All work

B2B · Fintech · 2021–2022

Pentview CRM

A CRM built from scratch to work as one system with a trading platform.

Role
Product Designer, design lead on the CRM. Started as the sole designer, then mentored a second designer.
Team
Product owner, ~8 engineers, 2 QA, 2 designers
Scope
End-to-end — research, information architecture, UI and the design system for the whole product
Impact
  • Sold to several companies
  • 78k+ leads in the system

Context

Pentview already had a trading platform, WebTrader, but client data lived in a separate system with its own logic and no real connection to it. We designed the CRM to work alongside WebTrader as part of the same ecosystem — built in parallel, not bolted on afterwards.

  • One system from day one. The CRM and WebTrader had to work as one from the start, not be stitched together later.
  • Built to last. CRM tech moves fast; the product shouldn't feel outdated in a year.
  • Scale with the business. More leads, more teams — same speed.

Research

22surveys12moderated usability sessions

We talked to users of different CRMs — mostly people with sales and finance backgrounds, aged 21 to 45, with 1 to 7 years of experience. The research combined qualitative and quantitative methods, with sessions run both online and in person.

Key findings

  • Sales leads need clear statistics on sales and leads to track how the team is performing.
  • Sales managers need a fast way to browse clients, leave notes and see clear, easy-to-tell-apart statuses.
  • Leads fall through the cracks — there's no clear way to prioritize tasks.
  • Users asked for call integration with automatic call logging and a real task system — most still tracked work in spreadsheets and sticky notes.

User personas

Portrait photo used for the Michael persona

Michael, 32

Sales manager

Goal
Increase sales efficiency
Pain points
  • Easily distracted
  • Complicated comment system
  • Navigation difficulties
  • Too many bright colors
Portrait photo used for the David persona

David, 35

Sales lead

Goal
Improve team performance
Pain points
  • Hard to track leads and sales data
  • Unintuitive interface
  • Not enough client info
Research questions we asked
  • Walk me through what happens when a new lead comes in — what's the first thing you do?
  • Tell me about a time something slipped through the cracks. What happened, and why?
  • When a client calls and you need their history right away, where do you actually look first?
  • What's the most annoying part of switching between WebTrader and the CRM during your day?
  • If you could change one thing about how you track follow-ups, what would it be — and why does it bug you?

Process

Timeline

  1. Phase 1 Discover

    • Business research & competitor analysis
    • User research
    • User personas
  2. Phase 2 Define & explore

    • Product architecture
    • Sketching
    • Low-fidelity prototype
  3. Phase 3 Test & iterate (iterative)

    • Usability testing with real potential users
    • Iterating on lo-fi
    • High-fidelity prototype
    • Design system
  4. Phase 4 Launch & beyond

    • Launch
    • Support
    • Outcome analysis
    • Roadmap for next releases
An iterative process: we tested with real potential users and refined the design in loops. After launch, the work continued — support, outcome analysis and a roadmap for what's next.

User flow — Michael, sales manager

  1. Log in with 2FA and land on the dashboard
  2. Open the Clients tab and search for a client by name
  3. Change the client's status from their details
  4. Schedule a follow-up meeting with a reminder
Product architecture diagram of the CRM showing many modules, tables and connections; the lower part is blurred.
Product architecture. Every module, table and action mapped before any UI — part of it blurred.
Sketches. Tables, charts and dashboard layouts on paper.
Low-fidelity prototype. Layout and navigation tested before visual design.

Key decisions

Three findings from the moderated usability study that changed the design.

Color scheme Accessibility

Problem
The old status colors failed WCAG contrast and looked too similar — users misread statuses and picked the wrong one.
Solution
A new status palette that meets WCAG AA contrast and is distinguishable at a glance — also for users with color vision deficiencies.

Viewing multiple clients

Problem
Only one client profile could open at a time, in a pop-up — comparing clients meant constantly closing and reopening windows.
Solution
Client profiles now open as separate tabs, so users can switch between them and compare data side by side.
Client profile opened as a full page with tabs for Information, Trades, Meetings, History and Documents, showing personal, address and marketing details. (open full size)
Client profile — sections as tabs (Information, Trades, Meetings, History, Documents).

Lack of focus

Problem
Users struggled to focus — the table felt cramped and key info was hard to find quickly.
Solution
  • Bigger cells — more breathing room, easier to scan.
  • Customizable columns — users rearrange the table to fit how they work.
  • Status moved to the front — it's the first thing people check, so it's the first thing they see.
Before and after of the leads table with highlights. Before: tight rows and the status column at the far right, partly hidden. After: taller rows, draggable column handles, and the status column moved next to the client name. (open full size)
Before (top) and after (bottom), changes highlighted.

Design system

Built from scratch alongside the product, so every new screen reused the same type, colors and components.

Typography

Montserrat

Regular Medium Bold

Colors

  • #2C2C2D

  • #39414E

  • #D6D7D9

  • #5AD4FB

  • #7EBFD3

High-fidelity screens

Outcome

Outcome

The CRM has been sold to several companies and keeps attracting new clients, with 78k+ leads in the system. For a product built from scratch, that traction is the clearest sign it works.

What I learned

This was one of my first big products, and I made my share of mistakes along the way — which is exactly why it taught me so much.

  • Speed and priorities. With a product this size, you can't design everything at once. I learned to work fast, ship in stages and prioritize what matters most to users and the business — then come back to the rest.
  • Launch is the start, not the finish. Getting a product this big close to ideal takes many more iterations after release: support, feedback and a long roadmap.

What's next

The product is large and keeps evolving in stages — after launch we moved into support and planned the next iterations:

  • Keyboard & screen reader support — extend the WCAG work from the status palette to full keyboard navigation and screen reader labels across the table-heavy views.
  • Mobile experience — the table-heavy structure needs a rethought layout for small screens, not a scaled-down desktop version.
  • Data visualization — with 78k+ leads, the analytics layer can be more readable and actionable.