A savvy marketer today is probably as technically advanced as a data analyst was 20 years ago

Despite championing the “data driven” approach - a lot of orthodoxies developed by the data community fall flat. Boris Jabes, CEO of Census, a company specializing in reverse ETL, offers a fresh perspective on how businesses should approach data utilization and integration.

The Illusion of Data Readiness

One of the biggest misconceptions Jabes encounters is the belief that data must be perfectly clean before it can be used effectively.

“‘My data’s not ready; we still need to clean it, manage it.’ I hear that from data teams all the time,” Jabes says, mimicking a familiar refrain. “Yes, governance & quality are important. But often, it’s more valuable to deploy something half-baked and get a feedback loop going — unless there are significant downsides.”

This approach prioritizes action over perfection, encouraging companies to start using their data even if it's not in an ideal state. The key, according to Jabes, is to tighten feedback loops and focus on rapid deployment and adjustment. 

“If you can speed up how fast new information gets deployed and acted on, you correct faster. And the company that corrects faster is the one that wins.”

The Shifting Role of Data Professionals

Jabes also sees a profound shift in the role of data professionals. The technical skills that once defined data work are becoming table stakes. 

"Joining 16 tables together in a complex query is still impressive but not as big a competitive advantage." Instead, the focus is shifting towards understanding what should be queried and why.

This change is driven by the increasing technical literacy across various business functions. "A savvy marketer today is probably as technically advanced as a data analyst was 20 years ago," Jabes notes. 

This shift is pushing data professionals to evolve. “The real value is moving beyond data wrangling and toward machine learning, advanced analytics, and actionable insights,” Jabes says. “It’s not about writing ETL by hand, but about understanding which insights drive the business forward.”

The Paradox of Infinite Choice

With the rise of modern data tools, companies have more power than ever to analyze their data. But with that power comes a new problem: decision paralysis.

Jabes likens the experience to streaming music. “We went from owning a limited CD collection to having all music available at our fingertips. And yet, we open Spotify (or Netflix) and feel paralyzed by choice.”

The same holds true in data analytics. Just because businesses can store & compute nearly anything doesn’t mean they know what’s worth computing.

“The challenge today often isn’t technical - it’s knowing which questions to ask and which metrics to care about. The real problem is curation to make useful decisions.”

Beyond Structured Data

While Census primarily deals with structured data, Jabes acknowledges the growing importance of unstructured data in business operations. "We've all learned that there's way more of it than we'd all been thinking about," he says. This realization is pushing companies like Census to explore new frontiers in data integration.

Jabes envisions a future where the interaction between different software systems is mediated by natural language instructions rather than rigid data structures. "The most exciting frontier for us to explore is building and interacting with software systems through natural language instructions - in English, not SQL or API calls.” 

“Imagine saying, ‘Tell me which open issues in Salesforce have been fixed in our app,’ without having to write a complex query or sync entire datasets between GitHub and Salesforce.”

This shift could fundamentally change how businesses interact with their data, making it more accessible to users and reducing the time spent on data wrangling.

Ultimately, Jabes believes the future of data management lies in balancing accuracy with speed. Companies that insist on perfect data before acting risk falling behind. Instead, success will come to those who embrace imperfection, tighten feedback loops, and prioritize actionable insights over endless data refinement.

“It’s not about having perfect data. It’s about having useful data,” Jabes says. “The companies that can act on insights faster - even if those insights are imperfect - will be the ones that win.”

As businesses navigate this evolving data landscape, the message is clear: Stop waiting for perfection. Start making decisions. And learn by doing.