We in the international development sector often have this perception that the private sector is way more advanced than we in the collection, analysis, and use of user data.
That somehow Facebook, Google, and every Mobile Network Operator has a God-like dashboard that instantly displays every data point and trend imaginable, at the touch of a button – or now with artificial intelligence, even before the well stock-optioned operator realizes they need it.
The Private Sector Reality May Surprise You
Our private sector compatriots are just as overworked as we are, often with the same basic tools and incompatible databases. That was a key surprise at the Technology Salon on How to Access Private Sector Data for Public Goods?
Unlike us, private sector staff also have a overriding mandate to create a profitable return on everything they do, which means they rarely can put mission first, like we do, and need to justify any non-revenue generating activity.
These triple realities of high workloads, low resources, and constant profitability pressures means that we need to rethink how we approach private sector players to use their data for public goods.
Step 1: Get Your Own House in Order
Holly Krambeck was very clear in her main learning from working with multiple transportation companies – we need to organize our own organization’s data needs before we even think about approaching private sector companies to ask for their data.
This includes several aspects of demand validation and segmentation, including:
- Appointing a point person or team to champion the potential for private sector data and lead the data sharing processes
- Standardizing the forms and processes across the organization, spanning everything from how someone can initiate a data ask, to how the data can be used and when it needs to be disposed and how.
- Organizing internal demand so that there is a coordinated ask from private sector players and coordinated usage across your organization.
- Developing an internal support ecosystem so that program staff, who may not have deep statistical or technology expertise, know where and how to find it to effectively use the data that is received
- Sorting out all the legal issues involved, which can range from the level of anonymization required, to GDPR and other data law compliance, to the level of sharing that’s possible internally, and if it possible, externally too.
Step 2: Strategically Approach the Private Sector
Laura McGorman was equally clear about how private sector organizations need to be approached if you want a successful partnership. We need to have mutual respect – for their efforts and their major risk in sharing any data at all.
Respect Their Efforts
Data costs money to collect, analyze, and yes, even share, yet we often have the belief that we should get things for cheap or even free, because we are mission-driven organizations. However, in addition to cost, there is great value in data, which is why there already are multiple commercial data sharing companies that will gladly share data with us – for a fee.
So if you are asking for data, start by respecting the efforts that private sector companies are investing in their data. Be prepared to intelligently ask for the exact data you need, and expect that you’ll need to do the heavy data analysis work. Also, free is rarely sustainable, so expect that if you do find value in the data you ask for, you should expect to pay for it in the future.
Respect Their Risk
As the ongoing Facebook data sharing drama shows, private sector companies are put at a great risk when they share data outside their organization. Cambridge Analytica based their Facebook interventions on data that came from a seemingly harmless academic researcher request, not a malicious hacker.
So if you are asking for data, be sure to respect the private sector company’s need for certain safeguards and legal requirements. In fact, shouldn’t we be even more conservative with others’ data, as mission driven organizations? Therefore, think through the data sharing risks for both organizations – legal and moral – and make sure you can proactively assuage their worries.
Step 3. Consider Their Competitive Ecosystems
Rachel Sibande clearly pointed out the unique position of mobile network operators in the countries where we work. As detailed in the DIAL report on Unlocking MNO data to enhance public services and humanitarian efforts, MNOs are faced with a flood of vague data requests, often have the weakest internal capacity, are operating under national regulators with shifting policy objectives, and are deadlocked in competition with other MNOs in their markets.
We Often Are Ambiguous in Our Needs
Governments and NGOs often don’t know what data MNOs have (or don’t have) and how to ask for data that can realize cross-sectoral benefits. They are also very reluctant to pay MNOs for the data or the analytical services to exploit the data they do get. That doesn’t stop them from continuously, and duplicatively asking MNOS for data to the point of MNO exhaustion.
They May Not Have Internal Capacity
At the same time, MNOs are mainly focused on customer acquisition, and MNOs in low-income markets don’t usually have analytical capabilities needed to develop data sharing products that are useful to mission driven organizations. MNOs also have trouble pricing their data appropriately for humanitarian and development sectors, especially since the latter often expect data to be given freely, and without restriction.
They May Be Wary of Regulators
Government regulators can be a strong deterrent to data sharing by MNOs due to justified citizen privacy, security, and sovereignty concerns. Also, data sharing could lead regulators to ask for or restrict datasets even more than we would think appropriate. While we may focus on restrivie new laws around online speech in Tanzania and Uganda, the new European GDPR law can be seen as even more detrimental to effective data sharing.
They Are Certainly Wary of Their Competitors
We sometimes forget that a company’s data is even more interesting to its competitors than to us, and something as seemingly beneficial as rush hour traffic flows or community usage levels can be a critical competitive advantage to a private sector company. We certainly see this playing out in the USAID ecosystem where few (if any) implementing partner willing shares constituent data with other IPs.
Step 4: Getting the Data Is Just the First Step
The Technology Salon conversation did have an overriding conclusion: getting a data sharing agreement in place, and receiving private sector data is just the first step in a greater data utilization journey.
Collected but unused data is a waste of everyone’s time and resources – and yet we all already have countless databases that are untouched, or at least underutilized. Adding private sector data, that might be magnitudes larger, will not magically create greater understanding or better programs.
We need to be agile data analysis of all our own data – big and small – before we can expect to be effective data stewards of private sector data, and create public data goods that the world can use.