Benn Stancil, a co-founder of Mode and a prominent writer on data and analytics, is known for his sharp, often contrarian, and widely-read blog, "The Benn Stancil Substack." His essays dissect the realities of the modern data industry, challenging hype and offering a pragmatic perspective on the roles of data teams, the tools they use, and the cultures they operate in.

On the Modern Data Stack and Tooling

  1. "The modern data stack is a solution in search of a problem." Stancil argues that the explosion of tools in the modern data stack (MDS) was driven more by venture capital and market trends than by a fundamental shift in the core problems data teams face.
  2. "The dirty secret of the modern data stack is that it’s not for analysts." He contends that the MDS was built primarily for data engineers, often creating more complex workflows for the analysts who are meant to be its primary users.
  3. "For all the talk of being ‘modern,’ the stack has a decidedly dated feel." Stancil points out that the component-based architecture of the MDS resembles the on-premise systems of the past, just reconfigured for the cloud.
  4. "The modern data stack is a_ Rube Goldberg machine_ of BI." This quote colorfully illustrates his view that the MDS often involves overly complex, pieced-together solutions for tasks that were once more integrated.
  5. "We aren't building the next generation of data tools; we’re rebuilding the last one with new logos and new database connectors." A critique of the incremental, rather than revolutionary, nature of many new data products.
  6. "The modern data stack is built on a lie... that we, the data people, are the chosen ones, destined to lead our companies to a higher state of being." He challenges the self-important narrative sometimes found in the data community, advocating for a more humble and business-focused role.
  7. "The biggest lie the modern data stack ever told is that it was a stack at all. It's a puddle." This metaphor suggests a lack of cohesion and a sprawling, often messy collection of tools rather than a structured, interoperable "stack."
  8. On the "unbundling" of BI: The MDS unbundled the all-in-one BI tool into dozens of specialized products, promising best-in-class solutions but often delivering integration headaches and higher costs.
  9. "The problem with the modern data stack isn't that it's unbundled; it's that it's incoherent." The issue isn't just that the tools are separate, but that they don't work together in a logical, seamless way for the end user.
  10. "The next data stack will be a single, cohesive product." Stancil predicts a "great rebundling," where the fragmented tools of the MDS will be consolidated back into more integrated platforms that are easier for analysts to use.

On Data Teams, Analysts, and Their Role

  1. "Data teams aren't support staff. They're a product team." He advocates for data teams to operate like product teams, with their own roadmaps, priorities, and focus on delivering value to the business, rather than being a reactive "help desk" for data requests.
  2. "Stop being a service organization." Stancil urges data teams to move away from a model where they simply fulfill requests and instead proactively identify and solve business problems.
  3. "Analysts should be businesspeople first, and data people second." The most effective analysts are those who deeply understand the business context and use data to solve business problems, not just produce reports.
  4. "The job of a data analyst is not to answer questions, but to ask them." This highlights the importance of critical thinking and curiosity in the analyst role, pushing beyond simple query fulfillment.
  5. "The most valuable skill for a data analyst is not knowing SQL, but knowing the business." Technical skills are foundational, but true impact comes from applying those skills to what matters for the company.
  6. "Data teams are often caught in a 'tyranny of the urgent,' constantly firefighting and responding to ad-hoc requests, with little time for strategic work." A common pitfall that prevents data teams from delivering their full potential.
  7. "The ultimate goal of a data team is to make itself obsolete." By empowering the rest of the organization with the tools and skills to answer their own data questions, the data team can focus on more complex and strategic challenges.
  8. "Data analysts are the translators between the language of data and the language of the business." This emphasizes the crucial communication and storytelling aspect of the analyst's role.

On Data Culture and Decision-Making

  1. "Data-driven is a destination, not a state of being." Becoming data-driven is a continuous process of improvement, not a switch that can be flipped by buying new tools.
  2. "A company's data culture is not defined by its tools, but by its behaviors." A truly data-driven culture is about how people make decisions, not the software they use.
  3. "The goal of a data team is not to make every decision data-driven. It’s to help people make better decisions." Sometimes the best decision is not purely data-driven, and the data team's role is to inform, not dictate.
  4. "Self-serve analytics is a myth." The idea that business users will simply and effectively answer all their own complex questions with a BI tool is unrealistic. It requires guidance, training, and a strong partnership with the data team.
  5. "Dashboards are where data goes to die." Stancil is critical of the overuse of dashboards, which are often built, looked at once, and then ignored, creating "data graveyards."
  6. "Metrics don't matter if they don't change how you act." The purpose of measurement is to drive action and improve outcomes, not just to report numbers.
  7. "The CEO is the Chief Data Officer." The commitment to a data-driven culture must start from the very top of the organization to be successful.

On Business Intelligence (BI) and the Future

  1. "BI is not a technical problem; it’s a social one." The challenges of BI are less about technology and more about communication, collaboration, and aligning on what matters to the business.
  2. "The future of BI is not about more dashboards, but about more conversations." Effective data work is collaborative and iterative, involving a dialogue between the data team and business stakeholders.
  3. "AI will not replace analysts. It will replace the parts of their jobs they hate." He sees AI as a tool to automate tedious tasks, freeing up analysts to focus on more strategic and interpretive work.
  4. "The future of the data stack is one that is built for analysts, not engineers." The next wave of tools will prioritize the user experience of the people who are closest to the business problems.
  5. "We need to move from 'data-as-a-service' to 'data-as-a-product'." This involves building durable, well-maintained data assets that serve ongoing business needs, rather than one-off reports.
  6. "The metric layer is the Holy Grail that nobody can find." While the concept of a centralized, single source of truth for metrics is appealing, it has proven incredibly difficult to implement and maintain in practice.
  7. "The best BI tool is a conversation." This underscores his belief that direct communication and collaboration are often more effective than any piece of software.
  1. "The data world is high on its own supply." A critique of the industry's tendency to get caught up in its own hype and jargon, losing sight of the fundamental goal of solving business problems.
  2. On the hype around data catalogs: While useful, he argues they don't solve the core "last mile" problem of translating data into actionable business insights.
  3. "We're all just building spreadsheets, again and again." A reminder that many complex data tools are ultimately trying to replicate the flexibility and accessibility of the humble spreadsheet.
  4. "The 'data mesh' is a solution to an organizational problem, not a technical one." He sees the data mesh concept as a way to structure teams and responsibilities, which may not be necessary or appropriate for all companies.
  5. "Stop trying to be Google." Stancil advises smaller companies to avoid copying the complex data architectures of tech giants, as their problems and resources are vastly different.
  6. "The 'citizen analyst' is a unicorn." The idea of a business user who is also a skilled data analyst is rare, and data teams should not build their strategies around this exception.
  7. "We confuse 'can' with 'should'." Just because we can build a complex data pipeline or a detailed dashboard doesn't always mean we should. The focus should be on value and impact.

Additional Learnings and Insights

  1. The importance of "The Last Mile." The most critical, and often most difficult, part of analytics is the "last mile"—the process of turning a data insight into a concrete business action.
  2. Pragmatism over dogma. Stancil consistently advocates for a practical, results-oriented approach to data work, rejecting rigid adherence to buzzwords or trends.
  3. Focus on the "why," not just the "what." Understanding the business reason behind a data request is more important than simply fulfilling the request itself.
  4. Simplicity is a virtue. The simplest solution that solves the problem is often the best one.
  5. Data work is a craft. It requires not just technical skill, but also judgment, creativity, and a deep understanding of the medium (data) and the context (the business).
  6. Narrative and storytelling are key. The ability to weave data into a compelling narrative is what separates a good analyst from a great one.
  7. Embrace the messiness. Real-world data is often messy and incomplete, and analysts need to be comfortable working with ambiguity.
  8. The value of skepticism. A healthy dose of skepticism towards vendor claims, industry hype, and even one's own analysis is essential.
  9. Build partnerships, not dependencies. Empowering business users is more sustainable than making them dependent on the data team for every request.
  10. "Data is political." Data can be used to support different agendas, and analysts need to be aware of the organizational dynamics at play.
  11. "The job is never done." The work of a data team is a continuous cycle of asking, answering, and asking again, as the business evolves and new questions arise.