As Exxact scaled, its sales team relied on NetSuite for CRM functionality—but the tool was slow, rigid, and misaligned with how sales actually worked. Reps tracked opportunities outside the system, resulting in inconsistent data, missed follow-ups, and unreliable reporting.
I led the design of a custom CRM focused on speed, clarity, and alignment with Exxact’s real sales workflows, reducing data entry time by ~50% and creating a scalable foundation for future systems like OMS and WMS.
Before the CRM, opportunity tracking lived in spreadsheets like this one. They were color-coded by status, maintained manually, and inconsistent across reps.
The project followed a structured design process to align stakeholders and reduce rework.
While several CRMs offered strong individual features, none aligned with Exxact’s sales structure or transaction volume. Most assumed:
After presenting findings to stakeholders, we decided investing in a custom CRM would better support the business long-term.
What we liked
Table-based data views for scanning many records at once
Predefined filters for common workflows
Ability to create entities inline while creating transactions
Custom dashboards and templates
What we wanted to avoid
Kanban boards (50-100 active transactions made them unusable)
Overly complex filtering systems
Where to innovate
Clearly defined sales stages
Status models for people and companies
Support for individuals purchasing without a company
Required loss reasons
Sales inbox and email integration
I interviewed 4 internal stakeholders across sales roles. These insights informed MVP feature prioritization and a post-MVP wishlist.
From research and discovery, I established three guiding principles:
The flow diagram above maps the linear path between people, companies, inquiries, and opportunity creation which includes edge cases where a person buys for a company in one transaction but as an individual in another.
Feature scoping surfaced 96 distinct requirements across 8 workflow areas. Rather than build everything at once, features were grouped by area and prioritized based on research findings, alpha readiness, and development feasibility.
Testing from iteration 1 revealed three requirements for the final pattern: keep critical info on one page, clearly differentiate edit and display modes, and handle edge cases like multiple companies, email addresses, and individual buyers.
The final CRM centered around four core patterns that balanced speed for power users with clarity for new hires.
The prototype launched to 14 sales reps. Junior reps logged leads and created opportunities. Senior reps managed accounts and tracked pipeline progress.
To support early adoption, I built and delivered an onboarding slide deck (Thank you Figma Slides) before the first session, giving reps enough context to push the system rather than just explore it cautiously. This accelerated usability feedback significantly.
A quick look at how this project felt—measured by challenge, teamwork, and personal sanity.
Research
Talking to users changed everything
Without interviews, we would have built a kanban pipeline no one would use. Watching reps manage 450–500 opps a year reframed our direction before a single screen was designed.
Complexity
Scoping is a design decision
With 96 features identified, prioritization required knowing which gaps caused the most daily pain, not just what was most requested.
Iteration
The blueprint matters more than the feature
Getting the core create/edit pattern right before individual workflows saved significant rework and gave the whole team a stable foundation to build from.
Adoption
Design doesn't end at handoff
Building the onboarding deck reminded me that helping reps understand the why behind the new system was part of the design problem, not an afterthought.
Collaboration
Stakeholder disagreement is a feature, not a bug
Tension around density, scope, and terminology forced explicit decisions rather than assumptions that would have caused rework later.
