Stack principles before you buy anything
Before adding tools, establish three things: your daily ticket volume, your top three ticket categories (WISMO, returns, product questions — for most stores), and your team size. These three inputs determine which stack layer matters most.
A one-person store handling 15 tickets a day needs a different stack than a 10-agent team handling 500 tickets. Over-tooling for your current stage wastes money; under-tooling means your team drowns when you scale.
Every tool in your stack must integrate directly with Shopify and with each other. A stack where tools do not talk to each other means your agents switch tabs, data does not sync, and escalations get lost. Integration is not optional.
Recommended stack by store size
These are starting points. Your specific ticket mix, existing tooling, and team preferences may shift the selection. What matters is that each layer covers a distinct job and connects to the others.
| Store tier | Daily tickets | Recommended stack |
|---|---|---|
| Solo / early stage | 1-30 | Bookbag (AI agent) + email escalation |
| Growing store | 30-150 | Bookbag + Gorgias (help desk) + Loop Returns |
| Scale / mid-market | 150-500+ | Bookbag + Gorgias Pro + Loop Returns + Klaviyo proactive flows |
| Enterprise / Shopify Plus | 500+ | Bookbag + Zendesk or Gorgias Enterprise + Loop + ShipBob/ShipStation |
The AI agent layer
The AI agent is the front door of your support stack. It handles chat on your storefront, resolves the automatable majority of tickets, and passes the rest to your help desk with full context. This is the layer with the highest ROI because it scales automatically — one agent at 50 tickets a day and 5,000 tickets a day.
Bookbag is built specifically for ecommerce. It connects natively to Shopify, reads order and product data live, and applies your return and shipping policies to resolve tickets autonomously. It does not charge per-resolution, which matters when your goal is to automate as much volume as possible.
- Bookbag: Shopify-native, flat-rate pricing, resolves WISMO, returns, product questions automatically.
- Key criteria for any AI agent: live Shopify data access, policy-based resolution, clean escalation to your help desk, and CSAT measurement.
- Avoid tools that only do FAQ matching without order data — they cannot resolve the tickets that matter.
The help desk layer
The help desk handles everything the AI agent escalates. For growing Shopify stores, Gorgias is the market leader — it was built for ecommerce and has the deepest Shopify integration available. For teams that already use Zendesk or Freshdesk for non-commerce reasons, staying on those platforms is valid.
Gorgias — best for Shopify-first teams
- Native Shopify order data in every ticket sidebar.
- Shopify actions (refund, cancel, duplicate) from within a ticket.
- Ecommerce-specific automation rules (auto-close delivered orders, auto-tag by AOV).
- Pricing: per-ticket, which incentivizes automation — you are rewarded for deflecting tickets to AI.
Zendesk — best for larger or multi-channel teams
- Mature platform with robust SLA management and reporting.
- Shopify integration via app (not native, but solid).
- Better for teams that handle support across products — not just ecommerce.
- Pricing: per-agent seat.
Freshdesk — best budget option
- Lowest cost per agent in the mainstream market.
- Shopify integration via marketplace app.
- Free tier for very small teams.
- Less ecommerce-native than Gorgias but functional.
Returns management
Returns are the second-largest ticket category for most Shopify stores (after WISMO) and the most process-intensive. A dedicated returns tool automates the label generation, tracks return status, and feeds refund approvals back to Shopify — reducing the human time per return significantly.
- Loop Returns: the most widely used Shopify returns app. Supports exchanges, store credit, and carrier-branded tracking pages. Integrates with Bookbag so the AI agent can initiate returns via Loop.
- AfterShip Returns: solid alternative, good for stores already using AfterShip for tracking.
- Shopify native returns: free, functional for simple stores, but lacks the automation and UX of dedicated tools.
Analytics and reporting
Your support stack should tell you three things: how many tickets are coming in and why, how many the AI is resolving, and what your customers think of the experience. Stack these metrics from three sources:
- 1Bookbag dashboard: deflection rate, escalation rate, top unresolved categories, CSAT per conversation.
- 2Help desk reporting: tickets by category, agent handle time, SLA compliance, repeat contact rate.
- 3Shopify analytics: correlate support contact rate with sales events — a spike after a shipping delay is a signal to add proactive messaging.
What to avoid in your Shopify support stack
The market for Shopify support tools is crowded. Several common mistakes lead merchants to overpay or under-deliver.
- Generic chatbots with no Shopify data access: they can answer 'what is your return policy?' but not 'has my order shipped?' — which is most of your volume.
- Per-resolution pricing for AI support: if you are automating 200 resolutions a day, per-resolution fees compound fast. Flat-rate pricing (like Bookbag) aligns incentives.
- Building a stack before you know your ticket mix: spend a week manually tagging tickets before buying tools. The distribution always surprises merchants.
- Duplicate tooling: an AI agent plus a help desk with its own AI features can create conflicting automations and confusing escalation paths. Pick one AI layer.
- Skipping a returns tool if returns are more than 15% of your tickets: manual returns processing is where teams waste the most time per-ticket.
Key takeaways
- Build your stack based on your ticket volume and mix, not what is most popular or most expensive.
- The minimum viable stack: an AI agent (Bookbag) with email escalation. Add a help desk at 30+ daily tickets.
- Gorgias is the best Shopify-native help desk for most growing stores; Zendesk for larger multi-channel teams.
- Loop Returns is worth adding once returns exceed 15% of ticket volume.
- Avoid generic chatbots, per-resolution AI pricing, and duplicate AI tools in the same stack.