A Ready Reckoner For Business Leaders
Saurabh Gupta — Co-Founder, CODEX
Going back in time, the phrase “Data Strategy” first originated in IT companies a couple of decades ago in an effort to store and retrieve their data more efficiently. Their focus was predominantly on data storage. Gradually, as businesses began to realize the true potential of data generated in their organizations, the scope of data strategy expanded to improve all the ways in which data is acquired, stored, managed, shared, and used.
Today, data-driven decision making is not just used by IT companies, but by businesses large and small in almost all industries around the globe. This brings us to the need for developing a strong and robust data strategy, which complements the business strategy and aids in improving top and bottom lines.
Primary objective of having a data strategy
Data strategy is the foundation and guiding document for business leaders to advance their data and analytics efforts in their organizations. A good data strategy not only provides direction but also helps business and data leaders in maturing their data capabilities. It helps them to shift from making decision based on gut feeling and hindsight to taking decisions based on hard substantiating data.
Data strategy must be looked as a long-term plan and it must include provisions that detail the role of people, technologies, and processes to solve data-related issues and challenges.
How Big Data is Transforming Telecom Industry
Data is no more a waste of storage space for telecom companies. It has now transformed into the most valuable strategic asset. Telecom companies are no longer considering revenue from the traditional telecom operations as their primary revenue source. It has become more of a commodity now. It is the data — considered as a strategic and corporate asset — that is driving value for telecom companies.
A lot of data is being captured through call logs and location-based settings and is being utilized by telcos for generating new revenue streams. Here are a few major areas where telcos leverage on big data:
- New product & service development: anonymized data captured from users is used for new product development. In the process, data analytics are used to verify which products performing and which are not.
- Customer value management: Customer data is used to market value-added services like bolt-on roaming, add-on data, location-based services, etc.
- Operational performance and maintenance: Operational data captured from various instruments and equipment helps in boosting network optimization.
Monetization: Can generate revenue by reselling anonymized data to third parties like retail, real estate, etc.
A little more on the importance of data strategy
When it comes to data strategy “one size does not fit all.” Every organization has its own set of strengths, weakness, opportunities, and challenges. Data strategy must, therefore, be unique to each organization or business depending on their specific requirements.
Efforts on the data front happen in multiple ways to meet specific organizational or business unit needs. In some organizations, multiple data projects run simultaneously to address those needs. Without a cohesive and enterprise-level data strategy, a lot of these efforts end up in duplicating the same data, creating process overlaps, replicating the same work, and finally wasting precious time and organizational resources.
A well-crafted data strategy can avoid all those potholes and tie up loose ends. It can bring cohesion and standardization across the organization in terms of acquiring data from various sources and managing it in much more efficient manner. An effective data strategy must be:
- Relevant and related to organization
- Integrated with the overall organizational objectives
- Actionable in terms of execution,
- Flexible for continuous revisits and updates
A data strategy establishes common methods, practices, and processes to manage, manipulate and share data across the enterprise in a repeatable manner.
So, where to start…?
Now that a strong case for data strategy is established, it brings us to the question — how to go about it? It is always prudent to start with the data strategy statement, which will be the bedrock for all the data related goals and actions.
A data strategy statement broadly defines three things — the strategic objective, scope, and the advantage. The strategic objective provides business and data leaders the direction to focus their efforts towards data success. While the scope outlines the boundaries for those efforts, the advantage is that distinctively unique position which an organization achieves to deliver better value to its customers.
Reshaping Automotive Industry Through Big Data
It is a known fact that Hyundai is a pioneer in investing in Big Data right from 2012 and has reaped several benefits before any other automotive company did. Many have followed suit since then and are leveraging their data in multiple ways to drive value.
Automobiles these days are loaded with sensors, cameras, and GPS to capture various parameters of related to vehicle functioning and performance. As per McKinsey, a connected car generated 25GB of data every hour. In this connected world, this data proves to be of high value as it can unlock new avenues for automotive companies to create new revenue streams and boost vehicle performance.
Some of the major areas of using Big Data in the automotive industry include:
- Vehicle maintenance: Machine learning models that capture sensor data from the vehicles can help automotive companies to understand vehicle wear & tear; predict and avoid vehicle breakdowns; and develop timely vehicle service schedules.
- Enhanced security: Sensors and cameras on the vehicle are interlinked and the data thus generated helps the users with enhanced security features. Users can remotely switch off their vehicles if stolen or act if a threat is detected.
Improved performance: Automotive companies and OEMs can push timely over-the-air updates to vehicles to optimize/boost their performance and subsequently improve the driving experience. They can update existing functionalities or introduce new functionalities and features.
Elements of Data Strategy
Now that the data strategy statement is taken care, let us go ahead in understanding various elements that go into building the data strategy. Business leaders are well aware of building their business strategies. Though developing a data strategy is a little similar in the process, business leaders must work in collaboration with data leaders like the Chief Data Officer / Chief Analytics officer in drafting the data strategy.
The following elements of developing successful data strategy are modelled based on the work of Bernard Marr, world renowned data guru, who helps Fortune 500 companies in their digital transformation journeys.
Aligning data with business priorities
As Bernard Marr says, the first step in data strategy development is to consider the organization’s strategic priorities and key business questions. Once these are factored in, then it becomes easy to understand how data can help in addressing/delivering those priorities. In this way, data strategy must always be aligned with the overall business strategy.
Initial focus on short term wins
Quick wins in the short term can help in the overall implementation of data adoption projects. Data adoption in many organizations is faced with organizational friction and resistance to change from the employees. As these projects usually take longer time to implement, such opposition could easily ruin the entire exercise in the absence of some quick wins in the short term. The data strategy must have a built-in mechanism to achieve some quick wins (even if they are small), to drive confidence into the rank and file of the organization, which eventually motivates people to support the bigger cause.
Identify data requirements and data sources
Your data strategy must include defining the data that is required to achieve your organizational goals. A lot of thought from both business and data leaders must go into deciding what data is required. There should also be clarity on the type of data — structured, semi-structured or unstructured. The strategy must not only mention the internal sources of data, but also include provision to acquire data from outside the organization (third party sources).
Governing the roles and responsibilities for handling data
Effective data governance helps people to think data as an asset rather than a by-product of business transactions. It motivates people to follow the set policies while working with data. So, as the next step, the data strategy must include who are responsible for handling, accessing, and using data.
Data strategy must provide this clarity on the roles and responsibilities. The ‘roles’ part must address questions like:
- Who should acquire the data and from which source?
- Who has access to what data?
- Who is responsible for data cleansing, analysis, and extraction of insights?
- Who is responsible for safety and security of data storage?
- Who are data champions?
The ‘responsibilities’ part must address questions like:
- What are the responsibilities of the CDO, data scientists, data engineers, data champions and business managers?
- Who is responsible for data accuracy and completeness?
- How to ensure the ethical usage of data?
- How to minimize and eliminate data duplication?
- How can the right data be made available to the right people?
Build the tech infrastructure based on business requirements
Many companies usually look at purchasing/installing the latest equipment and technologies for digital transformation projects. This can prove costly to managements if the tech infrastructure is not aligned with the business priorities.
Business leaders must not make the mistake of going with the hype. While it is true that one must build a flexible and scalable data architecture that delivers results, tech infrastructure must be based on hardcore business priorities and realistic expectations.
People: shifting to data-driven culture
It is a well-documented fact that people can either make or break even the most expensive digital transformation projects. Yes, the ‘people factor’ must never be underestimated while developing the data strategy. When an organization is spending so much on acquiring new tools and technologies, they may as well train their people to use them effectively.
A lot has been written about upskilling / training people in using data technologies. Making them data literate can reduce a lot of friction that naturally builds in organizations during the implementation of such large, long terms projects. Data literacy can go a long way in successfully shifting to data-driven culture.
Hence, a good data strategy must include the necessary roadmap or plan of action to make people accept the new norm of data-driven decision making without much friction and shift smoothly towards a data-first culture.
Pharmaceutical Industry Relies on Big Data and Analytics to Improve Drug Discovery and Clinical Trial Processes
Global pharmaceutical industry is currently undergoing rapid changes as traditional manual processes are giving way to smart AI-based processes that help pharma companies in reducing costs and improving revenues.
Big data analytics has now become the prime area of investment for pharma companies for its capabilities to identify patterns and enhance treatment efficacy.
Drug discovery: Traditionally, drug discovery process has been manual, iterative, tedious, and gulping down financial resources and increasing costs. The latest mathematical and computational models powered by AI and machine learning technologies has brought down the lead time in discovering lead molecules.
Clinical trials: AI is helping pharma and clinical trial companies tremendously in identifying ideal candidates for the trial and thus improving patient recruitment. Big data also helps in optimizing trial design by applying machine learning algorithms to clinical trial workflows and enable continuous operational improvement. It also offers deeper insights into patients recruited for trials, thus helping in trial output optimization.
