Building a Culture of Data Trust | Wabi Sabi Your Data

How I Watched $150K Slip Through the Cracks — And What I Built to Stop It Happening Again

Hello, data Shokunin-deshi!

Welcome to the final chapter in our special series on data-driven culture. I’m Lior, your guide in this journey of data enlightenment. This edition is packed with practical insights and hard-earned lessons, designed especially for leaders who need their data to be trusted fast.

Whether you're building data products, setting governance standards, or are simply tired of hearing “I don’t trust the data,” this one’s for you.

Being data-driven goes beyond saying, "We are Data Driven." It's about the practices you put in place, how you leverage data, and, above all, creating impact. In our last newsletter, we discussed Data ROI as the key to proving data is an impact creation tool. Today, we'll explore how to establish trust in data—the number one killer of Data ROI when absent.

---

What we will learn today:

In this culminating edition, we explore how to build a foundation of trust that transforms your data from questionable to indispensable:

• Why trust forms the foundation of truly data-driven organizations

• The essential elements of effective data governance

• Strategies for overcoming common trust barriers

• Steps for creating a culture of constructive questioning

---

Wabi Sabi Your Data: The Zen of Data Trust

"Trust is like a garden; it requires regular tending and cannot be rushed into bloom overnight."

It reminds me of last year when we tried something new: setting a scheduled time for watering our tomatoes in the greenhouse. The pleasant surprise was the great yield, but simultaneously, some tomatoes developed diseases from overwatering.

In the Zen tradition, two concepts are particularly relevant to data trust:

Shoshin (初心): The "beginner's mind" approach encourages openness, eagerness, and lack of preconceptions. In data governance, this means approaching each data challenge with fresh eyes, questioning our assumptions, and being willing to learn.

Ichigo Ichie (一期一会): This concept translates to "one time, one meeting" and emphasizes the uniqueness of each moment. For data professionals, it reminds us that each data interaction is a unique opportunity to build trust—and once trust is damaged, that moment cannot be recaptured.

In traditional Japanese tea ceremonies, participants trust the tea master implicitly. This trust isn't blind—it's earned through the master's meticulous preparation, attention to detail, and years of practice. Similarly, in the realm of data, trust isn't granted automatically but is cultivated through careful governance, consistent quality, and transparent processes.

Let’s ground this in something deeper—what ancient wisdom can teach us about modern data trust?

Wabi-Sabi teaches us to embrace imperfection—not as failure, but as a space for improvement.
In data, perfection is a myth. But transparency, iteration, and continuous care build trust over time. Just like a cracked teacup becomes more beautiful through its wear, trusted data systems earn their credibility through scars, scrutiny, and refinement.

The Cost of Distrust: A Personal Journey

Let me share a story from my early days at a research company. We supplied critical market data to clients who made multimillion-dollar decisions based on our reports. Account managers would check the data before sending it to clients, but as analysts, we were disconnected from how clients used our insights and whether they trusted them. From time to time, we would receive a request from the account manager to validate the numbers, but we couldn't QA it properly as the data was protected in a special file, which required reaching out to another team. This separation created a dangerous gap—we weren't witnessing firsthand the implications of data errors.

Years later, at an app publisher, I experienced a transformative moment. Sitting directly with marketing teams, I watched their faces fall when they discovered inconsistencies in dashboards I'd built. Seeing their frustration—and the cascading impact on decisions—changed everything for me. I realized that when data trust breaks, it's not just numbers that suffer but real business outcomes and relationships.

M, one of my most data-savvy marketers, started every morning the same way: exporting my dashboard to Excel and re-validating every single number. That was the only way she could trust it.

She spent one to three hours a day on this ritual. That wasn’t just wasted time—it was a quiet vote of no confidence. And when the data eventually led her to the wrong decision, the cost was real. Her team lost budget. She lost face. We lost credibility.

So I started coming in early—before the marketers arrived—to check and validate the data myself. I built my tracking sheets to spot anomalies. Later, together with a data engineer, we created an automated alert system that emailed users when something was off.

But many missed the emails.

This taught me a simple, painful truth: data trust is fragile.
It takes months to build and seconds to break.
Once a CFO spots an error in a financial report, their trust doesn’t just dip—it can disappear for quarters.

Recipe for Success: Building a Foundation of Data Trust

Building a Culture of Data Trust: The Five Pillars of Data Trust

Want to build trust fast? Start with these five pillars
You likely know them, but are they implemented? Operationalized? Visible to your users?

Now, much of data trust is built through proper governance, but in today's centralized and decentralized world, we face issues that cause loss of trust and are not always within the data team's control. We should work to create better transparency and control over data to reduce these frictions.

I'm sure you're already familiar with or implementing many of the items below, but sometimes it's important to revisit the fundamentals:

1. Transparent Data Lineage:

   • Create clear visibility of where data originates

   • Document how data transforms at each stage—yes, from raw data to the dashboard itself, every step should be documented and available

   • Enable users to trace insights back to their source

2. Proactive Quality Assurance:

   • Implement automated data validation checks throughout your pipeline

   • Establish simple and pragmatic "traffic light" systems that signal data health (green = trustworthy, yellow = requires investigation, red = do not use)

   • Communicate data issues before users discover them

3. Incident Response Protocol:

   • Develop a clear process for handling data quality incidents; try to identify the level of incident, the circles that should be involved, the communication and documentation, as well as potential damages and extra costs

   • Maintain transparent communication during resolution

   • Conduct thorough post-mortems to prevent recurrence

4. Trust Through Governance:

   • Clearly define data ownership and accountability

   • Establish data standards and certification processes

   • Create documentation that builds understanding, not just compliance

5. Culture of Constructive Questioning:

   • Encourage appropriate skepticism about data

   • Focus questioning on the process, not the people

   • Celebrate when questioning leads to improvements

I would even add a sixth pillar called “Data Feedback Loops”

• Draft guidelines for communication handeling

• Establish regular feedback from users to the data team to identify and resolve trust issues.

• Establish communication channels and methods

One story from my past about ownership and trust: an organization I joined was deciding to move toward data decentralization, following aspects of the data mesh framework without fully applying it. Some product teams and product owners I interviewed explained that one day, they discovered they now needed to own the data they produced and ensure its correctness, but they lacked the knowledge and skills.

They mentioned receiving a template to use for ingesting data into the data lake. One product manager shared that during an update, an engineer overlooked the need to adjust the template, so for 5 months, they didn't send any data to the data lake—and no one noticed. I said, "Great, it means no one needed the data.", we can save here easily 2K a month, but he replied, "No, they just also skipped it, and no one noticed until the CFO realized the numbers didn't make sense."

When I investigated further, fixing this issue cost the organization more than 70K in reruns and another 150K in lost opportunities, plus a very confused team wondering why the data was so important but no one spotted the issue for so long.

Let’s zoom into a practical example of how trust can be built into daily workflows.

Case Study: The Morning Confidence Dashboard

At a mid-sized e-commerce company I worked with, we implemented what we called the "Morning Confidence Dashboard." Before business users accessed their analytics each day, this special dashboard would run, showing:

• Data freshness (when was the last successful update?)

• Pipeline health (were all scheduled processes completed successfully?)

• Anomaly detection results (were there unexpected patterns or outliers?)

• Recent changes documentation (what modifications might impact interpretation?)

This simple addition transformed the organization's relationship with data. Decision-makers started their day with appropriate context, knowing whether they could trust what they were seeing. The number of "Is this data correct?" support tickets plummeted by 78% within three months.

I'm always concerned about my data consumers because their job isn't to QA the data—their job is to consume it and perform their assigned tasks, whether in controlling, marketing, or product. Their job is to look at the numbers, and our new tool allowed them to have more trust and, worst case, someone or something to blame for the use of bad data. We kept finding issues and adding them to our checks, but this pragmatic solution provided a boost of trust.

Most importantly, when issues did arise, they were discovered proactively by the data team rather than by confused business users in the middle of critical decision meetings.

“Since using the Morning Confidence Dashboard, we’ve cut data issue escalations in half and increase data trust score by 40 points.” — App Publisher company

Reflection:

Ask your team—or yourself:

1. How do we currently signal that data is safe to use?

2. Where have trust breakdowns occurred in the past quarter?

3. What’s the cost of that mistrust (time, budget, decisions)?

💬 Want a second set of eyes on your answers? Hit reply, and let’s talk.

---

🔒 For Premium Readers: From Insight to Action

🔒 Premium Only: Data Trust Toolkit
This week’s premium content includes:

• A Data Trust Audit you can run with your team

• A template for incident response

• A 30-day plan to boost trust signals

👉 Not yet a premium member?