Search the term “data-driven enterprise” and you’ll get close to 300 million results. Why? Executives like the sound of it. They’ll even tell you their companies use data to create value – whatever that means. Here’s the truth: data lacks value until it births expertise.

Helping human analysts transform data into expertise that’s efficiently applied and widely shared is massively important. The cost of failing to do so is too high and may be rising at a staggering rate. We explain why in this blog post and offer forward-thinking managers a way to get ahead of the problem.

Data workers are at the heart of the issue. We need more of them because their work has never been more important. And yet, despite the platitudes, executives appear to be massively underinvesting in tools and talent to turn data into expertise.

Consider: of the 24,000-plus U.S. businesses known to employ at least 1,000 people as of this writing, an estimated 75,000 workers at these companies perform data analysis as part of their core job function. That is just 3 data workers per 1,000-plus-person company. That’s a lot of potential insight concentrated into a precious few experts. For leaders intent on building a “data-driven enterprise,” it’s a structural problem that needs solving.

Why ambitious data leaders come up short

Expertise is a human trait developed with effort over time. Creating more of it requires either more humans, smarter systems to find patterns in data, or best of all, humans and machines working in concert. In too many cases, leaders choose neither and instead simply ask their analysts to do more. That’s a recipe for burnout and attrition.

At the very least, it can lead analysts to conclude that they’ve no hope of growing in their current roles – they’re too busy trying to not drown in data requests. But they won’t stay stuck for long. A 2018 study by Global Talent Monitor found that 40% of departing employees cited a “lack of future career development” as a motivating factor in their departing. A culture that breeds burnout and turnover among their analysts will always have a difficult time developing and applying expertise.

High-turnover businesses also waste enormous sums. Data workers in the U.S. tend to make between $60,000 and $150,000 in annual salary. Losing them can cost anywhere from 33% to 75% of their yearly compensation, according to multiple studies. That’s a median cost of about $57,000 per analyst that leaves. Leaders can end up spending millions hiring and then losing experts before they develop any new expertise.

The challenge of developing expertise at scale

Human limits help to explain why the “data-driven enterprise” is still largely a myth. Organizations that can most benefit from data – specifically, the 1,000-plus person enterprises – appear to be having trouble finding and keeping the talent needed to gather, review, organize, and eventually analyze the mountains of information in their possession. Most simply lack the resources to uncover profitable insights and then share those findings (i.e., their hard-won expertise) efficiently.

But even large teams of data analysts are likely to struggle to make meaning from data sets. Why? The world is awash in information. At the dawn of 2020, the World Economic Forum estimated the world’s data at 44 zettabytes. That equates to 44,000 bytes taken to the seventh power – or, 40 times bigger than the stars in the observable universe. Surely the known data universe has continued to grow exponentially in the years since that report was published.

Investing in systems to help analysts become more efficient is the only choice for today’s aspiring data-driven enterprises. In doing so, they allow the precious few experts they’ve hired to spend the bulk of their time on forensic study of information that falls outside what’s expected. This process – human analysts carefully studying  machine-curated information to spot profitable patterns – is the essential foundation for developing expertise.

The best path to insight also avoids burnout

To be fair, truly data-driven enterprises are only now emerging. These are the companies, government agencies, and other organizations that cultivate institutional expertise into a knowledge asset.

Think of banks investing to improve anti-money laundering practices or consulting firms serving a multinational client base. Turning data into insight fast is crucial for these businesses; applying expertise helps them to narrow the range of data worth studying. Knowledge assets automate how this expertise gets applied, usually through the use of an emerging toolkit organized in a composable platform that combines the descriptive power of knowledge graphs with the processing capability of A.I. services.

These fully integrated environments have two primary components: knowledge graphs that put data in context by defining the relationships within, and A.I. systems that can perform repetitive tasks at machine speed. Composable Graph + A.I. Platforms apply the expertise codified in knowledge graphs to train A.I. systems to surface patterns that might never be found through a purely manual search.

In doing so they make analysts far more efficient and allow them to do their best, most creative work and make real the mythical data-driven enterprise we’ve long heard about. It’s about time.

Get Started for FREE – Schedule a Solutioning Session

GraphGrid is a Composable Graph + A.I. Platform that helps organizations deliver advanced analytic solutions using graph and A.I. in weeks instead of months or even years.

Schedule a free solutioning session with a GraphGrid professional. Tell us about your highest impact data-intensive challenge or opportunity so we can align our platform and professionals to accelerate your graph-enhanced solution journey.

In the meantime, feel free to download GraphGrid at our Freemium license level and check out our step-by-step tutorials, video tutorials, or read more about GraphGrid in our use cases, blogs, and documentation.