Support Hub
~50%
Of tickets were L1
3
Severity tiers mapped
3
Teams can update support
Replacing a broken help desk with a custom-built support hub that lets customers self-serve, gives teams autonomy over content, and keeps the human connection close at hand.
The problem
- Generic widget on one store, nothing on the other.
- L1 tickets were consuming agent time.
- No structured search or FAQ flow existed.
- Content updates depended on development.
How I approached it
- 1
Severity mapping
Defined the difference between self-serviceable L1 issues and agent-led L2/L3 support.
- 2
Third-party POC
Tested a lighter-weight solution and ruled it out because it would not support the UX or autonomy required.
- 3
Custom structure
Designed a Shopify-native support model with search, topic grouping, and editable content.
- 4
Platform resilience
Scoped the system so the FAQ layer could stay stable while support tooling moved from Zendesk to Sprinklr.
What I changed
- Custom support hub with search and topic categories.
- L1 questions answered through self-service.
- Chat available from the start of the journey.
- Teams can manage support content directly.
Before & after
The support experience moved from broken or absent help to a structured self-service hub that still keeps human support close at hand.
Before
- Generic widget on one store, nothing on the other
- L1 tickets consuming agent time
- No structured search or FAQ flow
- Content updates depended on dev
After
- Custom support hub with search and topic categories
- L1 questions answered through self-service
- Chat available from the start of the journey
- Teams can manage support content directly
The Problem
DFYNE had two broken support states at once. The US store relied on a generic Zendesk widget that pushed simple questions into a ticket flow, while the international store had no FAQ or self-service support at all after the old help desk was removed.
That meant a large share of incoming tickets were basic L1 questions about orders, sizing, shipping, and policies. The challenge was to reduce that support load without making the experience feel like deflection. It still had to feel like DFYNE.
Approach
The project started with ticket taxonomy. I worked with customer care to map L1, L2, and L3 support flows so the information architecture reflected the questions customers were actually asking rather than internal policy structures.
Solution
The final direction is a structured support hub with search, topic-based FAQs, and immediate access to chat. Instead of forcing customers into a ticket flow, the hub answers high-frequency questions quickly while keeping escalation close by.
Try it
What it did
- Reduce avoidable L1 ticket volume so agents can focus on complex issues.
- Keep a clear path back to chat so the support relationship stays personal.
- Make policy and FAQ content resilient as tools change behind the scenes.
- Keep the customer-facing support experience stable during platform migration.