Every business leader knows they need to be working with data, but with so many potential efficiencies and exciting use cases, it can be difficult to know where to start.
I have a straightforward process for this that I take my clients through when we sit down to consider how they could be using data more effectively.
Of course, every business should already understand its strategic goals. What are the most important things that need to be done for the business to grow and succeed? Everyone in the organization should be aware of these. Are we looking for profits? Growth? A better understanding of our customers? To reduce customer churn or wastage in our processes?
So when it comes to identifying data use cases, the best jumping-in point is to brainstorm initiatives that can help us meet one (or more) of these goals.
High-level use cases
At the top level, there are six areas where we can put data to use to help us hit our goals. They are:
- Improving decision making
- Understanding customers and markets
- Creating better products
- Offering better services
- Improving business operations
- Monetizing data directly
Netflix is often highlighted as a great example of a business that has pioneered intelligent data usage, using data across all of these areas. Using the data it gathers as its customers use the service, it makes decisions about what movies and shows to recommend, as well as decisions about what content to commission or create itself. It monitors video playback quality to ensure customers are receiving a good level of service, and it monetizes this information through its advertising partners.
Tying these high-level use cases into a strategic goal or two is essential simply because the wealth of opportunities offered by data make it easy to get distracted or side-tracked into projects that aren’t of immediate importance. This can be particularly true if you’re in an industry where levels of data literacy are high, and there are shiny and exciting use cases appearing every day. Considering use cases without first considering how they help you achieve your strategic goals is a recipe for wasted time and money!
Once this is understood, I recommend gathering all of the key stakeholders – those who will have a part to play in your transition to becoming a data-driven business and those who will be affected – and brainstorming ideas.
Keeping the strategic goals and high-level use cases in mind, come up with as many ideas as you can for where data could help you to bring them together. For each idea, try to answer the questions:
What is the objective? For example, are you trying to understand what age range your customers are in? To identify inefficiency in your design, manufacturing, logistics, sales or after-sales processes? To stop customers leaving for another provider when their contract expires? Where to most efficiently spend your advertising budget to grow your audience and customer base?
How will we measure success? Every idea should be tied to a metric or indicator that can be monitored to give feedback on how successful it is. There’s no point doing anything if we don’t know how to tell whether it’s working or not!
What data will we need? Do we have everything we need to know already, or are we going to need to implement new data monitoring and gathering initiatives to find it? Will it be internal company data, or do we have “data gaps” that need to be plugged with external data bought in from a provider? Will it be structured or unstructured data?
What technology and skills do we need? Is everything we need going to be available through cloud services, or if we have more specific needs or are dealing with very sensitive data, are we going to have to build bespoke infrastructure? Increasingly, the solution these days is a hybrid one. A basic understanding of the technology requirements is useful at this point, as, of course, is some insight into whether you have the capabilities to carry out your idea in-house or are going to have to look for partners or new recruits.
What are the governance requirements? When you’re dealing with data – particularly sensitive personal data – you have to be certain you are complying with all necessary laws and regulations. This is for two reasons – firstly, to avoid the severe penalties that can be incurred by misusing or misplacing data, and secondly, to maintain the trust of your customers. Being found in breach of the EU GDPR, for example, can potentially result in a fine of two percent of global turnover, or 10 million Euros – enough to kill most businesses!
Who will be the data customer? These are the end-users of the data you’ll be working with – they could be your actual customers, or they could be people within your organization who will take the insights you uncover and turn them into action. It’s important to have a clear idea of who this will be because it impacts the way that your data will be communicated – a critical element of any data strategy.
Who will be the use case owner? This is the person who takes overall responsibility for implementing the idea and monitoring the outcome at a tactical level. Every data use case should have an owner, although an owner may end up with responsibility for more than one use case. It could be the person who conceived the use case, or someone with an affinity for data, or someone who is close to where the data will end up being used. It might be someone who ticks all of those boxes.
Choosing your first use cases
Brainstorm as many prospective use cases as you can – as long as they fit a strategy goal, a use case category, and you can answer the questions above, then they can go on the list. But once you have as many as you can think of, it’s time to start thinking about which ones you want to start working on first. During the brainstorming, you may very well have come up with ideas that, realistically, are too difficult, too expensive, or too large-scale to work right away. Brainstorming them is still a very useful exercise, though, as it gets everyone thinking about the myriad of ways that data flows through an organization and how it could be put to better use.
This is the point, though, where I recommend shortlisting around five realistic, achievable and worthwhile use cases to move forward with first. Any more than that, and they risk becoming unmanageable. These should fall into one of two categories – “majorly transformational” or “quick wins”. Majorly transformational initiatives are those that are likely to have a direct and long-lasting impact on your most important business goals, metrics, and KPIs. Quick wins are just as valuable, though, because they are easier to implement and should show a quick return – either in how they impact KPIs or, just as importantly at this stage, on how they impact data literacy and buy-in within your organization. These are projects that can speedily demonstrate the value of data and data-driven change and help develop an organization-wide appetite to move on to bigger and more transformative projects.
Choosing the right use cases for your data strategy is just one of the topics covered in depth in the second (and completely updated) edition of my book ‘Data Strategy – How to Profit From a World of Big Data, Analytics and Artificial Intelligence. Click here to buy it!
Who will be the use case owner?
At this point, every use case needs someone who will take overall responsibility for ensuring that it can be delivered. As we already know it’s in-line with business strategy and we understand how to measure it’s effectiveness, we need someone whose responsibilities purely lie with making sure it is planned and implemented at a tactical level. They will be responsible for overseeing the deployment of the technology and the data, ensuring that performance metrics are monitored, and reporting on the outcome of the use case. Of course, particularly in a smaller organization or one where buy-in for data-driven transformation is still being negotiated, there may not be enough people involved for every use case to have an individual owner, so one person might take responsibility for several use cases. Alternatively, it could be a member of the team that the use case will impact – marketing, for example, if it's a marketing use case, that understands the importance of shifting towards a more data-driven culture.
As our ice cream shop is currently a very small enterprise, consisting of the owner (us) and a couple of part-time sales assistants who help out when it gets busy, it makes sense that we will be the data owners ourselves.