Building Blocks of the Digital Revolution: Data

In today’s digital age, which is fast-becoming defined by force-multiplying technological advances such as Artificial Intelligence (AI) and the latest machine learning techniques, the value of data and prevalence of enterprise-wide data literacy are redefining what it means for organizations to be “data driven.”

As organizations strive to embrace digitalization in attempt to broaden competitive moats and protect market positions, unlocking the full potential of data-supported decision making has become paramount to achieving material differentiation.

At the GIC’s Bridge Forum Summit, tech leaders and entrepreneurs including Barr Moses, Co-Founder & CEO of Monte Carlo, Barry McCardel, Co-Founder & CEO of Hex Technologies, Tristan Handy, Co-Founder & CEO of dbt Labs, and Christian Kleinerman, SVP of Product at Snowflake, discussed some of the characteristics they have found common to the most successful data-driven enterprises and the strategies those companies have adopted to harness the power of data to fuel sustained success.

A Customer-Centric Approach

Improved tooling, aggregation, and data observability technologies have led to an unprecedented proliferation of consumable data. From customer preferences and market trends to operational efficiencies and predictive analytics, organizations are now in possession of more valuable information than ever before. However, access to data alone does not equate to improved outcomes. To extract actionable insights and drive meaningful outcomes, businesses must adopt a customer-first mentality around data utilization to achieve differentiated value.

“The biggest gap for data teams is being far from customers. The number one question that we try to ask is how is this data used? And by who? And how is it servicing the business? The question seems trivial but it’s actually extremely hard to answer for many data teams. The path to unlocking value has to do with figuring out that answer.”

Focus on customer-centricity also requires that organizations think of “data as a product,” Moses explained. Product managers and data teams need to ensure that data assets are structured, organized, and properly distributed within companies like other, more traditional products in order to successfully meet the needs of customers, both internal and external.

Governing Data

Another critical aspect of proper data utilization and productization is ensuring an organization develops, and consistently revises, a robust data-governance operating model centered around data quality.

“You want to be able to empower your teams with the right data and the right tools to extract value from that data, but it’s equally imperative to keep security and governance top of mind. That’s a big part of the challenge that I think most organizations face today.”

Data must be reliable and trustworthy to instill confidence in decision-making processes. To achieve this, organizations can draw upon best practices from the field of software engineering, such as employing rigorous processes of review and staging before products are rolled out, rigorous A/B testing on different data methodologies, and “debugging” exercises aimed at uncovering latent inconsistencies in outputs.

“One of the trends or themes that I’m seeing teams adopting, that’s actually helpful and practical, is thinking of data as a product: having product managers for it, having requirements, defining the customer, prioritizing the governance around it, and data quality,” said Moses.

Kleinerman agreed, adding: “In the same way that you have processes and controls to build trustworthy software to meet security and compliance requirements, that’s how organizations should be thinking about data itself. And once you change the mindset, that data quality is not an afterthought of some department on the corner, it becomes a priority with a full lifecycle repeatable process.”

Unlocking Data Through AI

The rise of data driven technologies, powered by predictive systems and AI, will inspire organizational innovation and provide companies powerful opportunities to enhance their products and realize accelerated impact. Barry McCardel, Co-Founder & CEO of Hex Technologies, believes the availability of these tools will rapidly proliferate in the near future:

“I think in two years, AI is going to be coming out of your ears, and we’re going to see all the major cloud info providers and data platform providers in a race to make it as easy and as fast as possible to spin up large language models internally.”

Handy, Co-Founder and CEO at dbt Labs, agreed, adding that the most value is likely to accrue to organizations who develop specificity in their incorporation of AI technology: “I think that the thing that is going to unlock all of this is the ability to take these large models and shrink them down for specific contexts. Your model needs to have purpose, just as individual employees have purpose.”

At the end of the day, businesses need to figure out which AI models have the best capabilities for their data needs and which can most empower their people to focus on what’s important. McCardel summarized: “I’m very optimistic about AI. And I’m optimistic about in the domain of data to liberate humans, and especially data teams, to allow them to focus on those really high-leverage, creative, uniquely human tasks, so long as they remain focused.”

Importantly, these models should be built with the right data infrastructure and people-support strategies in place, according to Handy: “While customized, fine-tuned large language models are important, you need to be able to ship the data to those models to make sure that you can do those fine tunings. Therefore, the people and technologies involved in creating those high reliability, high performance pipelines will be increasingly important.”