Stop Trying to Monetize Your Data

How data can help your customers take actions based on insights

Every week in our briefing center, we talk to companies about unlocking the potential of their data, and the same question crops up. I often find some version of it in the pre-brief document I get as a regular speaker for these data overview sessions. But it’s usually the same: “How can we monetize our data?”

While I’m excited to talk to companies about the possibilities for their data, I find that trying to answer that question directly misses the real question: What problems can you solve — or solve better — with data?

Looking at data in isolation from the problems it could help solve is a widespread misstep made by organizations. Nick Heudecker, research director at Gartner, recently noted that “the big issue is not so much big data itself, but rather how it is used. While organizations have understood that big data is not just about a specific technology, they need to avoid thinking about big data as a separate effort.”

This is not to say that monetization doesn’t or can’t happen. Rather, consider how your data can help solve specific business problems. This indirect but potent path can eventually lead to data monetization.

Start with the problem, not the data

Approaching your data with the sole purpose of spinning it into gold is — surprise! — the stuff of fairy tales. Starting with your data and trying to find the money is actually more like panning for gold: labor-intensive, stressful, and inefficient. You can spend a lot of time sifting through data for something promising, but you still have to validate that it’s something someone will pay for.

Instead, start by identifying business problems that exist and treat them as opportunities. Here are some broad categories of ‘problems’ and opportunities that data can help address:

1. Customer problems you’re already solving: Your data will have the greatest impact on the core customer problem you already solve. Whether it’s delivering packages or providing retirement plan services, this should be the reason you exist as a business. Your data can help improve how you solve these core problems, and, in some cases, instantly provide a path to scale.
2. Customer problems you haven’t solved yet: No matter your industry, your customers have many times when they could make better decisions about using your product or service, and, as a result, have a better experience. For example: a student at a university might enjoy knowing which courses to take to earn a higher GPA. A car owner would appreciate understanding how fast to drive to optimize fuel efficiency. Or in general, you could use data could tell your users how long it will take to have a problem with your product or service addressed.
3. Operational problems that cost organizations money: Many enterprises have lists of inefficiencies — like a telco or bank with high customer churn, a utility company that’s losing revenue due to service theft, or unexplained application outages that reduce productivity. Issues like these can become high priority areas based on the estimated impact to the business.

As you think about current and potential problems to solve, consider what type of person needs to do something differently as part of that solution: Is it a field technician? A support center rep? A specific customer persona? A user?

A Recipe For Smart Applications

“Monetize” through the application

Once you know whose problem you are trying to solve, understand how they currently solve this problem. Some questions to guide you in this process include:

1. Where do they go for their information as they make their decisions or take actions?
2. What steps do they take?
3. What applications are they already using in that process?

You either already have an application in front of this target user or you don’t. A lot of enterprises have existing applications that could be augmented. Other times, it makes more sense to build an app from scratch.

Regardless, the application needs to let the user take an appropriate action as part of solving that problem. It can’t simply be an information portal. Why? Users generally come back to apps that allow them to take meaningful action. For example, if the target user is a bank customer, they should be able to transfer funds, not just look at their bank balance. If it’s a customer service rep, they should be able to dispatch a technician or issue a credit to their account.

Users generally come back to apps that allow them to take meaningful action.

Now, consider that action as the one you want to optimize. Sometimes it’s relatively simple: If the target user is a consumer and they are able to purchase a product or book a service, you are probably trying to drive more of those actions. Sometimes, it’s a little more complex: If the action is to dispatch a service technician, perhaps you want to dispatch the closest technician that isn’t likely to require overtime pay. Influencing these actions as they are happening is what can drive additional revenue or lower costs.

And that’s precisely where data comes in.

Use data to improve the experience

At this point you should know the user, the application, the action, and which way you want to influence that action. You can influence that action with data by offering decision support directly in the app at or preceding the action in question. This is not just data surfaced in an analytical report — most users don’t have time to interpret a bar chart or scatter plot. Instead, it’s a feature that directly offers a tip, recommendation, or estimate.

Data was used to improve Netflix recommendations (left), Uber trip estimates (middle) and give Airbnb hosts a tip on what to price their rooms (right).

Unpacking what goes into one of these data-driven features reveals a number of requirements. Predictive analytics and optimization math is nothing new, but the scale of data that is available to help build those models strains legacy systems. Being able to run analytical queries and libraries of analytical functions, like R or Python, in parallel across a horizontally scaled database, markedly changes the sophistication and speed of your data models.

For Cloud Native, Data is the New Black

Analytical models — however sophisticated — are still in essence just math that produces a table. Tables by themselves are not enough to prompt a user to take an action. Getting there through analytical models means asking and answering two common questions:

• Approximately when is something going to happen?
• What is the best option range of potential actions?

Notice that neither question seeks an exact answer. Predictions and optimizations can have a range of confidence levels that leaves some margin for error. Given what users on Netflix or Amazon are trying to do, a recommendation reel functions best as a range of curated option. AirBnB’s “price tip” feature offers a slider bar to show a range of suggested listing price points and relative degrees of booking likelihood.

Finding the right way to surface the output from an analytical model to drive behavior and better meet user needs can take several iterations. This is where an agile approach — test-driven, iterative releases; a platform for continuous delivery; and blue-green deployments — all support data-driven applications. Further, user feedback and analyzing clickstream data are essential to fine-tuning the experience for users.

It’s about value

Anticipating a user need at the right moment and changing a user’s behavior is no trivial task. But if you can present an easier, faster, smarter alternative, you have an opportunity to promote a new or different action that is relevant to the user.

In the end, your data capabilities and models are only as good as how well you get those insights into the hands of users at the moment they need them. And it is this very place where the user finds value — in smart alternatives and features — where you can finally fulfill that promise of monetizing your data, by making it useful.

Change is the only constant, so individuals, institutions, and businesses must be Built to Adapt. At Pivotal, we believe change should be expected, embraced and incorporated continuously through development and innovation, because good software is never finished.