Three models for determining the true value of data
Companies today sit on a treasure trove of enterprise data. This includes structured data that they have always been able to leverage, as well as unstructured data that they can make usable now through innovative techniques and tools around data processing, extraction and classification.
When data scientists apply analyses to these data points and values, they can extract strategic insights that help their businesses make smarter decisions and edge out the competition. They can make timelier decisions when faced with radical changes in the marketplace, improve the accuracy of these choices, and mitigate their associated risks.
What’s more, they can gain a better understanding of their customers as well as possible gaps or issues within their existing operations to drive improvements.
With all the possible business outcomes with enterprise data, it is up to data scientists to identify which data points or datasets possess the value they need to target their analyses and generate meaningful change. To do so, they can structure the value of their data in three ways: intrinsic value, derivative value and algorithmic value.
Companies can aggregate their data, monetize it and sell it through a data-as-a-service model. Equifax does this every day by compiling consumer financial records and selling them to banks, credit unions and financial institutions looking into credit histories to guide lending decisions.
Likewise, Dun & Bradstreet has built its business on possessing a large database of commercial data that it sells to other companies for the purpose of evaluating potential partners. Moreover, many retailers possess large volumes of data on consumers that they will sell to manufacturers looking to get knowledge on the marketplace and buying behavior. The monetized price for these datasets provides a measure of their intrinsic value.
Beyond the direct monetization of the data, an organization can gain new value out of the information by combining it with another dataset, looking at different correlations and making inferences.
For instance, an agricultural company can take weather data and combine it with data on soil and crops – all organized geographically by zip code. Data scientists can analyze and make connections between the datasets to help the business figure out the best fertilizer and pesticide combinations that will optimize crop production in different regions.
Also, banks can apply graph analytics, which can help identify connections among documents or individuals in a dataset. They can look at financial transactions to identify a possible network of money launderers. They can then combine this data with sensor data, such as point of sale charges and phone calls, and data scraped from the web, like product catalogue searches, to find these individuals.
This derivative value requires companies to apply business model thinking, evaluate adjacencies between datasets, and use domain knowledge of the industry.
Natural language processing can help extract key entities in unstructured data as well as help identify sentiment. Connecting once unstructured data with other datasets demonstrates how this once intangible information can provide newfound insights.
For instance, a company could comb through company emails and extract data, such as sender, subject and time stamps. Then, data scientists may find a connection between certain email patterns and employee performance. This information can help the company then find indicators of top employees and catch early signs of waning performance among its staff.
It is also important for data scientists to consider which data combined with which data provides lift to a business application. Take into consideration whether the additional cost of procuring and maintaining the data is worthwhile for a particular objective.
One of the top ways that companies are capitalizing on their data is prediction, namely making more predictive business decisions and anticipating customer needs.
For instance, Amazon uses data on what consumers have previously bought and correlates the information with similar customers' purchases to make new recommendations. Through machine learning, the suggestions become more accurate as people make more purchases over time.
Thus, Amazon can increase its order numbers and size through smarter cross-selling and upselling, driving growth. In the same way, a commercial lender can use data from previous loan transactions to develop a recommendation model. Sales agents can then determine other financial products to suggest to a borrower.
In another example, insurance companies can collect photographs from car accidents and claims information for each case and apply them as labeled data with a machine learning engine. Computer vision algorithms can help classify the images. Using these records, the company can better predict claim payouts for future cases.
Whereas before each accident would require a long, multi-step assessment, the insurance company can instead quickly determine the claim payout or determine if they need to do a more in-depth review for a special case. It simplifies the process for both the organization and customer.
As proclaimed by The Economist and others, the world’s most valuable resource is no longer oil, but data. However, unlike resources such as oil and capital, the value does not come from simply having or purchasing large amounts of datasets.
Data scientists are vital to their companies’ future successes, as they hold the skills needed to identify and capitalize on the true value of data. With targeted analysis, they can uncover meaningful insights to guide decisions that impact products, customer experience, risk mitigation, processes and profitability.