Becoming Customer Zero: A Journey to Scalable Data Products

Sigma Software’s Data Center of Excellence (CoE) was launched over 15 years ago and built on the same principles that shaped our AI, Information Security, and other CoEs over the past two decades – the strategy many now associate with the “Customer Zero” concept. At its core, it means that any product, capability, or service intended for clients must be adopted internally and prove valuable to our own teams first.

Thus, we relied on the Data CoE Business Intelligence capability to collect, monitor, and analyze data on our operations, supporting both day-to-day and strategic decisions with clear, data-driven insights. Those insights depend on high-quality data in the first place, and that’s where we ran into the same obstacles many companies face today. The IT department, typically responsible for early-stage data governance, identifies challenges that are not always easy to turn into opportunities:

• While data is being collected across the organization, managing it becomes increasingly expensive.
• The use cases – and even more so, the business cases – for data are not clearly defined. There’s often no compelling short- or long-term understanding of how data will return the investment.
• Data is not unified or harmonized.
• There are data duplicates and weak governance, leading to inefficiencies and escalating costs.
• Business intelligence and data analytics efforts are siloed and typically driven by the requests of individual departments.

Further, we’ll walk you through the practices that helped us overcome these challenges, turning internal capabilities into services now used by our clients. These insights are especially relevant for Heads of IT, CIOs, and COOs looking to shift data from a cost center to a source of business value.

First Data Product

People generally dislike building business cases within the internal IT. It’s hard to convince anyone that moving from Excel to a structured data storage and online analytics tools is worth the investment. How do you justify such an investment when it might take a team of costly engineers working full-time for six months just to get the analytics up and running?

To be completely honest, our first data analytics platform, which was a foundation for data products and business cases, as well as the investment behind it, wasn’t a complete success. It was fully rewritten within five years of the initial pilot because it simply could not scale and adapt to the changing needs of the fast-growing organization. However, the product we launched proved value to the stakeholders, which was the key outcome.

Looking back, some unexpected elements from those early pilots turned out to be even more valuable than the outcome itself:

1. Two major success factors for any future data product:
   a) A solid business case
   b) Stakeholders’ commitment
   Get these right first – everything else will follow.
2. A business case isn’t just an artifact – it’s a mindset.
   Start with building that mindset through training and cultural change. Don’t expect business cases to appear on their own. Talk to people about what a business case is. Run workshops, seminars, even hackathons. If you simply wait for business cases to come to you, people will eventually abandon the idea of a data product altogether. It just feels too hard to prove its value.
3. Presentation matters.
   The way your analytics looks and feels plays a critical role in gaining the buy-in. The devil is in the details of how data is visualized and how intuitive the tools around it are. It doesn’t matter if the data is static or backed by a complex pipeline. What matters is that it looks professional, it is easy to interpret, and it earns trust. Once you show analytics saves money or generates revenue, and once people enjoy using it, you’ve earned the right to go deeper and invest in the full architecture.

“At the end of the day, the technical platform can always be rebuilt. What really matters is educating and aligning people around value, commitment, and credibility. Right from the start” – emphasizes Anatoliy Kochetov, COO at Sigma Software – “this is a key foundation and the best investment you can make early”.

Design Thinking, Open Source & PoC

Our early business cases focused on cost optimization. One of the first successful data products we built involved tracking office space utilization to identify savings opportunities well ahead of the lease renewals. Even with static data, the initiative delivered at least $200K in annual savings.

While the savings forecast was promising and the business value clear, it was still critical to ensure that collecting and processing this data wouldn’t cost twice as much as it saved. That’s why we applied two key principles for our data products development:

Cost-Effective Proof of Concepts (PoCs)

Even though PoC became more or less a buzzword in the world of data and especially AI, it didn’t lose its true sense and actuality. We piloted different approaches, but the one that worked best for us at the early stages was focusing on building PoCs on top of open-source tools and affordable commercial solutions. This allowed us to validate concepts quickly without committing to heavy investments.

Applying Design Thinking in Data Platform Development

We didn’t apply design thinking in our early attempts at building a data platform and products. We focused too much on technology and tools, going all the way from creating a dedicated data warehouse to purchasing commercial analytics tools. Many of the solutions were implemented simply because they were considered state-of-the-art, had worked elsewhere, or came recommended by well-known consultants. And, perhaps, they worked to some extent for that time. The only problem was that we were looking for the right tools without fully shaping the concept behind why we actually needed them.

The key takeaway is that design thinking applies to most, if not all, aspects of Data Platform and new Data Product design, and it inevitably pays off with better product fit and fewer re-dos.

The next time around, we applied design thinking from the start, shaping a new approach we’ve relied on ever since:

• Evolve the organizational data model

A well-structured data model is the backbone of any scalable data platform you would need to have in place for building your data products successfully. Start by breaking it down into sub-domains, validate each with the stakeholders, and iterate domain by domain. This also lays the groundwork for cleaner data pipelines and more meaningful analytics.

• Simplify and improve business processes

Data should reflect how the organization works today or will work tomorrow, not how it worked years ago. Revisit the process layer before locking in your data structures. When needed, go back to the ideation phase of the Design Thinking cycle.

• Test foundational data architectures

Don’t theorize – prototype. Try out different architectural patterns: Data Warehouse, Data Lakehouse, or hybrid architectures like Data Fabric.

• Evaluate your data access layer and serving mechanisms

Ask yourself: Do you need data marts or a semantic layer? Should your platform support APIs, reverse ETL, or direct-to-app analytics?

Here’s how Stas Samko, who leads our Data CoE, reflects on it: “When building a modern data platform, technical tooling is only a part of the picture. You need lean experimentation, strong process understanding, and a design-led approach to ensure the platform delivers real, measurable value. Start small, iterate fast, and design with people, not just infrastructure in mind.”

Data Governance and Scaling

Once your initial business cases show results, the foundational architecture is in place, and a core data budget is secured, the next critical question is: how do you ensure time and money are spent wisely, with the maximum return on investment?

Here are several best practices we found valuable at this stage:

1. Focus on Data Product Management

As your data initiatives scale, don’t expect business units to independently define the use cases or build data models. You need a dedicated team of Data Product Managers – ideally within your IT or Data organization. These professionals bridge the gap between data engineering and business domains. They drive alignment, surface real needs, and turn ideas into successful data products. Their domain expertise and stakeholder engagement are critical to unlocking the high-impact use cases. Many people we engage pass our Product Management trainings, so it is not always mandatory to rely on extremely experienced staff only.

2. Adopt Lean Data Governance

You don’t need a complex, multi-layered data governance structure to get started. Begin lean: assign a Data Architect, ideally someone who also understands and maintains the corporate data model. Yes, this creates a temporary bus factor risk, but it’s essential to set the foundation. This person becomes your internal compass, guiding data quality, structure, and reuse. You can build broader governance frameworks as your data maturity grows.

3. Introduce Internal Billing for Data Products

One way to drive accountability and perceived value is through internal billing models. When departments are “charged” for the data products set-up and use, they treat them with more consideration, just like any external service. This model helps justify future funding, reinforce product value, and make data products ROI more visible. This also brings more discipline to the data services provision processes – people simply won’t pay for poor quality or zero impact.

4. Explore External Data Marketplaces

Not all data must remain internal. If you have datasets that don’t expose sensitive IP or commercial secrets, consider monetizing them externally. Numerous data marketplaces allow you to list datasets that can be accessed and purchased through APIs. This can create an additional revenue stream to fund further IT or data investments, while positioning your company in the data economy.

5. Enable Self-Service Analytics

Your best-built data products should be the flagship examples. Though you can’t build everything for everyone. Empower your users with self-service BI capabilities. Establish well-designed Data Marts and implement a permission-based access layer, where access is automatically controlled based on roles or departments. This creates a balance between flexibility and governance, and unleashes analytical potential across your organization.

6. Engage Hyperscalers for Co-Financing

If you’re using cloud infrastructure, don’t overlook funding opportunities from hyperscalers like Microsoft (Azure ECIF), Amazon (AWS MAP), or Google Cloud. These companies often co-finance up to 50% of your project costs if you commit to using their services and work with approved partners. These programs can dramatically improve your cloud ROI and support larger-scale data platform rollouts.

Conclusion

At Sigma Software, we believe in the Customer Zero strategy – testing and refining every data capability internally before offering it to clients. Now, after more than 15 years of Data CoE evolution, our internal best practice has become our professional offering, which we share here.

Key Best Practices from Sigma Software’s Data Excellence:

• Start with Value: Secure stakeholder buy-in and embed business case thinking across the organization through training and cultural change.
• Design for Impact: Use Design Thinking to align data models with real business processes and iterate quickly with cost-effective PoCs using open-source tools.
• Build Trust Visually: Focus on intuitive, high-quality data visualizations to drive early adoption, even before the backend is fully mature.
• Establish Lean Governance: Begin with a dedicated Data Architect and grow governance organically, without over-engineering.
• Empower and Monetize: Enable self-service analytics through Data Marts and permissions, explore internal billing and external data monetization strategies.
• Leverage Hyperscaler Support: Tap into cloud co-financing programs (e.g., ECIF, MAP) with Sigma Software as your trusted implementation partner.

Through this approach, we have transformed internal innovation into real-world data solutions that reduce costs, generate revenue, and unlock business insights – helping our clients turn their data from a cost center into a scalable model that powers competitive advantage.

• #BI & Big Data