Bridging the Analytics Divide
The ability to access and analyse data is unevenly distributed. Bridging the analytics divide to encourage innovation and empower independent businesses will be one of the major challenges of the next decade. Data trusts — models for sharing sensitive commercial information among competitors — may prove to be a useful part of the solution.
We now live in a society that is both rich in data and equipped with tools that can extract useful information from that data.
The past decade has brought major advances in data collection — distributed, connected sensors are everywhere, turning ordinary interactions and activities into data points. At the same time, rapid advances in machine learning have given us the tools to mine these data points to generate new insights of great value.
For businesses operating in this data rich environment, enormous new opportunities for value creation have been unlocked. The patterns that algorithms and machine learning systems identify in large data sets can be used to develop targeted advertising campaigns and personalised recommendation services; to enhance marketing strategies; to optimise products and services; to reduce operational inefficiencies; and to support product and business model innovation.
Randy Bean, CEO of NewVantage Partners, business and technology consultants to Fortune 1000 companies predicts, “the next wave of big data will be all about leveraging the power of AI and machine learning to deliver business value at scale.”
But which organisations are actually primed to take advantage of data and machine learning opportunities?
It’s easy to find examples of large corporations using advanced analytics. General Electric uses advanced analytics to manage pipeline risks — combining weather, dig-reporting, seismic, geospatial and site inspection data. Bridgestone, the multinational tire manufacturing and sales corporation, employs data analytics to determine store locations, to develop more efficient processes for inventory and human resource management, and to predict customer service requirements. Capital One Financial Corporation uses data analytics to predict which customers are most likely to sign up for new credit cards and repay debt.
Certainly, these are notable examples of advanced analytics at work in the corporate setting. But they are also examples of large multinational corporations with deep pockets and significant technical expertise.
For the majority of businesses — independent, small and medium sized operations — the benefits of advanced data analytics remain out of reach.
The Analytics Divide
To extract useful knowledge from data, a business must have access to sufficiently large data sets and the resources required to analyse that data. A minority of businesses possess both. There is an analytics divide.
The analytics divide is clearly apparent in markets such as internet search, social media and retail. Companies like Google, Facebook, Amazon and Walmart privately own and have exclusive access to exceptionally large and self-perpetuating data sets from which they can analyse user or customer behaviour and make strategic decisions that continue to expand and embed their position in their respective markets.
But in a digitally connected society, an analytics divide has the potential to arise in any sector.
A 2018 survey found 97% of the sixty Fortune 1000 companies that were surveyed are investing in big data or AI data analytics projects. These companies included American Express, Bank of China, Bloomberg, Capital One, Citigroup, Credit Suisse, Morgan Stanley, JP Morgan, Goldman Sachs, Freddie Mac, Fannie Mae, VISA, Wells Frgo, Astellas, OptumLabs, Estee Lauder, Ford Motors, IBM, LinkedIn, Motorola and Verizon.
As these large corporations continue to amass data and develop advanced systems of analysis, there is an increased risk across all sectors of the global economy that opportunities for innovation and growth will be unequally distributed.
Data has become an extremely valuable commodity. There are thriving markets in selling data and analytics services in some sectors. Monsanto sells analytics services to farmers. Google sells website usage data through Google Analytics. Twitter sells to application developers access to its data.
Unfortunately, in many sectors, companies with access to vast stores of data will often restrict access. Without data, new startups can’t enter the market. Without good analytics, competitors aren’t as competitive. There are powerful incentives for businesses that collect data to protect their competitive advantage by trying to restrict what others can do. Large, vertically integrated companies also have access to data along each stage of their supply chains that are not as easily available to independent businesses.
Bridging the Analytics Divide — A Collective Action Approach
The Analytics Divide hurts small and independent businesses the most.
Independent businesses only have access to data they generate in their operations — they often don’t have a good view of the entire market, and they can’t easily benchmark against their competitors or predict future conditions.
Without good data, independent businesses cannot effectively compete with large, industrial-scale companies that have access to much larger and broader datasets.
This is where data sharing arrangements can help. When independent businesses band together to share their data, they can all benefit. In an age of advanced analytics, the key to encouraging innovation and making independent businesses more competitive lies in finding ways for competitors to share access to industry data.
How Can Competitors Share Data? Three Simple Rules for Successful Data Trusts
Sustainable industry data sharing is a collective action challenge. It requires encouraging a diverse range of actors to share commercially sensitive information about their assets and operations.
Our research suggests a knowledge commons model for data sharing can support sustainable industry data sharing. Under a knowledge commons model, or what we call a data trust, participants agree to share standardised business data and an independent organisation compiles this data and provides regular analytics to members.
Under a data trust arrangement, participants that share data create and control the resource. The rules of access are governed according to mutually agreed principles. Together, participants create a new resource that is owned and shared for mutual benefit, even if they are competitors in the marketplace.
Our model for a successful data trust is underpinned by three core principles: fairness, usefulness and security.
- FAIRNESS. All trust members should have equal access to aggregated data and analysis. Limitations on data access must ensure that the data trust will not compromise the commercial interests of its participants.
- USEFULNESS. A successful data trust must create value for members by collecting, analysing and sharing industry data to support strategic decision-making. Often, by pooling resources, participants can unlock new opportunities that are out of reach of individual businesses.
- SECURITY. The trust must ensure the security of information that is commercial in confidence, and must take steps to ensure the privacy of any personal information that is collected.
If carefully designed to meet the needs of participants, data trusts offer a solution to some of the biggest challenges of the analytics divide. Data trusts promise to provide an important model for sustainable industry data sharing that can to help under-resourced businesses and sectors take advantage of advanced data analytics.
Of course, it first requires that businesses be bold enough to act cooperatively.
All images Pexels (CC)