Data transformation is an integral component of every digital transformation initiative. Unfortunately many fail to deliver for the business.
The most challenging aspect of a digital transformation is the scale of change. It often entails a new way of working with data for the entire organisation.
Having worked at organisations large and small, many of them iconic brands, we've observed some traits that differentiate a successful data transformation from one that fails to deliver. What the successful ones have in common are a cogent data strategy, an organisational embrace of agile ways of working and putting in place a team with the right skills and tools.
At its heart, a Data Strategy is the organisational approach to using data as a strategic asset. That means that any data initiative is considered in terms of the value to the business and alignment with strategic goals. This is coupled with cross-functional ways of working and architectural principles that marry enterprise consideration with the flexibility to be agile.
A data strategy acts as a true north and helps anchor the emotional biases towards technologies, processes and business use cases that inevitably arise during the execution phase.
Once a Data Strategy is in place, the data transformation piece becomes vastly easier. In fact it materialises as a by product of delivering data products which have a huge business value.
Agile ways of working
Agile is an overused term which can encompass a plethora of concepts - Scrum, Kanban, Lean, DevOps / CICD, TDD etc. Each has its associated practices and rituals.
What's often missed is the point of agile - to learn quickly. The meaningful metric of agility is the speed of those learnings. A truly agile culture uses those learnings to fail quickly and thereby minimise lost time and costs. Or it will find ways to continually optimise and improve business performance.
For a data transformation to be agile it must have the right:
- Framework (tools and processes)
- Ways of working (highly cross functional / collaborative)
Skills and tools
Putting in place a modern - typically cloud based - data platform will simultaneously test the enterprise along several dimensions. For mature organisations with long standing data functions this can be challenging.
The skills and concepts required for a cloud based data platform will differ significantly from traditional on-prem, proprietary technologies. Open source frameworks prevalent in data analytics and data science such as Python, R and Scala need to be mastered. And the concept of using cloud based services requires developing a different mindset to that of on-prem based resources.
Rather than see these as a threat, existing employees should be enabled through a mix of training and working alongside experienced consultants.
A data transformation requires far more than an architectural road map. For the transformation to successfully land, a congruent data strategy, embrace of an agile culture and the right skills need to be considered as well.