-centric team, CLICK HERE.


Also Read 

Train Your Employees in Data Analytics Using E-Learning Platforms

Building Analytics Competency Within the Team – Common Pitfalls to Avoid


Share with:

FacebookTwitterGoogleLinkedIn


RECOMMENDED FOR YOU

Don't Forget To Sign Up For Exclusive Updates!

POST COMMENT