By some estimates, AI is set to add $13 trillion to world GDP by 2030. These results will largely be realized by the early adopters of AI with laggards losing business share.
AI’s ability to achieve this level of impact rests largely with access to high volumes of quality data. The most valuable companies in the world already know this. Microsoft, Amazon, Apple, Google, and Facebook are all data-driven companies.
In order to remain competitive, companies must view data as a strategic asset and leverage it as a differentiator. Aligned with business strategy, companies must articulate a cogent data strategy that outlines what data is fundamental to business operations. It is important for businesses to understand how they will obtain, maintain, and leverage data, cost effectively, to reduce costs through automation and increase revenue through generating business insights.
As implied earlier, AI is only as good as the data fueling it. Therefore, understanding how the business will obtain the right, quality data in sufficient volumes is foundational. To this end, there are three primary aspects to consider: data you have, data you can acquire, and data you can buy.
Most companies, unless launched the day before, do not achieve their present form overnight. They incrementally develop increasingly sophisticated processes and systems, often haphazardly, to run the business. As a result, the data businesses have, often becomes siloed in isolated departments or units where it cannot be effectively leveraged by the organization. To unwind this mess, companies must implement the policies and infrastructure required to drive accessibility to data that is critical for enabling business operations.
This can be an involved process, but the return on investment can be worth it. If properly executed, it will enable companies to understand their existing data inventory and what data they might need to achieve additional business benefit. Once the gaps in the data inventory are understood, companies can then move forward with acquiring or buying required data sets as appropriate.
Acquiring data can be accomplished through multiple mechanisms depending on the need. For example, publicly available sources can be scraped and integrated with internal data sets to drive incremental benefit. Natural language understanding tools can be used to transform unstructured data into tangible insights that improve business decisions. Machine vision solutions can also be used to acquire insights from the actions taken by customers and employees throughout every day. However, companies will often have to rely on buying proprietary information about their existing or potential customers that can’t easily by obtained through other means. For instance, it is possible today to purchase GDPR compliant data that gives you hundreds of data points each day about where people are physically located for over one billion people across the world – pretty much everybody that has a smart phone. This data can be used in a variety of ways that range from optimizing marketing expenditures to making data-driven decisions about individual lending or even capital investments in specific locations.
With the acquisition focus of the data strategy understood, companies can begin the process of extracting real value from their data. In this way, data becomes a true strategic asset empowering AI-enabled solutions solving specific business problems and generating real tangible results.
I’ll explore the other core components of data strategy – maintain and leverage – more fully in future blogs.