Training for a Company-Wide CRM Migration
Role
Learning Experience Designer (end-to-end owner)
Organization
Wix — Customer Care, Learning & Development
Timeline
14 weeks
Tools
Articulate Storyline, Articulate Rise, NotebookLM, Claude, OpenAI Codex
Manager training for a company-wide CRM migration at Wix, focused on helping managers operate in Microsoft Dynamics and coach their teams through the change.
The Challenge
Wix's Customer Care org was migrating off its legacy CRM onto Microsoft Dynamics. Managers and Team Leads across four functions (Delivery, BPO, PSG, QA) needed more than tool training — they needed to operate in Dynamics and coach their teams through the change.
Two constraints raised the difficulty: there was no prior Dynamics training to build from, and Dynamics was still being actively developed while the course was being designed. There was no stable source of truth for 14 weeks — content had to be revised in real time as features changed.
A generic tool walkthrough wouldn't cut it. It had to be scoped to what managers actually do.
My Role
I owned this end-to-end — problem definition, SME management, design, development, visual direction, assessment, and post-launch measurement. One designer, one course, full accountability for the outcome.
Process
Designing for the manager, not the tool
01
Discovery under a moving target
Structured SME interviews across three stakeholder groups — a BA Team Lead, a Delivery Manager, and the Product team building Dynamics itself. Collaboration ran continuously through meetings, Slack, and iterative draft reviews for all 14 weeks as the product kept changing.
03
Design and build
A 1-hour course in Articulate Storyline and Rise, with custom visual assets built to spec. Success metrics were defined up front — satisfaction ≥80%, pass rate ≥80%, retakes ≤2.0 — before a single slide was built.
02
Designing for the manager layer
The course was scoped to the manager layer: platform proficiency plus the coaching context managers needed to support their teams. I built it around the CCAF framework (Context, Challenge, Activity, Feedback), putting learners into realistic manager scenarios where they had to make decisions, not just click through information.
04
An engineered AI workflow, not casual AI use
With SME input arriving simultaneously across recordings, transcripts, notes, and multiple channels, I built a deliberate multi-tool system:
• NotebookLM as a structured knowledge base — every SME recording, transcript, and note, used to validate content against primary sources.
• OpenAI Codex to draft content and track information across Slack, email, and stakeholder threads.
• Claude, running a self-designed content-review skill, to check every draft against the org's style guide.
I served as editor throughout. Every AI-assisted draft was reviewed and refined against professional instructional judgment before it shipped — the workflow accelerated production, it didn't replace the design decisions.
Results
95%
Assessment pass rate
93%
Learner satisfaction
1.3
Average attempts to pass
Why It Worked
Three decisions did most of the work: designing around CCAF so managers practiced real decisions instead of passively consuming content, building a validation loop (NotebookLM) that kept content accurate while the product was still changing, and treating AI tools as a production system with a human editor at the center — not a shortcut.