The what, why and how of effective strategies to transform data

Open the transcript in a new tab

CHASEY DAVIES-WRIGLEY: Welcome everyone to this discussion about data strategy. In particular, the what, why and how of effective data strategies.

My name is Chasey Davies-Wrigley, and I am a Principal Data Engineer at Made Tech. I am a technology leader with a highly technical background including an MSc in Computer Science. I have over 22 years industry experience in both the private and public sectors.

During my many years in the industry I’ve been responsible for distributed multi-disciplinary teams, building, architecting, delivering and maintaining highly reliable and scalable systems across a variety of platforms and technologies.

At Made Tech, I am focused on empowering the public sector to deliver and continuously improve digital services that are user-centric, data driven and free from legacy technology.

I have an interesting agenda for you all today and it’s an area that I’m really passionate about – data strategy. We will be looking at what we actually mean when we are talking about an organisation’s data strategy. Lots of organisations do have data strategies, but there is still a fair way to go to make sure they have an effective one.

I’m going to start off today with a look at what we actually mean when we talk about data strategy. What is it aiming to achieve, and what function does it play in the bigger picture that is your organisation?

That should naturally lead us in to discussing the benefits this strategy brings and the importance of investing in creating a strategy that really works for your organisation. After all, time is always a commodity in contention, so why should your organisation invest its precious time in devising a data strategy?

I’m then going to focus on how to build not just a data strategy but an effective one. Where do you start, how do you progress, and how do you maintain that momentum?

As you will see, an effective data strategy is actually a journey.

Without further ado, let’s kick off with a look at what we mean by a data strategy. Well, there are so many different definitions of what a data strategy is and they all attempt to group different aspects under a different number of key headings, areas or pillars. I’m going to start off with Gartner’s definition, that a data strategy is a highly dynamic process, employed to support the acquisition, organisation, analysis and delivery of data in support of business objectives.
Let’s unpack that statement and then take it a step further, make it tangible and build on it. As a stand-alone definition I think it’s rather lofty, and doesn’t easily lend itself to being able to think of it in concrete terms.

I would totally agree that the strategy should be a dynamic process, as it needs to change, to evolve. The important aspect of a data strategy is that it supports the organisation’s aims and objectives, it supports the business strategy. This means that in the same way organisations differ in their business strategies, so too will they differ in their data strategy.

There is no one data strategy that is fit for purpose across all organisations. It’s also important to recognise that as an organisation’s objectives change, so too must the data strategy.

Let’s dig a little deeper into the other aspects noted in Gartner’s definition. It’s a process employed to support the acquisition, organisation, analysis and delivery of data. Okay, but what does that actually mean? What it’s saying is that it’s a process to support how we get, store, look at and then serve that data to areas of the business where it will have an impact on business objectives. However, Gartner’s definition for me only tells you what strategy is for but not really what a data strategy consists of.

For me, a data strategy is a clear vision of how an organisation intends to manage and use its data. It contains a framework that explains how the organisation’s data can deliver value. As I’ve already mentioned, that value is based on the organisation’s goals and objectives and its business strategy.

It’s important to remember that your data strategy will be a tool used to advocate for data use and investment in projects, teams, technology and other resources. It’s so important to get this right because an ineffective strategy can really damage the reputation of data within an organisation. Data should be an enabler to the business function and not a blocker that prevents other key services being delivered. You might say, “Well, that’s pretty obvious, Chasey.” But I’ve seen even recently, an organisation where all their digital services sent logs to be ingested by a central data platform. That platform required several weeks of development before it could ingest new services. This resulted in new services being delayed in their release to the public until the development had been completed.

Obviously a case where the data engineering function is actually blocking the delivery of new services. So, it’s actually hindering the business objectives.

Next comes people and culture. Your data strategy will require ownership, accountability. It’s success heavily relies on the backing from all levels of your organisation. Aligning the data strategy to the business strategy should incentivise this backing, but the strategy is going to need an owner. Usually, someone like the Chief Data Officer. They will need to really champion the benefits, identify the value for everyone in the organisation to get involved and to drive the delivery of the strategy. They will likely have a team of people to help them with this made up of business stakeholders, solution, delivery and of course, governance. It’s really important to understand how much the success relies on all of the people in your organisation.

Ask yourself, does your culture support what you want to do with your strategy? Just documenting a strategy isn’t going to have any tangible effect. It needs to be acted upon, results need to be seen and momentum maintained.

We also have the operating model. What am I talking about when I say operating model? In part, this refers to how you intend to structure your organisation. How will you organise those people involved in driving your strategy forward, like the business stakeholders? Those people who are needed to deliver your strategy, like the data analysts, and those who build the foundations for your data, like the data engineers. Do you put all those people in one central team? Do you distribute them across your organisation, or do you have a hybrid model with some services provided by a central team like data engineering, but then you have analysts embedded in functional teams working side by side with their subject matter experts?

In addition to the organisational structure, as part of the operating model you also need to think about the methodologies you want to use. No doubt you will want to ensure that you are agile and responsive but also proactive. That you are iterative but also innovative, with the value and the insights that your data yields.

Next up, we have data governance. This is a large area. This is not just about what you intend to do to ensure you are compliant with laws and regulations like GDPR and the right to be forgotten. It’s also about how you ensure the quality, the resilience, and the accuracy of your data. People really need to have confidence in the outcomes, the analysis and the numbers. The data needs to be correct and people need to be able to trust it.

We also have technology and architecture. Technology is about deciding on which tools you want to advocate using, and architecture is about designing how you use those tools together, what applications you should be building. At a macro level, it involves ensuring that you align your technology choices with your organisation’s digital strategy.

Which cloud platforms and technical skillsets does your organisation want to invest in? It also involves looking at specific areas like ingestion. How does data come into your organisation. Lineage – how does data move through the system, what changes are made to it along the way, and how are those changes tracked? How is the data transformed? How do you manage, govern and look after your data?

What about analysing, visualising and gaining insight from your data? How will you build models, algorithms and machine learning to deliver real, high value output. How can you share your data, potentially to other systems, other departments, other organisations? How can you democratise your data?

Lastly but not least, there should be a realistic roadmap with achievable milestones. Identify the low-hanging fruit and deliver value as soon as possible. That will really bolster the support from stakeholders, and help maintain the momentum in the delivery of your strategy. If people have to wait too long to see the benefits of anything, they lose interest, focus, and willingness to invest both time and budget into any strategy. Your data strategy is no exception, so a roadmap with quick wins is an essential element.

You could take the three horizons approach here, where you set out what you aim to deliver within the first three months, then six months to a year and then three years, for example. It’s really important though, to ensure that your CDO or whoever is responsible for your data strategy is kept accountable to deliver against the goals laid out in the roadmap.

Having a roadmap that is regularly monitored to check progress against will be a huge success factor in your data strategy.

Let’s look at why your organisation needs a data strategy. The usual benefits cited in order to get buy in go from monetising your data, sell more, sell quicker or gain a competitive advantage in the marketplace. These are not the same driving forces for the public sector.

For the public sector, the benefits are seen more in terms of the benefits it can deliver to citizens. How can we use data to make a citizen’s journey through services easier, more relevant to them and ultimately, more impactful for our society?

The benefits are centred around providing more efficient, effective and trustworthy public services. That doesn’t mean that cost-saving benefits achieved through better use of data are not relevant to the public sector. Having the access to the right data at the right time can really help stretch a department’s budget. It can be used to help inform decision makers on areas where budget can be applied more effectively on where it can have the most impact.

A data strategy can increase public value in three main types of activity. In forecasting and planning – by this I mean the role more efficient use of data can play in the design of policies, planning of interventions, anticipation of possible changes and the forecasting of future needs.

Service delivery – how the use of data can inform and improve the implementation of policy, the responsiveness of government and the provision of public services.

Finally, through evaluation and monitoring. The approach to data involved in measuring impact, auditing decisions and monitoring performance.

So I’ve described what a data strategy looks like, and why organisations should have one. How do you ensure its effectiveness? Documenting a strategy and publishing it will mean nothing if it’s not implemented. The driving force behind any implementation is people. One of the most difficult areas to address and bring about change is when it involves changing how people work. Your data strategy is likely to challenge how people have always done things. For example, subject matter experts may be used to making decisions based on their intuition, whereas your strategy is likely to be championing data driven decision making. SMEs could see this as the organisation dismissing their experience as no longer valid or good enough. They may actually become actively hostile to ensuring the success of the strategy.

A great way to combat this would be through increased communications and data literacy training. Be really mindful of the terms that you use, and how you convey them. For example, the term ‘data-driven decision making’. That could be explained in more detail so that people in your organisation understand that in practice, it’s actually more like data-informed decision making. You should never be in a situation where you are disregarding the opinions of your SMEs. It’s likely they will know more about socio-economic factors that come into play and influence the situation differently. They are experts in their fields for a reason, and the data should be there to help them and to inform them even further.

It’s really the responsibility of the CDO, or whoever is responsible and accountable for the strategy to champion the benefits of the strategy to everyone in the organisation. They should translate the strategy into the benefits for each stakeholder, data producer, department and team. People in the organisation need to know what is in it for them. Make the strategy relatable to them by understanding their business function and their goals. Always translate the data benefits into the reason why that data is needed and what can be done with it, to get buy-in from all levels.

After all, everyone’s goals will be aligned to the business goals at some level. As the data strategy is aligned to that, there should be some identifiable benefits for everyone.

In order to gain executive level buy-in, it won’t be enough to just promise them dashboards. Benefit really needs to be communicated in business terms of time, budget and improved outcomes for citizens. When talking about data value and benefits, we need to understand an organisation’s context. What is this data, how are you using it, through what process and for what purpose? All of these questions have the added benefit that they also help to focus on what the critical data is. Otherwise you could end up just bringing across all the data, and you could just end up being totally swamped by it all.

Actually, the business value and context is understood better when technical departments aren’t isolated, when business and technology functions work closely together to deliver outcomes.

Implementing a data strategy effectively is going to take a high degree of collaboration, transparency and inclusiveness. This also means there needs to be clear roles on who does what, and why they should be part of the journey. Focus on how to help everyone in the organisation. Really trust the data. Have access to it easily when they need it, in an agile and scalable way. This culture really needs to be built in parallel to build in with your data strategy. A great starting point is by assessing readiness. Sometimes this is called a data maturity assessment. At Made Tech we prefer to call this discovering your data opportunities.

Discovering your organisation’s data opportunities really centres around an in-depth look into three main areas of your organisation; your people, your processes, and your technology. Your organisation can actually bring in specialist consultancy services to help analyse those areas, synthesise the results and provide an action plan to follow.

This analysis will involve ongoing effort, talking with different stakeholders, looking into your organisation’s culture. To what extent are people open to accepting change? How data literate are they? How many people across all the departments in your organisation really buy-in to the idea of using data to help them make decisions, and to what extent are you building data skills across the organisation as a whole?

When looking into your organisation’s processes, it involves focusing on identifying to what extent your processes ensure confidence and trust in your data. What does your data governance look like? This is not only important to ensure data quality, but it also includes looking into the data security, privacy and regulations, and even covers areas like what data you keep and how you archive it.

How closely are your organisation’s processes aligned to the fair principles, where data is findable, accessible, interoperable and reusable? When we take a look at your organisation’s technology, w will be focusing on to what extent your organisation’s data platforms help or hinder your business objectives. Do they support the fair principles I just mentioned? How modern are they and your data engineering capabilities? It’s really not advisable to develop a data strategy that jumps straight to advanced AI and machine learning without first having the appropriate foundations in place.

After the analysis when your organisation has its action plan to follow, it is important to recognise that recommendations will not6 be achievable overnight. They really need to be tracked and measured to ensure progress is being made. The analysis can also be repeated again further down the road, and perhaps a different approach and set of recommendations will be needed. Realising an organisation’s data opportunities is a journey. Ensuring your organisation is not standing still, is moving forward, is really important.

Before we get to questions, I would like to just give you some contact details for Made Tech and for myself. Please feel free to reach out and ask anything that you may think of at a later date. Made Tech can help work with your organisation in defining an appropriate data strategy. We can also help you to discover your data opportunities and provide the relevant recommendations to help you along your journey. Thank you for your time today, I hope you have enjoyed my insights and that you have gained something of value.