Addressing 60% Abandonment: The Strategic Redesign of Striim's Flow Designer
01.THE CHALLENGE
A Powerful Tool That Users Struggled to Use
Imagine stepping into a high-tech kitchen, equipped with the best tools for gourmet cooking—yet you’re left fumbling just to find a knife. That was the experience for Striim’s trial users in early 2024.
Flow Designer, Striim’s no-code, drag-and-drop interface for building data pipelines, was supposed to be the gateway to Striim’s powerful capabilities. Instead, users—primarily data engineers and architects found themselves stuck, facing an unintuitive interface that didn’t align with their mental models.
In the first half of 2024, a staggering 60% of trial users abandoned the platform before creating a single pipeline.
The impact was clear: poor adoption, low conversions, and an over-reliance on 1:1 sales demos to retain technically skilled users who should have been self-sufficient.
This case study explores how I uncovered key friction points and reimagined the experience to simplify the learning curve while helping users harness Striim’s capabilities, establishing a foundation for its next phase of AI-driven innovation.
What is Striim ?
Striim is a platform for building data pipelines—connecting sources, transformations, and targets. Like a conveyor belt, it moves data between systems (such as databases and cloud services) in real time to power analytics and AI.
02.my role
Design & Beyond
As the sole UX designer & researcher I:
Drove UX research in a company where it isn't an established practice
Aligned business goals with user needs
Led the end-to-end visual & interaction design
Lead cross-team collaboration
Secured executive buy-in for UX as a strategic business driver
The Case for investing in a redesign
The biggest obstacle? Convincing leadership that research was worth the investment.
03.The Research Journey
From Skepticism to Support
When I proposed investigating why users were abandoning Flow Designer, I hit resistance. With no formal research budget, I got creative:
Starting with what we had: I leveraged a colleague's heuristic evaluation that identified 11 usability violations.
Looking outward: I evaluated 17 competitor tools to establish industry standards and identify gaps in our interface.
Mining support data: Analysis of 151 support conversations revealed 74% basic navigation issues—suggesting fundamental usability problems
This preliminary work secured approval for deeper research through several interviews with the customer support & success teams, shadowing support sessions with users and direct user interviews.
Key Insights
Through affinity mapping of the qualitative data, three critical insights emerged:
Misaligned workflows: Users thought in terms of Design Time, Run Time, and Troubleshooting, but our interface didn't reflect this mental model.
Frustrating interactions: Basic actions like zoom, pan, and version control were missing or inconsistent with accepted norms.
High learning curve (Paradox of the Active User): Striim-specific concepts created unnecessary barriers to entry.
A snapshot of the affinity diagram grouped by stages in the user's workflow
Lack of scroll, pan and zoom in the old experience
Some comments on the old experience
04. designs
From Insights to Innovation
It was around this time, that my mentor recommended UX Magic by Prof. Daniel Rosenberg - which focuses on semantic interaction design and profoundly shaped my approach.
The Evolution
Defined the grammar (a semantic relationship between functionality and user goals modeled as objects, attributes & actions).
Visualized core objects and their behaviors.
Explored various UI controls, widgets, and layouts.
Refined the layout through multiple iterations.
The Designs
The Layout
The Design Time Layout
The Run Time Layout
Semantic Interaction Design
Mapped user goals (objects, actions, attributes) to UI elements to create a layout aligned with user mental models.
Recognition over Recall
Replaced generic icons with company logos to reduce cognitive load and enable quick identification of sources and targets.
Visibility of System Status
Used color, layout, actions, and metrics to communicate system state and provide immediate feedback—building trust through clear, continuous communication.
Canvas Usability
User Control and Freedom
Pipeline creation is inherently iterative. Version control support creates a safe space to explore, make mistakes, and easily recover from errors.
Ease of Navigation and Precision
Zoom, pan, and scroll make it easy to navigate large pipelines, while alignment guides help users organize components with clarity and precision.
Connecting & Grouping Components
Jakob’s Law
Follow familiar patterns from workflow and diagramming tools to align with users’ existing mental models and expectations.
Flexibility and Efficiency
Offer multiple ways to connect components, so users can choose the method that works best for them.
Configuring Components
Divide the long form into sections & allow users to visually select their tables to prevent errors
Help users validate their inputs to prevent errors
Allow users to expand the configuration panel
Chunking & Progressive Disclosure
Split long forms into manageable sections and hid advanced settings by default to reduce cognitive load.
Error Prevention
Added validation and test options to catch issues early—like visual table selection to reduce input errors, and checks for text-based fields.
User Control and Freedom
Enabled users to save partially completed forms and return later, supporting non-linear workflows.
THE CURRENT EXPERIENCE
Comments & Annotations
Recognition over Recall
Reduce cognitive load by providing an external memory aid within the interface.
Support for Workflow
Enable contextual documentation of logic and decisions, making it easier to onboard new team members and preserve knowledge.
Support For Efficiency
Fewer context switches and minimizing friction keep users in a state of flow and maintain momentum in complex tasks.
Canvas ↔ Code Flexibility
Power users often prefer coding their pipelines. By enabling seamless Canvas ↔ Code conversion, we reduce context switching and support diverse working styles—aligning with the principle of Flexibility and Efficiency of Use. Users can choose the mode that best suits their task and expertise.
Accelerators for Power Users
Keyboard shortcuts streamline common actions in the Flow Designer, helping experienced users move faster and stay in control.
AI assisted Pipeline creation & Management (Concepts)
AI assisted pipeline creation
The redesigned Flow Designer and interaction patterns lay the foundation for AI-driven pipeline generation and maintenance in future releases. Features like version control, draft-saving, and contextual comments not only enhance usability today—they also provide the structure needed for AI to generate, edit, and explain pipelines in ways that feel natural and trustworthy to users.
05. testing the designs
Guerrilla Usability Testing: Creative Constraints
I ran regular informal testing sessions with customer success team members, bypassing approval barriers to gather quick behavioral insights for some of the less critical changes. We plan to test some of the more critical changes with our customers.
06. stakeholder management
Building Cross-Team Alignment: Align → Connect → Involve
I quickly learned that securing buy-in meant tailoring communication to each stakeholder. Using the Align → Connect → Involve framework (from an NN/G course with Nancy Dickenson), I held regular 1:1s with PMs and engineers to align on priorities. Ongoing feedback with dev teams ensured feasibility, while collaboration with the Product Led Growth team tied our redesign to trial conversion metrics. This approach turned resistance into cross-functional momentum.
My initial pitch for the redesign to the executive team hit an unexpected wall.
Digging deeper, I discovered they were prioritizing AI-driven features to justify costs to investors. I reframed the conversation to show how our redesign wasn't just necessary—it was the foundation for the AI capabilities leadership wanted. By aligning with the company priorities, I secured the support needed to move forward.
07. outro
Impact and Next Steps
The first phase of implementation is currently underway. The phased delivery approach balances quick wins with structural improvements. We're measuring success & failures through:
Trial conversion rates
Time to first successful pipeline
Drop-off points
Support ticket reduction
User satisfaction scores
Reflections
This project reinforced that UX leadership goes beyond design:
Data powers persuasion when stakeholders resist research
It is essential to connect UX improvements to business outcomes
Cross-functional alliance building is essential for success
Adaptability in messaging can make or break executive buy-in
The journey isn't over, but the foundation is set to transform Flow Designer from a hurdle to a true enabler of data integration.