“Your Company’s Data Could Help End World Hunger”The possibilities that data philanthropy opens up can be immense. As part of the popular TED Talks series, Mallory Soldner, advanced analytics manager for logistics company UPS, gave a striking example. In her presentation “Your Company’s Data Could Help End World Hunger”, she explained how 17% of costs were saved in a humanitarian program to feed 500,000 people, making it possible to feed tens of thousands more with the same funds. Access to the right data, tools, and skills made it possible to optimize food assortments, purchases, storage, and distribution.
3 Essential Steps to Useful ResultsCompanies have a major role to play in solving such major problems, says Soldner. She points out that data philanthropy must often go further than simply handing over datasets. There are three key steps to getting results that can help populations in any region of the world:
- Donation or sharing of useful data
- Availability of data/decision scientists
- Access to requisite technology
Protecting the End-Users and Companies Providing the DataHowever, sharing data has its challenges too. Companies that have collected the data may fret about the possibility of nullifying a competitive advantage or even exposing themselves to counter-attacks from competitors in their market. Even more serious, the Internet privacy of the user, consumer, or customer, whose data is being used, may also risk being compromised. “Re-identification” in either case is always a concern. Accordingly, techniques exist to anonymize data and thus protect the confidentiality of the people to whom the data relates. Global Pulse also offers four different tactics to help ensure confidentiality for enterprises making data available:
- Share data sets for analysis using a nondisclosure agreement (NDA)
- Under NDA, allow researchers to analyze data within the perimeter of the donating company
- Aggregation of data from multiple companies in the same industry to prevent re-identification and protect competitiveness (“Real-Time Data Commons”)
- The company analyzes the data within its perimeter and shares the indicators (“Public/Private Alerting Network”).
- Gathering and Contributing Data Science Skills
Examples of Technology in Data PhilanthropyApplications now in use that rely on data philanthropy style tactics include HealthMap, a freely available website and mobile app software to track the outbreak of diseases. A sister website called Flu Near You gathers weekly reports from users about their own health conditions. Conventional flu monitoring may require up to two weeks to confirm outbreaks, but Flu Near You accelerates warnings via data-mined insights. Elsewhere on the Net, Twitter has started a trial program, “Data Grants,” to give preferential access to its real-time APIs to certain academics and researchers. In this instance, the social network is foregoing its normal price for access, which would exceed most public sector research budgets.
Why Data Philanthropy is Good for BusinessGiving back, corporate social responsibility, stakeholder pressure, all these notions are reasons why a private enterprise would want (or feel morally obliged) to become a data philanthropist. After all, enterprises are often getting a free ride on the back of the massive amounts of user-generated content on the open web, which they collect in addition to any proprietary data. Yet many companies consider that their data philanthropy is less an act of charity, and more an opportunity to mitigate business risk and to foster innovation. They understand that the wellbeing of a population is usually of critical importance to the continuity and growth of their own business. As a general example, a company may invest in an emerging market with good potential, but find that a food or health crisis is threatening to leave customers in that market unable to afford what the company offers. Yet prior analysis of data possessed by the company might have given early warning signs of problems before they got out of hand. This is particularly relevant for mobile phone operators, usually able to mine their user data to detect shifts in population, trends in usage, and so on.
An Advantage in Acquiring Talent“Not just a paycheck” may be a fashionable motto, but it also reflects the desire of many employees to accomplish things at work that they consider to be worthwhile. In their own words, they “want to make a difference.” A corporate data philanthropy policy shows them that this is possible. Data and decision scientists may be more attracted to an enterprise with such a policy, suggests Soldner, engaging better at work and staying longer with the company as well.
Steps for Launching a Data Philanthropy InitiativeEnterprises looking to begin data philanthropy can do the following:
- Catalog the data the enterprise gathers, creates, and analyzes, and rank different datasets by ease of sharing.
- Determine possible beneficiaries of data, in the same community or further afield
- Identify any showstoppers in terms of data privacy, either for the enterprise or for individuals
- Discuss with appropriate public sector entities to see which data and formats they prefer
- Apply fair business criteria to decide the best entities/projects with which to partner
- Figure out the best way (given current resources) of distributing its data
- Engage proper privacy protection and anonymization data (remember the possibility of “Real-Time Data Commons”) before releasing data
- Allow researchers to publish results from the data, possibly also making its data philanthropy conditional on their sharing of findings.