My Role
Role: Senior Product Designer II
Team: Collaborated with PM, Developers, Technical Writers, Customer Success, Early Customers, and Implementation Engineers.
Responsibilities:
• Led end-to-end product design from discovery to delivery
• Conducted user research and usability testing
• Synthesized findings and presented recommendations
• Created user flows, prototypes and high-fidelity UI
• Facilitated cross-functional workshops, design critiques and design system feedback
• Worked closely with engineering to ensure design feasibility, design system alignment, and quality.
Role: Senior Product Designer II
Team: Collaborated with PM, Developers, Technical Writers, Customer Success, Early Customers, and Implementation Engineers.
Responsibilities:
• Led end-to-end product design from discovery to delivery
• Conducted user research and usability testing
• Synthesized findings and presented recommendations
• Created user flows, prototypes and high-fidelity UI
• Facilitated cross-functional workshops, design critiques and design system feedback
• Worked closely with engineering to ensure design feasibility, design system alignment, and quality.
The Problem
Growing Cloud Data Warehouse Industry
Cloud Data Warehouses (CDWs) like Snowflake, Databricks, and BigQuery had become foundational in enterprise data ecosystems, with over 40% of Fortune 500 companies adopting Snowflake alone. Tealium’s enterprise customers increasingly relied on CDWs as their central source of truth for customer, event, and behavioral data.
Lack of CDW Integrations in Tealium
Despite the growing reliance on CDWs, Tealium lacked integrated workflows for customers to connect their Cloud Data Warehouses to the Tealium platform.
Despite the growing reliance on CDWs, Tealium lacked integrated workflows for customers to connect their Cloud Data Warehouses to the Tealium platform.
The problem was that Tealium lacked native, first-party integrations with major CDWs, causing two issues:
• Customers faced friction in syncing and activating data stored in Snowflake and similar platforms.
• This gap increased time-to-value for marketing and analytics teams, becoming a roadblock for existing customers and companies evaluating Tealium’s offering.
• Customers faced friction in syncing and activating data stored in Snowflake and similar platforms.
• This gap increased time-to-value for marketing and analytics teams, becoming a roadblock for existing customers and companies evaluating Tealium’s offering.
The Goal
Bridging the Connection Gap
Our goal was to create a scalable way for customers to connect their Cloud Data Warehouses to the Tealium platform. Tealium prioritized Snowflake as the first CDW integration based on market share, customer demand, and strategic alignment.
We focused on three key objectives:
• Connection Configuration: Build scalable, secure, and easy-to-configure connections to leading CDWs.
• Targeted Queries: Enable querying of customer data to map into Tealium for more targeted marketing activations downstream.
• Data Monitoring: Deliver clear visualizations and metrics to help users monitor ingestion health and troubleshoot issues.
• Connection Configuration: Build scalable, secure, and easy-to-configure connections to leading CDWs.
• Targeted Queries: Enable querying of customer data to map into Tealium for more targeted marketing activations downstream.
• Data Monitoring: Deliver clear visualizations and metrics to help users monitor ingestion health and troubleshoot issues.
Discovery
Establishing the Connection Workflow
To establish a high level workflow of configuring a connection to Snowflake, we researched existing patterns within the product and landed on a general workflow consistent with our pattern of screens to View, Create, Edit, and Monitor.
Our goals for the workflow were to make it:
Reusable: A workflow pattern that is reusable with future CDWs added.
Intuitive: Minimize the steps involved and reduce complexity.
Flexible: Handle the wide variety of use cases and possible configurations.
Consistent: Be seamless in the existing experience without increase scope and redesign.
Our goals for the workflow were to make it:
Reusable: A workflow pattern that is reusable with future CDWs added.
Intuitive: Minimize the steps involved and reduce complexity.
Flexible: Handle the wide variety of use cases and possible configurations.
Consistent: Be seamless in the existing experience without increase scope and redesign.
Comparing existing workflows: Aligning early wireframes to existing configuration flows in the product, aiming for consistency, consideration of potential touch points, and reusable patterns.
Establishing the Big Picture: Landing on the key workflows of the configuration process involving Viewing, Creating, Editing, and Monitoring.
Early Feedback
Mapping Insights to Opportunities
Mapping Insights to Opportunities
After establishing the core workflow and initial low-fidelity designs, we conducted feedback sessions with both internal Customer Success teams and external customers. We then gathered and organized the insights into groups and mapped them to opportunities.
Synthesizing this input revealed a common concern: users needed more control over what data was ingested and when, especially given the massive scale and unpredictability of their datasets.
We learned that:• Customers often manage millions of rows across large database tables.
• Ingestion timing varies widely with some tables updated hourly, others only weekly.
This feedback reframed our focus and prompted a critical design question:
“How might we give users control of what data they want to ingest and when?”
Design Iteration
Addressing issues with the Query
Leveraging insights from early research, we began iterating on design solutions that would give users more control over how and when their data is ingested, while balancing flexibility with ease of use.
We focused on addressing three key user needs:
We focused on addressing three key user needs:
1. Seamless query-preview flow: Users needed the ability to move quickly between writing a query and previewing results without losing context or progress.
2. Guided query building with guardrails: A powerful yet accessible query builder was essential, one that supports customization but prevents overly complex or system-breaking queries.
3. Flexible scheduling: Ingestion timing had to match the diverse update cadences we heard from customers, ranging from hourly to weekly updates.
1. Tackling the disconnect with the query and preview: Users rarely get their query right on the first attempt, so separating query selection and data preview into different steps created unnecessary friction during trial-and-error.
After: The query and preview are brought closer together, allowing for faster trial-and-error to get to the right data.
2: The existing query selection is too simple: If the query is overly simplified, the data query will pull in too much data. However, if we enable a full advanced SQL editor, we run into technical complexities if there are no boundaries. We explored different ways to handle this balance of simplicity and flexibility of the query.
Solution Exploration: Building a query with condition drop downs.
Drawback: Very technically complex to build, difficult to set logic boundaries.
Solution Exploration: Offering a more flexible SQL based query
Drawback: Difficult to set boundaries within the logic and only works for SQL experts.
Final Design: Striking a balance with rigid selection of columns but flexible "where" clause section to input SQL for targeted data selection logic if needed.
3: Many Customers Need to Specify Ingestion Schedule: A default, set ingestion in many cases will be unnecessary and could incur additional processing costs.
Solution: Add a scheduling section to the configuration that covers the time ranges we heard about in our feedback and interview sessions.
Design Iteration
Helping Users Monitor Inbound Data
To help guide designs around connection monitoring and observability, we leveraged the mapped opportunities from our research to define key aspects of the user experience.
We focused on the these key user needs to help monitor and analyze data coming into the Tealum platform:
1. Visibility into the status of configurations.
2. Awareness of errors and the ability to see them at a glance.
3. Ability to view and troubleshoot errors with error log views and exporting.
1. Visibility into the status of configurations.
2. Awareness of errors and the ability to see them at a glance.
3. Ability to view and troubleshoot errors with error log views and exporting.
Solution Exploration: Leveraging the existing accordion detail views to include metrics.
Drawback: Cramped area for insights, additional clicks to get error logs and exporting.
Final Designs: Slideout with tabs dedicated to insights, error logs, and exporting.
Improvements: Prioritization of error messages, larger screen area for insights, more scalable user interfaces for future enhancements.
Impact
Establishing a Scalable Workflow
The feature first launched as an Early Access release to a small group of customers, allowing our team to stress test performance, query logic flexibility, and ingestion reliability under large data volumes. After several weeks of iterative feedback and refinement, the feature moved to General Access and became the foundation for Tealium’s broader CDW integration strategy.
Key Outcomes:
Key Outcomes:
• Established a scalable integration pattern that is now being reused to accelerate future cloud data warehouse (CDW) connection flows.
• Doubled data ingestion speed by enabling a direct-streaming connection, resulting in up to a 100% increase compared to previous ingestion methods.
• Closed a critical platform gap, positioning Tealium more competitively within the rapidly growing $9B CDW ecosystem.
• Closed a critical platform gap, positioning Tealium more competitively within the rapidly growing $9B CDW ecosystem.