- What is your understanding of digital transformation?
Not all the projects labelled as digital transformation are truly transformative!
The “Digital Transformation” label is applied to a large range of IT projects today, but we believe that truly transformative projects have some common defining characteristics:
- The project should provide an outcome which enables the business to either become significantly more profitable, better positioned competitively or significantly change areas like customer or associate satisfaction
- The project should leverage technologies which were either not available or were too expensive until very recently. While this time span is typically set between two to three years, this can vary depending on how volatile the business environment in a particular industry is.
- Finally, the project should have unique IP or elements which will separate it from the competition for at least six months
For example, speed and the cost of the solution, are the differentiators for a project we are working on to provide inputs to farmers on the productivity of milch animals by tracking cows with low cost designed by UST IoT devices.
- How impactful is data-driven digitisation and how can enterprises and start-ups leverage the technology? How does it bring better decision-making, game-changing efficiencies, or a better customer experience with more personalisation?
Data is absolutely critical for decisions, but data possesses a number of critical characteristics which together define usefulness and usability.
- Data management processes – this is based on the confidence an organization has about the accuracy of available data (data quality) as well as the ability to track lineage as data is transformed and allow governance around who can see what information at what time and the mastering of data standardization.
- Timeliness and granularity – this is a varying metric, measuring the cost of acquisition and storage vs value in terms of impact on important decisions
- Data Enrichment and continuity – the usefulness of data is enhanced by the presence of additional information and qualifiers that give the data semantics and additional meaning, this helps derive much-needed insights that aid in making actionable decisions
- Compliance – data must comply with regulations such as (HIPAA, CCPA and GDPR) as well as national restrictions like the ability to store data in a geography
Meeting all these needs adds to the cost of running compute workloads for data-intensive applications. While it is important that all data is useful and valuable, a Digital Transformation exercise will look at all these features – including total cost – and consider the value of data use case in terms of business impact and then drive forward the projects which generate true differentiation and ROI.
A great example of a promising project which stalled because of costs is one that we were working on involving the digitization of production equipment in a factory. For one of our clients in the tire manufacturing industry, the goal was to improve productivity by collecting real-time reliability and optimization data from the equipment as well as at various points in the store. Unfortunately, many of the devices that were necessary to obtain this information do not have adequate sensors and even those that do lack the ability to communicate and do not provide standard protocols for sharing data. The cost required to retrofit this equipment presented a hurdle that was too high to overcome. Ultimately, a solution which finally worked was the use of existing video feeds which were used for security operations and the use of visual analytics to identify optimization opportunities and perform predictive maintenance.
Personalization is another area where the categories outlined above can impact effectiveness. While Personalization can be great when people ask for it (think Netflix or Amazon), in most cases, customers want to be anonymous, and personalization becomes a drawback that actually depresses brand value. Applications like Waze which leverage the fact that your phone moves with you to get accurate traffic averages are the exception – the use case is the arbiter of whether data in an enterprise can be used transformatively.
- What are the key trends in data-driven transformation?
- Increasingly decisions are taken based on data, not on gut feeling. Even the greenlighting of media projects is based on a number of data types ranging from target demographics to timing of releases and even the way that the story ends
- Using large volumes of data (like petabytes and billions of rows) has become commonplace with tools like Spark running on the cloud
- Data is being used for not just better customer satisfaction but also employee / associate productivity as well as vendor and partner productivity – e.g. early views into supply chain volatility provide suppliers with greater ability to adjust
- The number of times that data is reconciled and corrected across the transformation pipeline is significantly reduced through automation and techniques such as anomaly detection using machine learning
- The cost of acquiring, storing and using data is rising and the intelligent use of data has become critical – a number of clients who were collecting data first and worrying about what do with later have been forced to change to a diametrically opposite approach
- Do companies invest in upskilling and reskilling employees for managing transformation?
In IT, employees who can learn are key to the continuity of business knowledge and have the proven ability to enhance the culture of your organization. We heavily invest in training, reskilling and skill level enhancement. It is important to consider understanding the levels of transformation and breakthroughs, specifically as it relates to the industry. However, it is important to understand other barriers, or even the glass ceiling and how a specific transformation could help launch the company on an exponential upward curve.
- Why is security important in digital transformation and how can companies ensure robust data security?
Speed in all activities is fundamental to successful Digital Transformation. Whether it is the reinvention of a fundamental process like ‘order to cash’ or we are working to deliver faster time to market for new features, speed is the single best measure of digital transformation. This includes suppliers to backend processes like manufacturing to front-end experiences like enabling a customer group to accurately discuss a disease with the members of a healthcare insurance company. Edge compute will be an area where companies may be open for security exposure, where data is collected, aggregated, or passes through. Thus, security at the edge and along the data fabric must be maintained at the same level to ensure continued security when the data is in transit, when data is at rest or when it is at transient points.
Integral to this is the ability to connect across systems and enterprises in a safe and dependable manner. All aspects of safe governance of data are encapsulated in data governance. There are a number of tools to manage data governance that covers data in rest and in motion. These could be tokenization based or encryption but will also need robust role-based access management. Companies are increasingly formalizing processes for data governance and stewardship and the smart use of tools (don’t buy a tool without commitment to outcomes from the product company is an emerging golden rule), the last thing that is needed is more software becoming shelf-ware or “cloudware”
- What kind of cost is involved in data driven digital transformation, the cost difference over regular digital transformation. And what’s the overall return on investment?
There is no real difference in costs – almost every Digital Transformation project needs smart and effective use of data both inside an enterprise and outside an enterprise. The ROI is again based on context and our approach is to prioritize projects based on ROI, speed to value and impact on the market – some of the projects will need more data and some will be a little less data-intensive. However, we do not separate projects as data-led or not data intensive. It comes down to insights that are driven by the data, and how that helps with the transformation.
- Why digital transformation fails for organisations?
The most common reason is the lack of a cultural alignment on the problems being tackled and the extent of involvement required from constituents.
There has to be a visible commitment from the corner office and this has to extend to an extensive change management process that ranges from retraining and reskilling to establishing clear goals which measure success. Too many Digital Transformation efforts run on a ‘best effort’ basis and that is a recipe for disaster because a lack of data-driven insights will lead to failures. While there is an initial hypothesis of what the business should consider for transformation. Some of this is based on customer needs, opportunities, industry gaps and other criteria that companies identify when they embark on a transformation path. However, the changing landscape must be viewed closely with course corrections made at the right time, before too much money and effort is spent or it becomes too late.
Most of the time Digital Transformation efforts are required to run and grow a sustainable business and continuous Digital Transformation efforts have to be built into the DNA and culture of the organization.
- Which industries were the early adopters of data driven digitization and which ones are catching up now.
Telecoms and Banking were some of the earliest adopters of greater digitization while the public sector, healthcare and traditional retailers are among those who are catching up.