A data strategy defines how an organisation will use data to achieve its objectives.
It also describes all the data related activities that need to take place for this to be possible.
Within some business sectors, e.g. Transport and Logistics, the case for having a data strategy is a strong one. They typically have a great deal of data, and potential uses for that data that range from the basic (e.g. finance, payroll, keeping track of vehicle maintenance and fuel economy) to the advanced (e.g. supporting investment decisions, predicting when a driver’s performance might lapse).
However, working with data can be time consuming and expensive. So, develop a data strategy only if a clear use-case presents itself.
That said, businesses that use data well are better placed to respond to the challenges ahead than those that do not.
A massive challenge facing many industries involves reducing the carbon footprint whilst keeping services running and remaining competitive – all under the shadow of advancing legislation.
G7 leaders have just announced their commitment to mandatory climate-related financial disclosures. This will do much to pave the way for climate risk metrics to be incorporated into corporate governance and strategy. Meaning that it will only be a matter of time before all businesses will have to report, and be held to account for, their carbon footprint.
When this happens the use-case for data will quickly become very clear, and very demanding – summarised as follows:
To demonstrate how real reductions in carbon footprint will be achieved whilst maintaining customer service quality and achieving profitability – all under the watchful eye of legislators with teeth, and shareholders who will want to see profitable operations continue.
The task required of data will be to present options for reducing the carbon footprint whilst operating sustainably, and to reveal which of the options works best.
Those options will include the big and the bold e.g. a transition to carbon neutral manufacturing and warehousing, as well as the incremental and the marginal e.g. finding the optimum mix of transport type, loading and route choice for minimum carbon emissions. Whilst also adhering to customer schedules and operating at the lowest possible cost.
The pivotal word is optimum – i.e. achieving the best result obtainable under specific conditions.
It involves finding relationships between different sets of data and using them in a calculation that maximises a particular measure of benefit.
Businesses able to do this will clearly have an advantage. Their decisions are likely to be the right ones because they will be underpinned by evidence and indicators of likely success. Remove the evidence and indicators and the risk of getting it wrong becomes greater.
A data strategy will therefore become essential as it will define the data needed for the search for the optimum to be possible, where it will come from, and how it will need to be processed.
It is not the only element within the strategy. Data is needed across all parts of the business. But it is going to become one of the most important.
The steps to create a data strategy are relatively simple to describe.
- Develop the questions that data needs to help answer. These are called Critical Use-Cases.
The need to demonstrate a steady improvement in carbon footprint will be one of those use-cases.
- Translate each use-case into a definition for the data needed – whether or not the data exists.
Even if data doesn’t appear to exist it might easily be found – or derived from data that is already within the business.
- Develop a data governance policy and define, for each set of data sought, how the terms of the policy will be met.
Key considerations should include steps to address weaknesses in data quality, how to achieve secure and reliable access for the right people, uses to which the data can and cannot be applied, adherence to governance principles such as the General Data Protection Regulation (GDPR).
- Identify the tools, infrastructure and steps involved in sourcing and processing the data needed.
This includes an outline description of any specialist algorithms required.
- List the skills needed at each point in the sourcing and processing of data and hire or train accordingly.
- Establish who is responsible for the successful completion for the steps outlined above.
- Define how success will be measured at each step in the sourcing and processing of data and develop a simple set of KPIs that will ensure that everyone stays on track.
It is very easy to throw resources at data in the hope that something useful will be achieved. A data strategy encourages exactly the opposite. It ensures that resources are applied to the task only when it is clear what they need to do and how success will be measured.
For the use-case related to reducing the carbon footprint whilst maintaining customer service and minimising costs, the approach will be all about finding an optimum combination of ‘variable settings’ within the business.
Taking Transport as the example once more, some variable settings will already be within reach e.g. the maximum loading of a vehicle or limits to driving time to ease driver fatigue. Others will be projections based upon possible investments to reduce the carbon footprint e.g. the likely range of electric vehicles that haven’t yet been acquired along with the likely availability of charging points that are not yet in place.
The task will be to decide upon the measures to minimise (e.g. cost and carbon emissions) and those to maximise (e.g. on-time deliveries) – along with the variables to be tested (e.g. number of electric vehicles and their location given assumptions about range and charging opportunities) – and the constraints to be applied (e.g. customer delivery requirements, limits to driving time, prevailing driving conditions).
It will then be to perform a series of iterative calculations that reveal the variables and constraints that matter most, which are then reviewed and tested to the point at which the ‘the best result obtainable’ starts to emerge.
The methods are not new. They have been the prevail of operations research for over a century.
The data required, however, is expanding and evolving all the time – as is the choice of technology with which to extract and transform data and host the algorithms involved.
The skills needed are therefore in short supply. But they can be learned. And if that investment in skills takes place it will do a huge amount to lay a foundation of data capability that will serve the business well going forward.
The key to success is to start small.
- Build a data strategy around just one use-case, one that has a tight scope, involves easily available data, and skills that you either have in-house or that you can easily acquire – and use that experience as a platform from which to build.
- But keep focusing on that one use-case for as long as possible and use it as a training reference point for as many people in the business as possible.
- Once you have a group of people starting to feel confident about working with data and in possession of a common set of well referenced skills, other use-cases will start to fall into place.
A good time to start is now. If the writing on the wall is to be believed, Environment, Society and Governance legislation (ESG) is going to gather momentum.
Large businesses will come under the spotlight first and will be expected to have a carbon reduction strategy that they can show to be working. Financial penalties for getting it wrong are likely to be significant.
Nevertheless, no matter what the size of the business, if you are confident and at-ease working with data, you will also be confident about making choices that will demonstrate to legislators and shareholders alike that your drive to improve environmental sustainability is authentic and effective.
It will serve you well.
Tel: +44 333 301 0302
We help organisations to develop data strategy, and we train your people to build and use data to unlock value.
If you would like to know more about how we can be of help please complete the form below.