A common mistake that organizations make when mapping out their data strategy is conflating data with the systems that track it. An effective data strategy is not about technology per se, it is about ensuring that your organization is making decisions based on information rather than on guesswork.
Technology helps to make things like data acquisition, reporting, and analysis more effective but in the same way that it makes other business functions smoother. It is the vehicle not the destination. If your decisions are not the right ones, technology may just end up getting you to the wrong place more quickly.
Or to put it another way: useful data tracked poorly is better than the useless data tracked well.
So how do we determine which data is useful? What do we need to focus on when developing our data strategy? Almost no matter the organization the ultimate goal is to make your product or service better, meaning make it more accurately address your customer's needs.
In other words your Data Strategy needs to also be about marketing
A two way street But isn't that the tail wagging the dog? What does marketing have to do with data lakes, big data, artificial intelligence*? Don't we need to learn what the hell hadoop is? What about that vague sense that we need some of that "predictive analytics" stuff?
*which, for the record, is neither
We often naively think of marketing as a push process, ie telling the story or giving information to sway a decision. But that is just one direction of the data flow.
Sure, you do need to educate the customer in how your product addresses their needs but in the end what we have direct control over is the design of our product or service not the thoughts and feelings of the customer.
This means that to be truly effective, marketing should be thought of as a two way street. It must not only provide to your customers the information on the features of your product that will help drive a sale but also provide to your design team the information on customer needs that will help inform the design of those features in the first place
Of course building features costs money and takes time so what we are really after is not perfection but instead getting the maximum return on investment (ROI). That is to say the maximum alignment to your customer's needs with the minimum necessary investment in time, resources and budget.
Profit is a lagging indicator of dissatisfaction
This is where your data strategy comes in. At it's essence it should be about providing the data for the metrics underlying that ROI calculation, measuring how well we are aligning to customer needs and how much that alignment is costing us.
A first order metric for ROI is to simply measure the revenue (or successful outcome for a non-profit or government entity) and the resources needed to achieve that revenue level. Having that basic data capture in place is certainly a foundation.
However this revenue-centered approach is too late in the process. By the time your sales start suffering because you're building the wrong product or offering an irrelevant service, it's already too late to change. All you can do at that point is prevent further loss.
Taking a marketing point of view when developing your data strategy helps to shift focus to earlier in the process on both the resource and sales side. It helps you to come up with more sophisticated measures of customer satisfaction and costs incurred, so that you can begin to move from reaction to prediction. In a perfect world you may even be able to tailor your offerings to what the customer doesn't yet even realize themselves that they need.
For example, instead of measuring customer satisfaction indirectly via sales, why not ask? Gather data explicitly. Surveys can be scientifically designed to suss out true sentiment, focus groups can be convened to get more qualitative feedback. Combine this with financial data on sales and you start to get a more nuanced view of what is actually driving your customer.
Depending on your product or service, you can take this a step further and instrument your offerings to gain usage data on the fly and in real-time. And with a more robust and comprehensive data set you can start building models whose output accurately tracks historical sales. A first step towards a more predictive and less reactive design process.
Predictive not Reactive In the end your data strategy should be about accurately and precisely knowing the needs of the customer so you can feed that into the design process from the beginning. That way you have the chance to avoid wasting money building features or services that your customers no longer care about.