Case Study: Building Salesforce’s global events platform
by reimagining the Dreamforce attendee experience.
CRM
B2B
Global Events
0→1
Platform
User Research
ROLE
Lead UX Designer
COMPANY
Salesforce
TIMELINE
9 months
TEAM
Solo, 20 Engineers, 12 QA, PO, Scrum
BY THE NUMBERS
77,000
Dreamforce attendees on 0→1 platform
3 to 1
registration, hotel, & payment: unified
24/7
web, tablet, & mobile app community
Visuals
Registration
Registration: Training
Tablet: Sessions
Tablet: My Session Calendar
The Problem
Three systems, one very stressed attendee.
Salesforce needed a new web, iOS, and Android platform for Dreamforce. Registration, hotel booking, session planning, and an event social network had never lived in one place before. In 2011, attendees juggled separate registration and hotel systems with no shared calendar and no on-site companion experience, while San Francisco hotel inventory sold out within hours of registration opening. Attendance was roughly doubling every year, so the fragmentation was only going to get more painful.
Role & Constraints
OWNED
UX design across registration & hotel booking, session planning & calendar, & the Chatter-based event social network, for web, iOS, & Android.
PARTNERED ON
A UX Researcher, 20 Developers, 12 Quality Assurance Engineers, 2 product owners, and a scrum master throughout the build.
CONSTRAINTS
Roughly half of each year's attendees were first-timers, so the flow had to work with zero prior context; ride-sharing barely existed yet.
MID-FLIGHT CHANGES
I started as a contractor, which shaped how much weight my recommendations carried against executive preference early on.
The Approach
4 steps, one vision.
1. DISCOVERY: Tested wireframes with 20 past attendees and catalogued 2011's pain points.
2. DEFINITION: Proposed a multi-screen signup with minimal fields per screen, grounded in established form design research.
3. DESIGN & ITERATION: Built and tested the flow; testers flagged field count as a concern but completed it anyway.
4. DELIVERY: Shipped a single-screen registration with 23 required fields after executive leadership overrode the phased proposal.
Key Decisions
Overruled in year one. Right in year two.
2012
Proposed a multi-step form with minimal fields per screen
Why: Established form design research showed that shorter, phased forms reduce drop-off, and our own research sessions flagged the field count as friction.
Result: Overridden. Leadership wanted every field on one screen, and that's what shipped for 2012.
2013
Repitched the same design, this time with a year of usage data
Why: The 2012 launch gave me real numbers to back the original recommendation instead of just precedent.
Result: Implemented for 2013, saving Salesforce over $100,000 in customer service costs.
Memorable Quotes
“It won’t matter how many fields we throw at them, they’re going to sign up!”
Outcome
What shipped, and what finally landed.
The platform launched in time for Dreamforce 2012, which drew a record 77,000 attendees and, for the first time, ran registration, hotel booking, and payment through one flow instead of three. A year later, the reduced-field registration I'd first proposed in 2012 finally shipped, and it saved Salesforce more than $100,000 in customer service costs. The same data that got overridden in year one became the case for change in year two.
Reflections
Being right in the room isn't the same as being right on the record. I learned to lead with data and precedent before the pitch, not after the pushback, and that staying through the contract-to-full-time conversion was what eventually gave the original recommendation room to ship.