Is Your Company Making Use Of It's Data

There are many ways to put data to work, and companies, especially their leaders, are advised to explore as many of them as they can. Each presents distinct opportunities for profit and competitive advantage, from product improvements to new revenue streams to possible industry game changers. At the same time, each presents challenges that must be experienced to be appreciated.

While big data, analytics, artificial intelligence, and the internet of things garner the lion’s share of media attention, using data to its full potential is much more about management than it is about technology. A team of data scientists may employ a series of clever analyses to yield an important insight, but that insight will die on the vine if others in the organization don’t carry it forward by developing a deeper understanding of the implications, making a critical decision, building it into a product, or leveraging it in interactions with customers. Putting data to work includes the whole sequence, from data to insight to profit.

In working with companies on getting more from their data, I advise managers to explore seven methods to put data to work. I also urge all leaders to initiate department- or business unit–size trials of all these methods, so they can learn how the options work and which would be best for their business.

• Make better decisions. First, use better (more relevant, more accurate) data when making decisions, up and down the organization chart. I’ve not worked with or heard of a company that didn’t freely admit that it needed to make better decisions — and many push hard to improve. But incorporating more and better data into decision making can be difficult. You must learn to understand variation, to combine data from different sources, and to drive decision making to the lowest possible level. By taking the time to learn these skills, though, you can use data to reduce uncertainty, increasing the chances of making sound decisions.
• Innovate products, services, and processes. Use data to uncover hidden insights, and use those insights to create or improve products, services, and processes. For example, at Morgan Stanley, Jeff McMillan and his team aim to improve working relationships with their wealth management clients by analyzing everything from client goals and portfolios to available investment products to email. An algorithm then takes this information and suggests actions, at which point advisors choose the best ones to suggest to their clients. McMillan encourages advisors to “imagine you have a conversation at 6:00 PM every evening with a Harvard MBA with 800 years’ experience. You tell her what you’re thinking about, and she thinks through your clients’ opportunities all night long. In the morning, she presents you with a list of your 10 best actions for the day. Wouldn’t that help you make your clients happier?” Their goal is to develop personalized strategies for each client based on far more data and analytic horsepower than any financial adviser could marshal alone.
• Informationalize products, services and processes. Build more data into what you offer customers, so you make existing products more valuable. Automobile manufacturers have a history of working on this by adding warning lights, GPS, distance-to-empty gas tank notifications, and other features almost seamlessly. I’ve yet to run across a product or process that wouldn’t benefit from more data.
• Improve quality, eliminate costs, and build trust. Proactively address quality by finding and eliminating the root causes of errors. Virtually everything a company does, from delivering products to running the place, uses enormous quantities of data. But bad data makes this work more difficult and increases costs — up to 20% of revenue! You can’t expect someone to factor data they don’t trust into an important decision. Take steps to actively track down data quality issues and eliminate their root causes.
• Provide content. Sell or license new, richer, or more targeted data. All customers depend on content, and thousands of companies, such as Bloomberg and 23andMe, aim to fill the need. Still, most companies don’t think much about selling their data. But doing so can provide great opportunity. For example, car insurance companies discovered a relatively simple piece of data they could sell: the number of new policies written each day. New car sales reflect the health of automobile manufacturers and are of great interest to investors. But manufacturers release sales figures monthly — an eternity for investors. Since each sale requires a new insurance policy, the number of new policies issued each day provides a faster indicator. This becomes a profit stream for the issuers and for Quandl, which aggregates this data across the industry and packages it for investors.
• Infomediate. Connect data providers and those who need the data. Here, the goal is not to provide content but to provide direction toward content. Google is, of course, the best-known example, but Quora, too, helps people find answers when expert help is needed. And there is huge opportunity here for others. In both their personal and professional lives, individuals spend hours each week looking for documents, reports, and other data. Find ways to connect these individuals with others who can provide the answers they’re looking for.
• Exploit asymmetries. An asymmetry arises when one side of a transaction knows something that the other doesn’t. Exploiting this knowledge helps them drive a better deal. Hedge funds and used car dealers use such data to create and leverage asymmetries. More recently, sports venues, airlines, and others have begun using variable pricing to capture maximum revenue from consumers. All companies can examine sales and related data more deeply in search of such opportunities. Conversely, closing asymmetries, as Carfax does for used cars, can also present great opportunities.

Each of these seven options can help your company put data to work, and in many cases a combination of these approaches can create incredible value. For example, Liz Kirscher, head of talent acquisition at Morningstar, and her team are rethinking their hiring process, looking more closely at existing data, incorporating new data, and bringing more discipline throughout. Important prongs of Kirscher’s approach include innovating by using artificial intelligence to better screen resumes; informationalizing by using the Hogan score to better understand “grit” (a predictor of success at Morningstar); and making better decisions, bringing greater transparency to the measurement of hiring success and failure.

As you explore these approaches, you’ll find that some work better or provide greater value than others. You may discover, for instance, that your decision making improves with more data (especially once you’ve fixed some underlying quality issues), but that selling customer data goes against your company values. Or that it’s relatively easy to make incremental improvements to products through informationalization, but you don’t yet have the right talent for larger innovation initiatives. Test options and learn quickly! And as you do, crystallize those ways of putting data to work that create the most value and profit for your company, and implement them in your long-term data strategy.