Closr

I'm currently designing and building Closr — an AI-powered toolkit that brings structure and clarity to complex B2B sales. From zero to MVP, I'm shaping the user experience, design system, and product strategy to support methodology-driven selling.

Role:

Founder & Designer

Duration:

On going

Type:

SaaS Product

Skills:

Research, Design & Build

Part 1/4

Project overview

Closr began as a side project driven by frustration. In my previous career consulting on B2B sales initiatives, I saw how difficult it was to scale consistent, effective sales behaviours across teams. Most tools captured data or pushed leads through rigid funnels — but few supported the nuanced, high-context work of actually closing deals. I led the research, designed the interaction model and UI, and built the MVP using Loveable. Still in its early stages and developed alongside my full-time UX role, Closr is a focused attempt to bring structure and clarity to the messy reality of B2B selling.

Tools used

Figma

Loveable

Google Forms

VS Code

User research

With five years in B2B sales before moving into design, I knew the space was complex. I narrowed the focus by speaking with sales leaders at PwC, Askable, and agencies to uncover the key blockers to closing deals.

Qualification: Teams lacked a reliable process to identify strong opportunities early.

Engagement: Reps struggled to involve the right decision-makers at the right time.

Context: Long sales cycles led to fragmented communication and forgotten deal history.

Problem statement:

Sales professionals often juggle multiple tools to manage deals, track accounts, and apply methodologies—resulting in fragmented workflows, lost context, and inconsistent execution. This lack of

User story:

I’m managing multiple complex deals at once, and it’s hard to stay on top of what each buyer needs at every stage. Without timely, relevant guidance, I risk missing critical steps—costing me the deal, my quota, and ultimately, my commission.

Problem synthesis

With limited time and resources, I had to be highly targeted in how I framed the problem space. I prioritised problems where the user impact was high but build complexity was low — aiming for the simplest solutions to the most painful issues.

Creating simple criteria to score problems based on impact and effort

Prioritising features that could be designed and built quickly without sacrificing value

Writing user stories to validate assumptions and keep scope grounded in real sales workflows

Experience design

This was a 0–1 product, so I defined the core user flows and interaction models — from initial concept through to wireframing and UI design. Each flow was built to minimise friction, reduce cognitive load, and mirror the real-world behaviours of B2B sales professionals.

Product build

I used Loveable to build the product and Supabase for authentication and data storage. This setup allowed me to store real deal data and associate it with specific user IDs — a non-negotiable foundation for any functional sales tool. While not the flashiest part of the work, it was essential to delivering a usable, end-to-end experience.

Part 2/4

Design process

Translating insights from the problem space into a functioning MVP required a structured, iterative approach — especially when working within the constraints of tools like Loveable.

I moved quickly between flows, prototypes, and builds, constantly testing and refining based on technical feasibility. This process deepened my appreciation for the engineering side of product design and taught me how to prioritise when managing a lean product backlog.

Discovery

Research, problem validation & backlog mangement

Design

Information architecture, usability & UI design

Delivery

Requirement documentation, design QA & testing

"When I’m managing a complex B2B deal with multiple stakeholders, I want to track progress against key qualification criteria and next steps, so I can stay focused, avoid context loss, and improve my chances of closing the deal."

Capturing user context (JTBD)

The Jobs to Be Done framework is my go-to for managing lightweight user requirements. It helps me consolidate user tasks into clear, outcome-driven statements — especially useful when scoping MVP functionality and staying focused on real-world needs.

User flows

Authentication

Email/password login, session management

Dashboard

Email/password login, session management

Deal Management:

Create/edit deals, track stages, score confidence

Methodologies:

MEDDIC, SPIN support, with custom fields

Tasks:

Turn risks into actions, track task completion

User flows

Create Deal → Add info → Select methodology → Save

Assess Deal → Update fields → View confidence → Get AI tips

Manage Risk → Identify risks → Create tasks → Track progress

Tech Stack

Frontend: React

Backend: Supabase (auth + data)

Data: Real-time updates

Frontend: ShadCN

Success metrics

Increase close rates

Methodology usage

Time to identify risk

Confidence score accuracy

Product requirement scoping

To keep the MVP focused, phase one will deliver a streamlined experience that guides sales reps through an end-to-end opportunity, supported by a simple, contextual recommendations engine (V1). In phase two, we’ll enhance the engine to offer smarter, more personalised insights based on richer context.

Designing the interaction model

Once I had defined the MVP scope, I designed the interaction model to support it. This involved exploring relevant design patterns, mapping key user flows, building decision trees, and accounting for potential edge cases — all to ensure the experience felt intuitive, flexible, and grounded in real sales behaviours.

User Interface design

After stress-testing key user journeys, I had enough confidence to begin designing screens. To accelerate the process and reduce build complexity, I leveraged ShadCN and UntitledUI components — allowing me to focus on crafting the interaction layer while keeping the UI consistent, accessible, and easy to implement. I customised components with tailored variants, tokens, and interaction states to reflect the needs of a sales-focused product, while maintaining design flexibility for future iterations.

Platform build

After adding detailed design specifications to reduce ambiguity in Loveable, I was able to generate a full-stack web app with a functioning front and back end. While the MVP is still rough around the edges and needs refinement before public release, it successfully lays the groundwork for a scalable, end-to-end product experience.

Frontend: React + Tailwind (via ShadCN UI)

Backend: Node + Supabase (PostgreSQL)

Auth: Supabase authentication with user ID-linked data

Build: Visual logic, custom components, and deployment handled within Loveable’s platform

Part 3/4

Product features

Closr is a side project I created out of frustration. While working as a consultant involved in B2B sales processes, I saw first-hand how difficult it was to scale effective sales behaviours across teams. The tools available — CRMs and lead management platforms — either captured data or pushed leads through funnels, but neither supported the nuanced work of actually closing a deal.

User authentication & state management

UX goes beyond the screen — it’s also about preserving context. Using Supabase, I implemented secure login, persistent sessions, and state management to maintain deal context between visits.

Deal management

Closr allows users to manage the full sales lifecycle — from prospecting to close. While not a revolutionary feature, it establishes a solid foundation for future enhancements like LLM-powered insights and actionable prompts.

Contextual sales guidance

Fully functional AI-driven sales guidance is still a way off. But, I was able to build an MVP state where user actions contributed to a dynamic confidence score - providing contextual sales guidance to mitigate risk and support deal progression.

Part 4/4

Next steps

My goal for Closr is to design, build, and iterate my way to 100 paying customers. To get there, I’m tightening the loop between user feedback and feature development to validate value quickly. I’ll begin by onboarding early users with free access in exchange for usage data and insights. Alongside this, I plan to refine the product’s visual language and interaction patterns — aiming to strike a better balance between creativity and execution speed.

Product feature validation

Enhancing visual language and interaction patterns to create a more polished and cohesive user experience.

Product UI improvement

Running lean experiments to test value fast — tightening the loop between feedback and feature development.