S22 · Retention

Customer Support & Experience

Gorgias, AI Support, CSAT, Service Recovery

Section 22 / Retention / by Rob Ward

Founder's Principle

Support Is Where Trust Is Built

At Quad Lock, if a customer had a problem we'd fix it immediately and generously. A $15 replacement part costs you nothing compared to losing a customer who would've bought from you for the next decade. The brands that cheap out on support are optimising for this quarter's P&L at the expense of next year's revenue. Your support team talks to more customers than anyone else in the company. Treat them like the intelligence asset they are.

Key topics covered

The Support Stack by Stage
Launch (Under $1M): Support is you or one other person. Platform: Gorgias (Shopify integration saves meaningful time per ticket). Write macros for your top 10 ticket types in the first month.
Platform Comparison
---
The Top 5 Ticket Types (70-80% of Volume)
---
Macros and Knowledge Base
Build macros for your top ticket types on day one. Write them in your brand voice. A macro that feels personal gets better responses than one from a banking helpdesk.
The Human Side of Support
Your agents need defined limits they can act within without escalating. Slow approvals kill the service recovery moment.
AI in Customer Support (2025/2026 Reality)
---
Key Metrics
These ranges are directional - what's achievable varies by category, product complexity, and team size.
Support as a Retention Engine
At Quad Lock, customers who'd had a problem resolved well often became our strongest advocates - going 180 degrees from frustrated to loyal.

Rob's take

BFCM was the biggest test for support every year. We'd staff up the Business Process Outsourcing (BPO) teams in advance, refresh all macros and rules, and run training on the most likely issues before the sale kicked off. Getting ahead of problems was everything. Every question you could answer in an FAQ or a proactive email was one less ticket in the queue during the busiest week of the year.

On BPO structure: at Quad Lock we placed teams at strategic points around the world to cover 24-hour service. Australia handled escalations and complex issues. Mexico and the Philippines covered volume across the other time zones. That setup meant a customer could reach us at any hour, and the AU team wasn't burning out on overnight shifts. If you're scaling past $10M and selling internationally, a distributed support model is worth evaluating. The cost difference between AU/US wages and offshore teams is significant, and the coverage improvement is immediate.

Support Is a Revenue Channel, Not a Cost Centre

Launch (Under $1M): Support is you or one other person. Platform: Gorgias (Shopify integration saves meaningful time per ticket). Write macros for your top 10 ticket types in the first month. Only open channels you can staff (email + Instagram DMs at this stage). See metrics table below for response targets.

Growth ($1M-$10M): Dedicated Customer Support (CS) person at 50-100+ tickets/day. One rep handles 50-80 tickets/day well; above 80, quality suffers. Add live chat. Start measuring CSAT (Customer Satisfaction Score) on resolved tickets. See metrics table below for targets.

Scale ($10M-$50M): CS is a team with a CS Manager. AI support tools (Gorgias AI) can auto-resolve 20-30% of tickets (built-in, lower barrier). Siena AI worth evaluating - as a dedicated AI CS agent, it handles tier-1 queries autonomously with higher resolution rates (40-60%). Team: CS Manager + 2-4 agents + QA process. See metrics table below for SLA targets.

Established ($50M-$100M): VP or Director of CX, sub-teams by channel, dedicated QA, CS data feeding into product and ops decisions.

Enterprise ($100M+): Full CX organisation with regional teams, specialised functions, and CX as a strategic board-level function.

Platform Comparison

Macros and Knowledge Base

Build macros for your top ticket types on day one. Write them in your brand voice. A macro that feels personal gets better responses than one from a banking helpdesk.

Knowledge base: Customer-facing FAQ that reduces ticket volume 10-30%. Build progressively: every macro answer could be a knowledge base article. Enable early. Your cheapest support interaction is the one that never becomes a ticket.

---

The Human Side of Support

Your agents need defined limits they can act within without escalating. Slow approvals kill the service recovery moment.

Real product knowledge. Every new CS hire should spend their first week using the product, reading every review, and going through the purchase experience as a customer.

Brand voice guidelines. Generic corporate support language is off-brand for every DTC company.

Free access · No subscription

Keep reading in the full playbook.

All 30 sections, the diagnostic Health Check, 400+ checklist items, and 8 tools. Free and always will be.

Open the full playbook

What you'll walk away with

Helpdesk platform set up and integrated with Shopify (Gorgias recommended as default)
Macros written for top 10 ticket types in your brand voice
CSAT tracking live: survey sent after ticket resolution (85%+ target, 76-78% is average)
Self-service returns portal configured (Loop, AfterShip Returns, or similar)
Turn on AI triage for auto-tagging and simple resolutions, or document why not.
Agent authority limits defined: refund thresholds, replacement limits, discount caps documented
Knowledge base / Help Centre published with articles for top FAQ topics
Service recovery process documented: acknowledge → resolve exceptionally → follow up → convert to loyalist
Feedback loops built: Support → Product, Web, Social, and Ops channels live with real-time escalation paths
Response time SLAs set: email under 4 hours, live chat under 1 minute