The Future of Enterprise Data: Insights from a Global Data Leader

In a recent conversation with Sally Bashuan, Head of Global Data Governance at Federated Hermes, we explored the evolving landscape of data governance and the challenges organisations face in the age of AI. As one of the largest investment managers globally, Federated Hermes manages over $800.5 billion in assets and advises on $2.1 trillion, giving them unique insight into the complexities of data management at scale. Her insights revealed fascinating contrasts between UK and US data markets, while highlighting universal challenges in data democratisation and governance.

The Trans-Atlantic Data Divide

One of the most striking revelations from our discussion was the difference between US and UK data approaches in regulated sectors. In the United States, dedicated access persons and control environment puts the brakes on general democratisation to a certain extent potentially making it challenging for engineers and analysts to foster data literacy across their organisations or at least makes it more complex to implement. In contrast, the UK market tends to be more liberal with access to data within a company and the guide-rails of confidentiality and GDPR of course. Data literacy programs are required for both regions albeit they may have a slightly different focus to allow for the variance in regulatory environments. International companies of course require individuals on both sides of the pond to understand each other's requirements, so training on GDPR in the States is not uncommon. Globalisation of regulation would certainly help companies in this respect.

Beyond Regional Differences: Industry-Specific Challenges

However, there are other potential challenges to data democratisation rather than just governance and regulation - traditional sectors, such as real estate still grapple with digital transformation challenges where the variety of data and differences in maturity cause additional complexity. It is not just traditional financial data but also property level and ESG (Environmental, Social and Governance). Data collection is completed in multiple places through the real estate management process and by a variety of people making quality issues. All of which aptly demonstrates that all data is not equal.

Sandboxing and Governance: A Delicate Balance

A recurring theme in our conversation was the critical need for sophisticated data sandboxing capabilities. Data analysts and engineers frequently require separate environments to experiment with and manipulate data. However, this creates a significant challenge: how do you ensure that these professionals don't inadvertently violate data usage restrictions during their exploration and analysis?

This challenge points to a crucial gap in current data infrastructure: the need for built-in governance tools that can preemptively guide engineers and analysts about data usage limitations. Rather than discovering restrictions after significant work has been invested, these tools could help teams work more efficiently while maintaining compliance from the start.

Democratising Data Through Internal Marketplaces

Another fascinating concept that emerged was the potential for internal data marketplaces. These platforms could enable different departments within an organisation to access and share data more effectively, breaking down traditional silos that often hamper innovation and insight generation. This approach to data democratisation could transform how organisations leverage their data assets across teams and departments.

The AI Adoption Paradox

Perhaps most intriguing was the discussion about AI adoption in large enterprises. Counter to what many might expect, bigger companies are often slower to embrace AI technologies. The primary reason? Copyright and intellectual property concerns. These organisations struggle to differentiate between the features that are explicitly using AI within SaaS platforms, creating hesitancy in adoption.

The Power of Community in Technology Adoption

A particularly valuable insight emerged regarding technology adoption in enterprise settings. Traditional cold outreach methods are becoming less effective, especially for data and AI solutions. Instead, products backed by strong communities - particularly those comprising data analysts, engineers, and AI experts - tend to gain more traction. This suggests a shift in how enterprise software needs to be developed and marketed, with community building becoming as crucial as the technology itself.

Looking Ahead

As we continue developing Petals, these insights are proving invaluable. We're already working on implementing built-in governance and privacy controls within our data warehouse and sandbox environments. Our goal is to create a solution that not only makes data more accessible but does so in a way that respects and enforces necessary restrictions and regulations.

The future of enterprise data management clearly lies in creating tools that balance accessibility with governance, innovation with compliance, and technological capability with practical usability. As we move forward, these insights will help shape not just our product development but also our approach to building and engaging with our community of users and experts.