Guiding Data Mesh Lean Transformations: A Full Workshop Facilitation Example

Facilitating a successful Data Mesh Lean Transformation requires structured workshops, clear facilitation strategies, and alignment between business and technical teams. Many organizations struggle with onboarding teams into Data Mesh principles, defining thin slices, and ensuring incremental value delivery. This guide provides a step-by-step approach to facilitating Data Mesh transformation workshops, using proven techniques such as Lean Inception, the Double Diamond framework, and the DUTI methodology.

How do we identify the right domain to start with? How do we align teams on a shared vision? How do we structure workshops that lead to real outcomes rather than endless discussions?

Having facilitated and observed many Data Mesh transformations, I’ve seen firsthand that successful adoption is not just about technology—it’s about structured collaboration. Without clear facilitation, organizations risk slow adoption, misaligned teams, and data products that fail to deliver real value.

This article provides a step-by-step guide for facilitators leading Data Mesh Lean Transformations, following the structured Double Diamond Approach. It outlines the sequence of workshops and Mural boards, covering everything from selecting the right domain to defining thin slices and implementing incremental steps toward a use case.

If you’re responsible for facilitating workshops, aligning teams, or guiding an organization through Data Mesh adoption, this guide will equip you with the tools and techniques needed to drive alignment, execution, and real transformation.

DUTI: A Thin-Slice Approach to Data Mesh Transformation

Many transformations fail because they try to do too much at once. Instead of attempting a massive organizational shift, the DUTI approach structures the transformation into thin, incremental steps:

1️⃣ Domain → Start with a single, well-chosen domain. This ensures early success and reduces complexity.

2️⃣ Use Case → Identify the highest-value use case for that domain, ensuring business impact.

3️⃣ Thin Slice → Define the smallest possible unit of value that can be delivered and validated.

4️⃣ Increment → Expand step by step, refining and scaling as you learn.

The image sequence below demonstrates the steps on the DUTI approach.

Data Mesh Transformation About to Start for a Organisation

Data Mesh Transformation – Domain Selected (on the Data Mesh Exploration Phase)

Data Mesh Transformation – Use Case Selected (on the Data Mesh Accelerate Phase)

Data Mesh Transformation – Thin Slicing The Use Case (on the Data Mesh Discovery & Lean Inception Phases)

This thin-slice strategy ensures quick wins, faster learning cycles, and tangible business outcomes—without overwhelming teams with unnecessary complexity.

The Data Mesh Double Diamond Approach: Structuring the Transformation

Successful Data Mesh transformations require more than just technical changes—they demand a structured approach to collaboration. Without clear facilitation, teams risk misalignment, unclear priorities, and slow adoption.

One of the most effective ways to guide teams through a Data Mesh transformation is by using the Double Diamond Approach, a structured framework that balances problem exploration with solution execution. Originally introduced in design thinking, this model helps organizations move from understanding the challenge to defining clear, actionable steps.

Visualizing the Data Mesh Transformation Process

Below is a diagram I helped design for one of ThoughtWorks’ largest clients on their Data Mesh journey. Over time, this framework became widely adopted by other organizations.

The Data Mesh Double Diamond: image from the Getting off to the right start (Part I) article

This diagram serves as the foundation for the facilitation process, breaking the Data Mesh Lean Transformation into four key phases:

1. Data Mesh Exploration → Understanding the organization’s Data Mesh readiness and selecting the best domain to start with.
2. Data Mesh Accelerate → Aligning stakeholders, defining strategic goals, and identifying high-impact use cases.
3. Data Mesh Discovery → Deep-diving into the chosen domain, refining objectives, and mapping data product interactions.
4. Data Mesh Lean Inception → Structuring incremental delivery and defining the execution path for the first Thin Slice

Each of these phases involves specific workshops and facilitation techniques to ensure alignment across business and technical people. By structuring the transformation in clear, actionable steps, organizations can avoid the common pitfalls of endless discussions, unclear priorities, and misaligned teams.

In the next sections, I’ll walk through an example, applying this framework step by step—from Data Mesh Exploration to Lean Inception.

How to Read This Article

This article follows the Double Diamond approach, guiding you through each phase with structured explanations and a real-world example. To help you navigate, I’ll use visual markers to indicate the phase we’re in.

Data Mesh Double Diamond

Beyond just theory, you’ll see a full example unfold step by step, illustrating how these concepts are applied in practice. This article has been developed over the past two years, shaped by insights from my training sessions on : Guiding Data Mesh Lean Transformations Training, and refined based on real-world facilitation experiences.

I encourage you to explore the Mural boards provided, use them in your own sessions, and leverage this hands-on approach to accelerate your Data Mesh Lean Transformation.

Data Mesh Exploration: Identifying the Right Starting Point

A successful Data Mesh transformation begins with a critical decision: Which domain should be onboarded first (or next) into the mesh?

Data Mesh Exploration

Selecting the wrong domain can lead to resistance, slow adoption, and an unsuccessful pilot. That’s why the Exploration phase plays a crucial role in ensuring that efforts start where they are most likely to succeed.

One of the most effective tools for this phase is the Data Mesh Readiness Assessment—a structured approach that evaluates and compares different domains based on key factors such as:

• Organizational Complexity – Does the domain have multiple data sources and diverse use cases?
• Data-Oriented Strategy – Is there a strong focus on deriving value from data?
• Executive Support – Does leadership actively support Data Mesh adoption?
• Technology Readiness – Does the domain have the modern engineering practices required?
• Early Adopter Mindset – Are teams open to experimentation and innovation?
• Domain-Oriented Structure – Is the organization aligned around business domains?
• Long-Term Commitment – Is there a clear vision for continuous transformation?

By systematically scoring domains using these criteria, organizations can prioritize the domain with the highest readiness score, ensuring a smoother transition.

You can access and use the Data Mesh Readiness Assessment here.

Applying the Readiness Assessment: A Practical Example

Imagine you are working with an enabling team supporting a Data Mesh transformation for a large e-commerce platform. The journey has already started with the Product Catalog domain, which manages product details, categories, and inventory levels. Now, it’s time to decide which domain to onboard next.

Following the DUTI methodology (Domain, Use Case, Thin Slice, Increment), the organization assesses multiple domains and identifies two candidates:

• Customer Management – Handles customer profiles, preferences, and purchase history.
• Order Processing – Manages order placement, payment processing, and shipment tracking.

Assessment Results

After conducting interviews and scoring each domain, the Customer Management domain emerges as the top choice due to its higher readiness scores. Below is he result of both assessments.

Data Mesh Readiness Assessment Results for Customer Management Domain

Data Mesh Readiness Assessment Results for Order Processing Domain

By selecting the most prepared and strategic domain, organizations maximize the chances of success and create a strong foundation for scaling Data Mesh across the enterprise.

With the right domain selected, you now need to align stakeholders and define strategic goals before diving into implementation. This is where the Data Mesh Accelerate Workshop comes in, ensuring the domain business and technical people move forward with a shared vision.

Data Mesh Accelerate: Aligning Your Domain and Defining Strategic Goals

Now that you’ve identified the right domain to bring into the Data Mesh, the next step is getting everyone on the same page—from business stakeholders to technical people.

Data Mesh Accelerate

This is where the Data Mesh Accelerate Workshop comes in. This structured session ensures that the domain team aligns on a shared vision, defines clear objectives, and prioritizes the most valuable use case to start with. Without this alignment, Data Mesh adoption can stall, leaving stakeholders and teams misaligned and struggling to deliver value.

By the end of this workshop, you’ll have:

• A Nirvana vision—a clear picture of what success looks like for your domain.
• Defined Objectives and Key Results (OKRs) to measure transformation success.
• A prioritized Use Case that delivers business impact.
• An initial Data Product Interaction Map, showing how your data products will interact.

Read the original article on Data Mesh Accelerate here: Data Mesh in Practice: Getting Off to the Right Start.

Let’s go step by step.

1. Start with a Shared Vision (Your Nirvana Statement)

Before diving into specifics, align your team on the future you’re working toward. Without this, conversations can quickly become fragmented—business people focus on outcomes, while technical teams focus on architecture and technical details. You need both perspectives, but they must be anchored in a common goal.

Example Nirvana Statement for the Customer Management Domain

“Our Nirvana is a future where customer data flows effortlessly across the organization, empowering teams with on-demand, self-serve access to trusted insights. From product development to customer support, everyone operates with up-to-date, reliable insights, driving strategic decisions and personalized experiences.”

2. Define Your Objectives and Key Results (OKRs)

With your vision in place, you now need measurable goals to track progress.

Your OKRs should answer:

• What do we need to achieve for our domain to be successful in Data Mesh?
• How will we measure whether we are getting there?

Example OKRs for the Customer Management Domain

Objective 1: Successfully onboard the Customer Management domain onto the Data Mesh platform and establish a foundational data product.
✔ KR1: Publish and make accessible at least one customer data product within 12 weeks.
✔ KR2: Onboard at least two business partners to actively use the published data product.
✔ KR3: Achieve 80% compliance with data governance and security standards in six months.
✔ KR4: Ensure a team representative actively participates in the data governance team (attending at least 75% of bi-weekly meetings).

Objective 2: Improve efficiency and reduce lead time for accessing customer insights.
✔ KR1: Reduce the lead time for retrieving customer data for reports by 30% compared to the pre-mesh baseline.
✔ KR2: Implement automated data pipelines to reduce manual intervention by at least 40%.
✔ KR3: Achieve a 75% positive feedback score from initial users regarding ease of access and usability.

Now, your team has clear goals that align business and technical priorities.

3. Exploring and Prioritizing Use Cases

With your OKRs in place, it’s time to move from vision to action. But where should you start? Not all use cases are created equal—choosing the wrong one can lead to wasted effort, while the right one can showcase early wins and build momentum for your Data Mesh transformation.

In this step, you’ll explore multiple potential use cases, ensuring they align with your OKRs and strategic goals. But identifying use cases isn’t enough—you also need to prioritize the one that delivers the most value while remaining feasible.

Here’s how you’ll do it:
– Brainstorm use cases that align with your domain’s business objectives.
– Map each use case to the OKRs to ensure strategic alignment.
– Prioritize one use case that balances business impact, feasibility, and readiness, serving as the starting point for execution

To maintain focus, use the following use case template:

We believe that [USE CASE] will help achieve [OBJECTIVE].
We know we are getting there based on [KEY RESULTS].

Example Use Cases for the Customer Management Domain

In our example, we are working with the Customer Management domain within a large e-commerce company. This domain is responsible for customer profiles, preferences, and purchase history.

During the Data Mesh Accelerate Workshop, you and your team collaborate to explore how Data Mesh can create value for this domain. After brainstorming, you identify five potential use cases, each directly mapped to an OKR.

Use Case 1: Personalized Customer Recommendations

We believe that leveraging customer data to provide personalized product recommendations will help achieve improved efficiency and reduced lead time for accessing customer insights through the Data Mesh platform.

We know we are getting there based on:

• Reducing the lead time for retrieving customer data for business reports by 30%.
• Implementing automated data pipelines, reducing manual intervention by 40%.
• Achieving a 75% positive feedback score from initial users on usability.

Use Case 2: Customer Churn Prediction Model

We believe that implementing a customer churn prediction model based on behavioral data will help achieve improved efficiency and reduced lead time for accessing customer insights through the Data Mesh platform.

We know we are getting there based on:

• Reducing the lead time for retrieving customer insights by 30%.
• Automating data pipelines, reducing manual intervention by 40%.
• Achieving a 75% positive feedback score from internal users.

Use Case 3: Customer 360° View for Support Teams

We believe that consolidating customer interactions into a 360° view for support teams will help achieve successful onboarding of the Customer Management domain onto the Data Mesh platform and establish a foundational data product.

We know we are getting there based on:

• Publishing at least one customer data product via the Data Mesh platform within 12 weeks.
• Onboarding at least two key business partners to actively use the data product.
• Ensuring a team representative attends 75% of bi-weekly governance meetings.

Use Case 4: Real-Time Customer Engagement Insights

We believe that enabling real-time analytics on customer interactions across multiple channels will help achieve successful onboarding of the Customer Management domain onto the Data Mesh platform and establish a foundational data product.

We know we are getting there based on:

• Publishing at least one customer data product within 12 weeks.
• Onboarding at least two business partners to use the data product.
• Achieving 80% compliance with data governance standards within six months.

Use Case 5: Customer Feedback Analysis for Product Improvement

We believe that analyzing customer feedback across various touchpoints to identify patterns and trends will help achieve successful onboarding of the Customer Management domain onto the Data Mesh platform and establish a foundational data product.

We know we are getting there based on:

• Publishing at least one customer data product within 12 weeks.
• Onboarding at least two business partners to use the data product.
• Ensuring a team representative attends 75% of bi-weekly governance meetings.

4. Selecting the Best Use Case to Start With

Now that you’ve identified the key use cases, it’s time to prioritize and select the best one to implement first. Your goal is to ensure early business impact while maintaining technical feasibility and alignment with the Data Mesh strategy.

Prioritizing Use Cases with a Collaborative Approach

Now that you’ve identified potential use cases, it’s time to prioritize and select the best one to implement first. The goal is to maximize business value while ensuring technical feasibility and readiness.

To do this, use the Collaborative Prioritization Table, a structured approach that evaluates each use case based on:

• Business Value – Strategic importance, urgency, and risk reduction.
• User Experience – Reach, impact, and confidence in delivering value.
• Effort Required – Relative effort compared to other use cases.

Instead of assigning absolute values, this method compares initiatives relative to each other, ensuring a fair prioritization process.

Learn more about this prioritization approach in the article:
WSJF + RICE: Collaborative Prioritization Table

Applying Prioritization to Our Example: Customer Management Domain

Following our example, in the Data Mesh Accelerate Workshop, you assess each of the five use cases using a collaborative prioritization table. By applying the business value, effort, and readiness criteria, you determine that Customer 360° View for Support Teams is the best use case to start with.

Below is an image of the Collaborative Prioritization table used in this process.

sample Use Case Collaborative Priorization table

Why start with the Personalised Customer Recommendation Use Case?

The prioritization table highlights why this use case is the best starting point: it delivers high business value while requiring lower effort compared to the alternatives. By starting with a high-impact but feasible use case, you ensure quick wins that build momentum and confidence in the Data Mesh transformation. Getting this step right reduces risk, accelerates adoption, and helps your team validate the approach before scaling.

With the Domain (D) and Use Case (U) selected, your next step is to define Thin Slices (T) and Increments (I)—applying the DUTI approach (Domain, Use Case, Thin Slice, Increment) to ensure early and continuous delivery of value.

However, moving directly from use case selection to thin slices and increments can be too abrupt. To bridge this gap, the Data Mesh Double Diamond framework follows a structured flow: Accelerate → Discovery → Lean Inception.

Now, you’re ready to identify the key data products that will enable your selected use case. This next step ensures that your Data Mesh transformation moves from strategy to execution, setting you up for success in the Discovery phase.

Next Step: Identifying Data Products

Now that you’ve selected Customer 360° View for Support Teams as your first use case, it’s time to define the data products that will support it.

In the next section, you’ll:

• Identify key data products required for the use case.
• Map out the relationships between these data products.
• Define ownership and governance structures to support sustainable implementation.

This is where your Data Mesh transformation shifts from theory to execution, ensuring that real business impact is delivered incrementally and iteratively.

5. Discovering Data Products: From Use Case to Execution

With the Use Case selected– in our example,  the Personalised Customer Recommendation  —, you now shift focus to defining the data products required to enable it.

This step serves as the bridge between the Accelerate and Discovery phases, ensuring that the Data Mesh transformation moves from strategic alignment to practical implementation.

What You’ll Do in This Step:

• Start identifying which data products are needed to fulfill the selected use case.
• Start defining the producers and consumers of each data product.
• Draft an initial Data Product Interaction Map to visualize how data will flow.

This is not about designing every technical detail upfront—it’s about establishing an initial draft that will evolve during the Discovery phase.

Applying This Step to Our Example: Customer Management Domain

During the Data Mesh Accelerate Workshop, you and your team create an initial draft of the Data Product Interaction Map for the Personalized Customer Recommendation Use Case.

Below is a draft interaction map illustrating the key components and data flows required to fulfill the Personalized Customer Recommendation Use Case.

Sample draft interaction map from the Accelerate workshop

This draft interaction map helps the team visualize the flow of data to fulfill the use case needs and align on what requires further refinement in the Discovery phase. But more importantly, beyond the technical aspect, it provides stakeholders with a clear glimpse of the work ahead—helping them see how everything connects: the use case, the OKRs, and their overarching vision. By making these connections explicit, the team ensures that technical and business leaders stay aligned on priorities, reinforcing the purpose behind the transformation.

Data Mesh Discovery: Refining Understanding Before Execution

The Discovery phase is a structured investigation process that enables your team to refine the selected use case, validate assumptions, and define key requirements before moving into execution. It plays a critical role in the Data Mesh Double Diamond approach, ensuring that you gather the necessary insights to make informed decisions without getting stuck in lengthy upfront analysis.

Data Mesh Discovery

At its core, Discovery is about reducing uncertainty. Your team examines and aligns on the essential aspects required to implement the use case—setting the stage for a smooth and well-informed transition into Lean Inception and delivery.

Following the Double Diamond approach for Data Mesh, the Discovery phase helps your team answer key questions, such as:

• Further understanding how this use case creates business value.
• What data products are required to support it?
• What governance, security, and platform requirements must be considered?
• How do these data products interact with each other?
• What potential risks or dependencies should be addressed before moving forward?

Rather than spending months in analysis, Discovery should be lean but insightful—equipping teams with enough information to confidently transition into Lean Inception, without falling into the trap of big upfront design.

Applying Discovery to Our Example: The Personalized Customer Recommendation Use Case

Now, let’s bring these concepts into our example.

Context: Your are running Discovery for the Personalized Customer Recommendation Use Case, previously selected in the Accelerate Workshop.

During the Accelerate Workshop, your team drafted an initial Data Product Interaction Map (image above) to visualize how data flows through the system. Now, in the Discovery phase, you refine this draft by:

• Clarifying data sources – Where does the data originate?
• Defining the scope of work – What tasks and components are required to implement each data product and enable this use case?
• Mapping input and output ports – How will different systems access and contribute to the data products?
• Identifying gaps and dependencies – Are additional data sources needed? Are there existing platform constraints that must be addressed?

Expanding the Discovery Scope

Beyond refining the Data Product Interaction Map, Discovery dives deeper into essential aspects that will define the success of the use case implementation. A key area is understanding how consumers will interact with the data products. This includes:

• Usage patterns – How will consumers access and use the data?
• Service Level Objectives (SLOs) – What are the performance and availability expectations?
• Technical constraints – What platform limitations or architectural decisions need to be factored in?

By addressing these questions, your team ensures that data products are not just technically feasible but also aligned with business and user needs.

While some governance and security considerations may have emerged during earlier workshops and discussions, the Discovery phase provides a structured opportunity to dive deeper into their implementation details. This includes defining:

• Ownership and access control – Who owns the data, and how is access managed?
• Compliance requirements – What policies and regulations must be adhered to?
• Interoperability guidelines – How do data products integrate with existing enterprise data ecosystems?

Key Deliverable: The Finalized Data Product Interaction Map

The Data Product Interaction Map is the primary output of Discovery. It serves as a blueprint for Lean Inception, helping teams:

• Refine data product boundaries – Avoiding overlaps and ensuring clear ownership.
• Identify dependencies and risks – Making implementation smoother and more predictable.
• Align execution with business objectives – Ensuring that technical solutions contribute to meaningful business outcomes.

In our example, the updated map now includes:

• Newly identified data products needed to support the use case.
• Additional input/output connections based on business needs.
• Defined governance considerations, ensuring data reliability and compliance.

Below is our sample Data Product Interaction Map:

sample Data Product Interaction Map

From Discovery to Lean Inception: What Comes Next?

With a well-defined Data Product Interaction Map, your team is now ready to move into Lean Inception, where you will:

• Define Thin Slices for incremental delivery.
• Plan execution using the DUTI approach (Domain, Use Case, Thin Slice, Increment).
• Align the team around a realistic, value-driven implementation plan.

Next up: Lean Inception—Turning Discovery into Actionable Execution.

The Data Mesh Lean Inception

In this section, we will continue with the same example from the previous phases and showcase a sample artifact generated in each Lean Inception activity. This ensures continuity and provides a concrete reference for understanding how Lean Inception supports the execution of a Data Mesh transformation.

Data Mesh Lean Inception

The Data Mesh Lean Inception focuses on collaboratively thin slicing the selected use case and defining an actionable plan to experiment and deliver value incrementally.

Building on insights gathered during the Discovery phase, the Lean Inception workshop brings together business and technical perspectives to align on priorities, identify thin slices of value delivery, and establish a clear roadmap for data product implementation.

Let’s go step by step.

1. Analytical Use Case Statement

This activity aims to explain the target Use Case and its relation to the desired objectives or outcomes to be achieved with data. Use cases translate the desired outcome for a customer or a user. In the context of Data, a Use Case describes how analytics can be used to improve a user experience or operational efficiency of a business process. The implementation of a Use Case should enable a desirable Objective/Outcome, which is demonstrated by metrics.

Template: We believe that [Use Case] will help achieve [Objective / Outcome]. We know we are getting there based on [Metrics].

Below is an image of the selected use case, which was defined during the Accelerate workshop and refined in the Discovery phase. In this activity, the group answers any remaining questions before moving forward.

Analytical Use Case Statement for our example

2. Personas, Needs, and Their Questions

A persona represents a consumer of data, describing their role, needs, and the questions they expect data to answer. Once the use case is well understood, the next step is to identify relevant personas and their data needs and questions to be answered in order to fulfill the Use Case.

Example Persona:

Persona: Sarah, 34, E-commerce Marketing Manager
Goals / Needs:

• Increase customer engagement and sales through personalized recommendations.
• Improve the efficiency of marketing campaigns by targeting the right audience.
• Leverage data to understand customer preferences and buying behavior.

Questions to be Answered:

• What products are customers most likely to purchase next?
• How effective are our current recommendation strategies?
• What are the key factors influencing repeat purchases?

Current Pain Points:

• Limited access to real-time customer insights.
• High dependency on manual data analysis processes.
• Struggling to personalize recommendations at scale.

Sample Persona for our example

3. Decision Journeys (AS-IS)

The Decision Journey (AS-IS) describes the current sequence of steps a persona follows to make decisions. Some of these steps represent current usage of source systems, with pain points that characterize inefficiencies, manual work, and data limitations in the current process.

In our example, Sarah, an e-commerce marketing manager, wants to create a personalized email campaign for returning customers. Below is her current journey and the associated challenges:

1. Log into the marketing analytics dashboard

   • Sarah starts by accessing the analytics platform to collect customer insights.

2. Download recent customer purchase history reports

   • She manually extracts purchase reports, but this process is time-consuming.

3. Cross-check customer purchase trends manually in spreadsheets

   • Sarah compares data manually, making trend analysis slow and error-prone.

4. Identify high-value returning customers based on spending patterns

   • She determines which customers should be targeted, but lacks automation.

5. Analyze customer preferences using past interactions and browsing behavior

   • This step is limited due to a lack of visibility across different platforms.

6. Create a personalized email list with targeted product recommendations

   • She compiles the list manually, which increases workload.

7. Upload the list to the email marketing platform

   • This requires manual formatting, adding extra effort and reducing efficiency.

8. Design the email campaign with relevant product suggestions

   • The campaign is created based on the segmented customer data.

9. Send the campaign and monitor engagement metrics

   • Engagement is tracked, but delayed reporting makes real-time adjustments difficult.

10. Evaluate the campaign performance after a week

    • Sarah reviews performance data, but insights come too late to optimize the campaign.

This journey highlights key pain points, such as manual data handling, slow reporting, and limited visibility, which impact efficiency and decision-making.

Below is the image of a sample Decision Journey (AS-IS) for our example:

Sample Decision Journey AS-IS for our example

4. Consumer Data Product Usage Pattern

To effectively design Consumer Data Products that precisely meet user needs, understanding their data usage patterns is essential. This insight informs the setting of Service Level Objectives and the establishment of technical constraints, ensuring that products align with both user expectations and data limitations.

Below is the image of a sample Consumer Data Product Usage Pattern for our example:

Sample Consumer Data Product Usage Pattern for our example

To illustrate how a Recommendation Engine operates within a Consumer Data Product, we analyze key usage factors that influence its design and implementation:

Key Questions & Answers

1. How often would she use it?

   • Every second
   • Every hour
   • Several times a day
   • Daily
   • Weekly
   • Monthly+

2. How complete does the data need to be?

   • 100% complete
   • Mostly complete
   • Can tolerate missing data

3. How accurate does the data need to be?

   • Perfectly accurate
   • Some room for error
   • As long as it’s representative

4. How fresh does the data need to be?

   • Real-time
   • Within minutes
   • Within hours
   • Within the day
   • Within the week
   • Within the month

5. How often is it updated?

   • Every second
   • Every hour
   • Several times a day
   • Daily
   • Weekly
   • Monthly+

6. When does it need to be used?

   • 24/7
   • Awake time
   • At a specific hour

7. How many people or services use it?

   • Millions
   • Thousands
   • Hundreds
   • Dozens
   • A handful
   • Very few

5. Decision Journeys (TO-BE)

The Decision Journey (TO-BE) outlines the necessary steps for a user to effectively engage with a Consumer Data Product, ensuring that all components are in place to provide valuable insights and answers to critical questions.

Below is the image of a sample Decision Journeys (TO-BE) for our example:

Sample Decision Journey TO-BE for our example

Example: Personalized Email Campaign for Returning Customers

Sarah wants to create a personalized email campaign for returning customers. In this improved version of her journey, data is seamlessly retrieved, processed, and utilized through structured consumer data products, reducing manual efforts and inefficiencies.

1. View personalized recommendations for a customer

   • Domain Entity: Recommendation Engine (Consumer Data Product)
   • Process: Displays insights in the email marketing platform
   • System: Email Marketing Platform
   • Consumption Layer: Analytical Workflow (Email Campaign Reports)

2. Retrieve customer profile

   • Domain Entity: Customer Profile (Aggregate Data Product)
   • Process: Combines customer attributes
   • System: CRM System
   • Consumption Layer: Analytical Workflow (Campaign Setup)

3. Retrieve customer purchase affinity

   • Domain Entity: Customer Purchase Affinity (Aggregate Data Product)
   • Process: Analyzes buying patterns
   • System: CRM System
   • Consumption Layer: Analytical Workflow (Campaign Setup)

4. Retrieve product catalog

   • Domain Entity: Product Catalog (Aggregate Data Product)
   • Process: Enriches recommendations with available products
   • System: Product Management System (another domain)
   • Consumption Layer: Analytical Workflow (Campaign Setup)

5. Process survey and support interactions

   • Domain Entity: Customer Survey (Source Data Product)
   • Process: Analyzes customer opinions and trends
   • System: Survey System
   • Consumption Layer: Operational Workflow (Data Retrieval)

6. Process purchase history

   • Domain Entity: Purchase History (Source Data Product)
   • Process: Tracks buying patterns and returns
   • System: Order Management System
   • Consumption Layer: Operational Workflow (Data Retrieval)

7. Capture customer interactions

   • Domain Entity: Customer Engagements (Source Data Product)
   • Process: Collects behavioral and feedback data
   • System: Analytics Platform
   • Consumption Layer: Operational Workflow (Data Retrieval)

6. Solution Mapping

Having mapped the necessary steps to fulfill the Decision Journeys for the Use Case, it’s now time to focus on mapping out the solutions needed for development. Please concentrate on the scope of the current Use Case, including its Personas and their Decision Journeys, to identify the specific work items required to bring it to life.

Below is the image of a sample Solution Mapping for our example:

Sample Solution Mapping for our example

Solution Mapping for Sarah’s Personalized Email Campaign

In this phase, we break down the technical and process-related work items required to bring the TO-BE Decision Journey into reality.

Work Items for Each Step

1. View personalized recommendations for a customer

   • Develop Recommendation Engine API
   • Add recommendation module to email platform
   • Implement A/B testing for recommendations

2. Retrieve customer profile

   • Integrate Customer Profile with CRM
   • Develop data quality validation rules
   • Create API for profile retrieval

3. Retrieve customer purchase affinity

   • Build Customer Purchase Affinity Data Product
   • Implement affinity scoring algorithm

4. Retrieve product catalog

   • Expose product catalog API
   • Improve product metadata enrichment
   • Ensure catalog data compliance

5. Process survey and support interactions

   • Build survey data ingestion pipeline

6. Process purchase history

   • Develop ETL process for purchase data
   • Automate data archival strategy

7. Capture customer interactions

   • Integrate website feedback tracking tools
   • Develop customer behavior insights dashboard
   • Ensure GDPR compliance for tracking

7. Technical, Business, and UX Review

This review aims to assess the team’s confidence in their understanding of the technical, business, and UX aspects of each backlog item. By discussing these perspectives, the team can identify gaps, clarify uncertainties, and surface any disagreements that need resolution.

Below is the image of a sample Technical, Business, and UX Review result for our example:

Sample Technical, Business, and UX Review result for our example

8. Sequencer

The Sequencer is key in organizing and displaying the backlog items, aiding in the step-by-step realization of the Use Case. This is achieved through a carefully planned thin slicing strategy that focuses on delivering the smallest possible segments or features that still provide value to the user. On it, you will define the Use Case thinnest slice (MVP) and its subsequent increments.

Below is the image of a sample Sequencer for our example:

Sample Sequencer for our example

9. MVP Canvas

The MVP Canvas is a visual tool that helps the team align and define the thinnest slice (MVP) of the use case—the simplest combination of backlog items that can effectively improve a decision journey with minimal effort and maximum impact.

Having previously discussed and identified the essential elements that make up the MVP, it’s now time to consolidate and document them in a clear and structured manner, ensuring they align with business objectives and user needs while allowing for early validation and feedback.

Below is the image of a sample MVP Canvas for our example:

Sample MVP Canvas for our example

MVP Proposal

• What’s the Proposal for this MVP?
  • To validate if the newly implemented solution produces better results than the current manual process by comparing Sarah’s work using the Recommendation Engine Data Product with results obtained prior to implementation.

Segmented Personas

• Primary Persona: Sarah, E-commerce Marketing Manager
• Segment: Sarah’s email campaigns before and after the new system release
• Comparing Sarah’s manual process to the new system will provide a clear assessment of improvements in efficiency and accuracy.

Journeys

• What journeys are going to be improved with this MVP?
  • Sarah generates personalized email campaigns using the Recommendation Engine.
  • Sarah sends email campaigns and tracks customer engagement.
  • Data is analyzed to compare campaign performance before and after the new system.

Features

• Expose Product Catalog API (Simplified Version)

  • Expose a basic catalog API with only essential product attributes (e.g., product name, price, and category).
  • Limit API calls to predefined product segments rather than the entire catalog.
  • Use a static dataset instead of real-time integration.

• Develop Recommendation Engine API (Simplified Version)

  • Implement rule-based recommendations instead of complex AI/ML algorithms.
  • Focus on a single recommendation logic for “most purchased items” rather than multiple models.

• Ensure GDPR Compliance for Tracking (Simplified Version)

  • Implement only basic consent tracking (opt-in/opt-out mechanism).

• Create API for Profile Retrieval

• Develop ETL Process for Purchase Data

Expected Result

• What learning or result are we seeking in this MVP?
  • The new system should reduce manual effort for Sarah and improve the accuracy and efficiency of personalized email campaigns.
  • The goal is to demonstrate measurable improvement in campaign performance compared to the manual process.

Metrics to Validate the Business Hypotheses

• Comparison of email open rates before and after implementation (Target: +20% improvement)
• Increase in click-through rates and conversions based on personalized recommendations
• Reduction in time spent creating email campaigns (Target: -30% improvement)
• Compliance with GDPR regulations and company data policies

Cost & Schedule

• What is the expected cost and timeline for delivering this MVP to early customers?
  • 2 weeks for MVP deployment and validation
  • Validation Schedule: Compare campaign performance for one week before and one week after MVP release

Templates for Each Phase of the Double Diamond Data Mesh

To support your Data Mesh Lean Transformation, here are templates and example boards for each phase of the Double Diamond Data Mesh. These resources will help you structure workshops, facilitate collaboration, and guide teams through each stage of the Double Diamond Data Mesh.

These ready-to-use templates will help you facilitate each phase of your Data Mesh Lean Transformation, ensuring that teams stay aligned, focused, and effective throughout the process.

My Journey with Data Mesh and Facilitation

Facilitation has been at the core of my career for decades. While I may not claim to be “the best facilitator in the world,” as some colleagues have generously put it, I can confidently say that I am deeply experienced in guiding teams through complex transformations.

My journey with inceptions began in 1999 when I studied them for my master’s thesis. Years later, when I joined ThoughtWorks in 2006, I was introduced to Agile Inceptions and immediately saw their potential. I immersed myself in facilitation, leading as many inceptions as possible. Over time, I refined the traditional Agile Inception approach into Lean Inception, a structured methodology designed to help teams align and define their Minimum Viable Product (MVP). In 2016, I published a book on Lean Inception, and since then, I have facilitated hundreds of inceptions and discoveries.

As Data Mesh gained traction, I found myself increasingly drawn into more complex inceptions—not just for digital products but for platform teams and intricate domain transformations. While many of my colleagues focused on standard product inceptions, I worked on Data Mesh workshops, helping organizations navigate decentralized data ownership, domain-driven data product design, and cross-team collaboration.

After working with several large enterprises on their Data Mesh transformations, I recognized a gap: facilitators needed structured guidance to support teams in adopting Data Mesh principles effectively. While the technical aspects of Data Mesh were being expertly handled by my ThoughtWorks colleagues and the broader community, the facilitation side required a more tailored approach.

This realization led me to create a dedicated framework for Guiding Data Mesh Lean Transformations, specifically for those responsible for leading workshops, aligning teams, and onboarding organizations into the Data Mesh journey.

Acknowledgments

This article would not have been possible without the incredible people I have had the privilege to work and collaborate with, especially in the context of Data Mesh facilitation and transformation.

Even though I might unintentionally forget to mention some of these amazing contributors, I want to extend a special thank you to a few of them. (If I missed you, please ping me, and I will update the article!)

Zhamak Dehghani, Martin Fowler, Steve Upton, Darren Young, Ecem Biyik, Chris Ford, Emily Gorcebski, Ammara Gafoor, Ian Murdoch, Kiran Prakash, Danilo Sato, Carolina Vega, Luma Corrêa, Andherson Maeda, and Fernando Fernandes.

Thank you all for your insights, support, and contributions to the evolution of Data Mesh facilitation!

Take the Next Step

Looking to apply Data Mesh Lean Transformation in your organization? I offer effective online and real-time training on the Double Diamond Approach, DUTI methodology, and Data Mesh facilitation.

More details: Guiding Data Mesh Lean Transformations Training

Need expert facilitation for your Data Mesh workshops? I help organizations run Data Mesh Readiness, Accelerate, Discovery, and Lean Inception sessions.

Learn more: Workshop Planning & Facilitation