Desktop App
Figma & Loveable

Purl AI

Initial Hero Image

Project overview

Purl is an AI-powered platform designed to bring structure to the creative process. This project involved creating an end-to-end ideation tool that guides marketing professionals from raw concepts to actionable solutions, validated by a successful MVP launch to potential investors.

Business opportunity

While generative AI presents a massive opportunity, most tools offer a "blank canvas" that overwhelms users. We identified a clear business opportunity to create a platform that provides a structured, guided workflow. This approach helps users harness the power of AI effectively, turning its potential into tangible, high-quality business ideas.

Role & ownership

As the co-founder and sole product designer, I owned the entire product vision and design lifecycle. My responsibilities included foundational user research, persona creation, information architecture, UI design, prototyping, and coordinating all usability testing.

Outcome

The initial MVP successfully demonstrated the product's viability and core value proposition. User testing confirmed that the guided, Kanban-style workflow was highly effective, with participants highlighting the platform's consistency and clear differentiation. The MVP provided a strong foundation and a clear roadmap for post-launch feature development.

Final Designs

No items found.

design process breakdown

Competitive Audit

I conducted an audit of direct competitors (Ideanote, Trello) and powerful, indirect AI tools (ChatGPT, Bard) to map the current landscape. The analysis revealed a distinct market gap: organisational tools lacked deep AI integration, while pure AI tools provided powerful generation but no structure, often overwhelming users with a "blank canvas" problem. This created a clear opportunity for Purl to build a bridge between these two worlds.

Insight 1

Create a Guided, Step-by-Step Workflow: Most AI tools are open-ended and can be intimidating. We saw a major opportunity to differentiate by creating a structured Kanban process that guides users from their initial prompt to a final, analyzed idea. This makes the power of AI accessible and focused, reducing user confusion.

Insight 2

Position AI as an Analytical Partner: Instead of just using AI to generate a high volume of raw ideas, we identified an opportunity to use it for analysis and depth. This meant designing a system where the AI helps users evaluate, categorize, and deepen a curated set of ideas, giving the user control and turning AI into a strategic partner rather than just a content generator.

Insight 3

Leverage the UI to Provide Clarity: We observed that the Kanban board is a strong and familiar mental model for our target audience. This created an opportunity to use the UI itself as a navigational tool, with each column representing a clear stage in the ideation process. This approach provides a constant sense of place and progress, inherently reducing the cognitive load that was evident in competing products.

Problems Identified

No items found.

Wireframing

As we were defining a completely new workflow, I began mapping out the core user flow, to establish a shared vision with stakeholders. The feedback on these was crucial for refining the information hierarchy. From there, I transitioned into low-fidelity prototypes, which were used to test and validate the platform's primary navigation and the intuitive flow between the main Kanban board and the detailed "overlay" views for each idea.

Information to be added

Research Overview

Conducted unmoderated usability studies on the MVP to demonstrate viability to investors and identify improvements. Users completed key ideation tasks, providing feedback on reducing cognitive load, refining user flow, and enhancing AI-human interaction.

Insight 1

Initial information overload was mitigated by segmenting the process into manageable stages.

Insight 2

Users valued the platform's potential, noting confidence in tasks and clear differentiation between human and AI input.

Insight 3

Feedback identified features for post-MVP development, expanding functionality beyond initial scope.

Information Architecture

The information architecture was designed to directly solve the core problems we identified in our research. The structure is built on a simple, user-centric principle: Clarity first, depth second. To achieve this, we abandoned a complex, multi-page approach in favour of a single, unified Kanban board framework. This architecture organises the ideation process into distinct "thinking spaces," corresponding to the stages our research identified as critical to a user's workflow. By using an in-context "overlay" for detailed analysis, the IA keeps the primary view clean and focused, allowing users to progressively disclose information as they need it, which directly addresses the problem of cognitive overload.

Information Architecture

Experimentation

No items found.
Final Hero Image

Lessons learnt

As a co-founder of Purl, this project was a unique lesson in navigating the intersection of design, business strategy, and product management. The challenge was not just to design an interface, but to build a viable product from the ground up, making critical decisions about where to focus our limited resources to create a compelling and fundable MVP.

Lesson 1

The Strategic Value of an MVP Design System. I learned the immense value of building a scalable, React-compatible design system from day one. Even for an MVP, this upfront investment in systems thinking ensured our development was highly efficient and created a consistent foundation that will save significant resources as the product grows.

Lesson 2

The Art of Ruthless Prioritization. A lean startup environment teaches you to be ruthless in your prioritization. Every potential feature had to be justified against our core goal: creating a viable proof of concept. This meant making tough calls to limit functionality, forcing us to focus only on what delivered the most user value and best demonstrated the product's potential to investors.

Lesson 3

The Prototype is a Powerful Business Asset. This project cemented my belief that a high-fidelity prototype is one of the most powerful assets a startup has. Our detailed Figma prototype served a dual purpose: it was our primary tool for user testing and design validation, and it was also our most effective tool for communicating our vision and demonstrating the product's value to potential investors.

next project

SEVolution
PLEASE REACH OUT TO ME
CONTACT ME
Ben.
back to top
Sitemap
homeworksABOUTcontact
social
awwwardsbehancedribbblelinkedininstagram