Plutarch: Thoughts on People Management
Julian Elliott on data strategies, part 2: how to formulate a viable data strategy
About the author
Julian (Jules) Elliott is an experienced Chief Data / Analytics Officer with track record of successful data-driven transformation across multiple sectors. He has worked with Direct Line Group, British Gas, Lloyds Banking Group, L.E.K. Consulting and began his career with Schlumberger.
Most recently, Jules was Chief Data Officer at Dentsu Aegis Network.
Julian Elliot on data strategies, part 2:
How to formulate a viable Data Strategy
A serious data strategy goes far beyond buzz-words. Applying “AI”, harnessing “Machine Learning”, or analysing “Big Data” are techniques that might add value in context. But on their own, their mention often only serves to muddy the practical and sequential steps an executive team should take to address what is, above all, an issue of competitive strategy.
The two most crucial elements to consider when starting to formulate a data strategy are “vision” and “capabilities”. Both are fundamentally about looking beyond the current state-of-play in order to compete with known and unknown future competitors.
An executive team dramatically increases its chances of adopting a viable data strategy by concentrating on a compelling data vision, clearly linked to the corporate vision, mission and strategy.
This vision should then be elaborated by describing the key data-intensive capabilities needed to deliver it, and their corresponding business benefits, e.g. interactive BI for better decisions, real-time personalisation for all customers, AI enabled fraud detection etc.
“Good business leaders create a vision, articulate the vision,
passionately own the vision, and relentlessly drive it to completion.” - Jack Welch
A succinct and convincing vision
The "vision" of a data strategy sets out the scope and ambition necessary to support and enhance the corporate strategy. The path to get there is explained by the addition of new (different and better) organisational data capabilities in a roadmap – we explore this later.
Expanding the vision into a couple of sentences provides guidance and direction for the cultural journey the company will need to undertake. Getting the balance right between ambition and something the organisation can understand and buy into today – necessary for progress – requires debate and judgement.
The vision should seek to unlock value in three main areas for an enterprise:
- Using data and analytics to underpin and improve decision making
- Delivering on-going operational improvements / efficiencies
- Monetising data and new products and/or services derived from data
Emphasis should be determined by the existing corporate strategy and intensity/urgency of threats and opportunities. If it’s the first board-agreed or company-wide data strategy for the corporation, focusing on just one of the three areas will “de-risk” initial implementation.
Understanding your capability requirements
With the "vision" established, the next step is to unpack (e.g. at a one page of detail level) the most important data-intensive capabilities that will need to be built. Each identified capability should be fleshed out with additional detail, such as:
- the options for their development (e.g. MBA – Make / Buy / Ally), with timeframes and realistic budget ranges for their development and the benefits they will deliver
- the types of analytics needed to successfully deliver these capabilities
- the datasets or types of data required to fuel these analytics
- required resources and implications, if any, for the operating model
Picking up the pace
When refining the data strategy, it is necessary to identify which capabilities are dependant on others, and thus what sequence of development will be required for success. This will mean prioritisation calls. Ideally a long list of data capabilities will be distilled down to 3-5 priority capabilities that can deliver most of the data vision in an acceptable time.
These capabilities can also be developed and deployed at progressively higher states of sophistication, maturity and scale. It’s often helpful to think of each capability going through distinct stages (e.g. crawl, walk, run) as they are “ramped up” across the business.
A strong foundation
Capabilities that are shared or required by other capabilities should be grouped into a “foundation layer”. Adequate plans for strong foundations to underpin and support the data strategy are essential for sustaining success efficiently over a multi-year programme. These types of foundation capabilities will usually include: data procurement, data governance, data protection, licensing, privacy, big-data IT capability (dataflows and environments) etc and the overall cultural change.
In the data-driven digital economy, having strong data foundations can, and in many sectors will, make the difference between winning and losing. Simply put: without strong foundations, it is hard – if not impossible – to construct and operate efficiently a set of sophisticated data capabilities in the medium and longer term.
It is easy for executive leaders to see these foundations as “back-office” or “compliance” matters. Those who increase their engagement will be able to understand the practicalities of data strategy, to prioritise investment, and to predict when and where they will see returns on that outlay, and thus demystify data-strategy. Those who fail to understand their critical importance, however, will find themselves rapidly overtaken by the pace of change and left pondering why the commercial benefits materialise more slowly and patchily than forecast.
“Change is the law of life.
And those who look only to the past or present are certain to miss the future.” - John F. Kennedy
Change is the only constant
In today’s economy and technological reality, there is never a “final” data strategy: the benefits of its implementation will inevitably reveal business issues and opportunities that were previously out of view but now need a response. Executive teams should therefore iterate the data strategy as often as needed, at least annually, (re-)aligning with the technology roadmap and actively enriching corporate strategy on each occasion.
The data strategy should not just be for consumption by the exec and tech teams. It can be simplified for communication across the company and to other external stakeholders such as strategic partners, suppliers, investors and industry commentators. The internal narrative is important; having employees understand that quantity and quality of data are essential ingredients for future success is critical. Building awareness of the data strategy helps everyone to understand how the company operates and competes using its data assets.
An executive team that actively participates in development and communication of a data strategy will have far more confidence in their overall corporate and technology strategies. The organisation will be far better prepared for future success. Having built its data capabilities on strong foundations, it will be able to deliver strong financial results – and to grasp exciting new opportunities as they emerge, doing so far faster than competitors.
Plutarch in no way claims to offer comprehensive statistical reports – the absence of numbers reveals that much, and individual confidentially remains his priority. Nonetheless Hunter-Miller's vast network offers compelling anecdotal evidence, and some occasionally interesting insights.
What would you be interested to know? Get in touch.