How to create a data strategy
Every organisation needs one. Every organisation deals with data on a daily basis but so many don’t have a strategy in place to maximise the potential of data for their organisation. Help is on hand.
Over the past 6 years I have built an analytics team from one excel spreadsheet to a team delivering analytics for the entire business using multiple platforms. Having not undertaken such an endeavour before this there was a lot of learning along the way. We didn’t have a data strategy in place which in hindsight made things much harder to scale.
This post brings some of those learnings together. It should help anyone undertaking a similar exercise shortcut some of the pain. Whatever you do, you need is a data strategy.
What is a data strategy?
A data strategy is simply setting out how your organisation should interact with data. By having a strategy in place you can maximise the potential for data to transform your business but also take account and mitigate for some of the risk surrounding data.
What should a data strategy contain?
A good data strategy should not be a huge instruction manual on how to deal with different data sources. That would be an instruction manual. It’s also not a step by step guide to implementing the strategy. That would be a plan.
A good data strategy sets out the overall principles for how your organisation interacts with data. It should be a simple document but it should also be a live document that plays a regular role in decision making.
It should be a framework for all big data decisions.
There are six main areas the strategy needs to cover:
Vision and values
Technology and architecture
People and culture
Operating model
Data governance
Accountability
1. Vision and values
Start will your business strategy. Don’t start with the data. This is really important. You probably have a myriad of data sources available to you and so you could formulate a plan for each dataset and how it will be used. This would lead to a strategy that is not going to be aimed at delivering the overall business plan. Senior managers will start to question the value the business is getting from the data.
So the strategy needs to be tied to enabling delivery of the overall business plan.
There are four areas where you can focus your data strategy. The focus you give each of these areas will depend on your specific business and the goals set out in the business plan. However, you are likely to need to touch on all four elements as you compile the purpose of your plan.
Using data to understand your customers
Providing your customers a more customised experience
Improve your internal processes
Monetising your data
By setting out what you aim to achieve the rest of the plan can be targeted at that. A clear vision will help you sell the plan to senior managers but also to your teams as you begin to roll out the plan. A good vision allows every decision you make to be tied back to a clear purpose.
2. Technology and architecture
Once you have defined your vision you can begin to consider the technology and associated architecture.
This section needs to cover:
How will your business ingest data from different sources?
How will you process that data?
How will you automate as many processes as possible?
Where will you store you data?
Which systems across the business will need to accept and receive data?
How will the business interact with the data?
How will you visualise the data?
How you go about doing this will depend on your business. We would all love to build a data strategy for a new business that has no legacy systems. In that situation we can build something relatively quickly.
However, real life isn’t like that.
You need to deal with legacy systems and have a plan for that. Think through the different data sources that you are going to need to connect to and design your architecture and choose your technology with that in mind.
My background is in UK railways. The problem here is a plethora of systems dating back decades in some cases!
3. People and culture
Once you know the architecture and software you are going to be using you can decide the skills you will need in the team. However, beyond the pure technical skills of being able to operate the software you need to also consider some much more important attributes that you will need for your data strategy to thrive.
Leadership. To glue all the different facets together, make quick decisions and set clear objectives.
Project management. Ensuring delivery and to to keep all your projects on course for success.
Listening skills. Ability to ensure you understand what is required from your internal stakeholders
Stakeholder management skills. You can have the most technically gifted team but if there isn’t someone who can manage the stakeholders across the business you will quickly fall foul of not being able to communicate effectively.
Communication skills. Not just the ability to present but also to be able to communicate through their work. Good data visualisation, story telling and tailoring the output to the audience.
The right culture then ensures all these skills work in harmony. As you develop your strategy you need to decide what type of culture you want and then ensure the leadership exhibits that culture. Culture always comes from the top down.
You don’t need everyone in your team to have all of these skills. You just need the right mix across the team so that people complement each other. However, when you do find someone who can do all of this. Hold them close! They are rare.
4. Operating model
This is how you will organise your team on a daily basis. A question you will need to consider in your strategy is the type of model you are aiming for:
Centralised - your data capability sits within a central team and the business comes to that team for their data and analytics needs.
De-centralised - teams across your organisation build out their own analysis and control their own data sources.
Hybrid - some functions sit centrally with local teams drawing on that expertise to carry out their own analysis.
The key here is to look at your organisation and decide what will work for you. You should look to eliminate bottlenecks. Noting this is a very real risk with a whole centralised team.
You then need consider how the strategy will play out in the every day life of the company.
How will projects be initiated, scoped, delivered and iterated?
What processes will you need for quality checking outputs?
Who is going to be responsible for ensuring systems are running, maintained and upgraded?
How many licesss do you you need?
How will you fund the strategy and gain sign off for new developments?
How will you track benefits of the strategy?
As you implement your strategy you will probably need to refine how things are done. Things change. Businesses change. So you will need a planning process that enables new innovations be tried and tested. You may start off with the latest data architecture but how will you ensure you can pivot as new technology comes along?
5. Data governance
How will you ensure there is trust in the data. People need to believe what they are seeing. In addition, you need to have processes in place to hand sensitive and personal data with care. Your information security team will be one of your key stakeholders (rightly so give the impact of a data leak) so you need to have clear processes in place.
How will you keep an information asset register and ensure it is kept up to date?
How will you ensure you are compliant with your organisation’s data policies?
What review processes will you put in place to ensure you always put quality analysis to the business?
As you build you data architecture you need to consider data security as you go. This is not something you can consider later on. It’s too important for that.
Include a data architect in your build team.
6. Embedding in the business
The key to taking a strategy, gaining approval and investment and then keeping it alive as the wider business is buffeted by events is to ensure you have Exec and Board buy in.
If you have a director who can be your champion, they will become someone you can bounce your ideas off and test things before they go up to the exec team. This challenge from a senior person will ensure that when you do present to Board you have covered off the big questions and thought about it from different angles.
Keep the purpose and objectives alive in team meetings. This will ensure the team stay on course. As a senior leader it’s up to you to keep referring back to the plan.
Most importantly, ensure review the plan regularly. Check it against emerging changes to the overall company strategy and business plan. This will ensure it stays relevant.
In conclusion
All businesses are different and will be starting from different places. So, there isn’t a one size fits all approach. However, I believe if you follow the overall themes set out above you can’t go too far wrong. It doesn’t need to be complicated.
If you found this useful, please add a comment or share on your social platforms.