Start with the question behind the question
You are not really shopping for a chatbot. You are deciding how much of your support load to move off your calendar, at what cost per ticket, and whether that tool can see enough of your business to answer well.
"Support agent" quietly means two different products. One resolves conversations inside a single inbox. The other is an AI employee that works across your tools — and it answers support with context the inbox-only tools never have.
This guide gives you the honest comparison the listicles skip: what each layer costs per ticket, where each one fails, and how support ties back to per-order profit.
The two layers you are actually choosing between
Layer one: the single-surface support agent
This is the mature category. Order-status, returns-policy, and tracking questions resolve reliably because they come straight from structured data.
Pricing has moved to outcomes, not seats. Gorgias charges about $0.90 per resolved conversation on most plans, billed only when "the AI resolves a customer conversation entirely on its own," per Gorgias's own pricing explainer. Zendesk prices its AI agents "based on the successful outcomes they deliver," with Suite plans starting around $55 per agent per month on Zendesk's pricing page; third-party reporting puts the per-resolution rate near $1.50, per eesel's pay-per-resolution guide.
Both build in a handoff: the agent escalates what it cannot resolve to a human. That handoff is the honest admission at the center of the business model — these agents do not handle everything, and you are not billed when they punt.
Layer two: the AI employee
The newer layer works across your store the way a human hire would — reading the ad accounts and the store and the email platform together, then taking multi-step action. Analysts call the underlying capability agentic AI: a system that can "act in the real world and execute multistep processes," per McKinsey's definition as cited by Solo.io.
The scope is the whole point. A support widget that can only tell a customer how to request a refund is a chatbot; a system that can issue the refund in Shopify, check the supplier status, and log the outcome is an agent.
PodVector AI's Victor is an example of this layer. Victor integrates with Shopify, Meta Ads, Google Ads, Printify, Printful, Gelato, and Klaviyo; he computes true per-order profit, delivers reports to your Google Drive, and drafts customer-support email grounded in the actual order. The AI employees for ecommerce guide walks the full category distinction if you want the longer version.
The per-ticket math (this is the part the listicles skip)
Say your store handles 300 support conversations a month — mostly order status, returns, and product questions. A human agent takes roughly 8 minutes a conversation. Assume the AI fully resolves half (150) and a person handles the rest; that split is an assumption for the arithmetic, since Gorgias itself says the real rate "emerges from usage over time" in its pricing explainer.
Option A — a human virtual assistant handles everything. 300 conversations × 8 minutes = 40 hours. At a mid-level offshore VA rate (about $6–$10/hour for 1–3 years' experience, per DDIY's Filipino VA rate breakdown): 40 × $8 = ~$320/month. At a fully-loaded US rate of $28–$65/hour, per CallForce's VA rate guide: 40 × $40 = ~$1,600/month.
Option B — AI resolves Tier-1, a human takes the rest.
150 AI resolutions × $0.90 = $135, plus the helpdesk subscription. The remaining 150 conversations × 8 minutes = 20 human hours → $160 offshore or $935 US-hybrid** — and the Tier-1 coverage runs on the clock the model keeps, not a timezone.$800 US. Total: **$295/month offshore-hybrid** or **
Two honest readings. Against a US cost baseline, per-resolution AI is dramatically cheaper on routine volume. Against a $6–$10/hour offshore VA, the pure dollar gap on 300 tickets is small — the stronger case there is instant response and zero management overhead, not price.
And the math scales differently. Doubling ticket volume doubles VA hours (and eventually forces a second hire plus managing them); it only linearly increases per-resolution fees, with no hiring step.
Where every support agent still fails
Ambiguous, high-stakes answers. Air Canada's chatbot invented a bereavement-refund policy; a British Columbia tribunal held the airline liable for the misinformation and ordered it to pay CA$812.02, rejecting the "the chatbot is a separate legal entity" defense, per CBC's reporting and the American Bar Association's write-up. You own what your AI tells customers.
Scope it cannot see. A helpdesk agent reads the inbox and little else. It cannot tell you that the ticket spike this week traces to a supplier delay on one SKU, because it never sees Printful; it cannot connect a refund surge to a losing ad set, because it never sees Meta.
Novel strategy. Gartner's warning is blunt: current models "don't have the maturity and agency to autonomously achieve complex business goals or follow nuanced instructions over time," per its 2025 projection that over 40% of agentic AI projects will be canceled by end of 2027. The same release coined "agent washing" — rebranding plain chatbots as agents — and estimated only about 130 of thousands of self-described agentic vendors are real.
How to actually pick
Match the tool to the job, and keep both of Gartner's numbers in view: the category is real — Gartner expects agentic AI to resolve 80% of common customer service issues by 2029 — and it is also the most over-labeled software on the market.
- Pick an outcome-priced support agent if your need is Tier-1 inbox volume and nothing more. Confirm it is native to Shopify and that you pay per resolution, not per seat. Gorgias claims it powers "customer conversations for 40% of Shopify brands" on its pricing page — a vendor claim, but a signal of Shopify-first depth.
- Pick an AI employee if you want support answered and connected to the rest of the operation — the same system that drafts the reply also knows the order's true margin and the supplier's status. That is the cross-tool scope covered in the operational-efficiency breakdown.
- Favor tools whose work product lives in your accounts — your Shopify, your Klaviyo, your Drive. If more than 40% of agentic projects get canceled by end of 2027 per Gartner's cancellation forecast, the artifacts should survive the vendor.
Why support should not be a silo
Here is the operator's case for the AI-employee layer. A refund is not just a support event — it is a hit to the per-order profit on that SKU, and sometimes a signal that an ad is driving the wrong buyers.
Victor sees every order, every ad dollar, and every supplier cost together, so he can tell you the true profit on each product and draft the support email with that context already in hand. He also makes the downstream changes a siloed chatbot cannot: he pauses the losing ad set in Meta, reprices the thin-margin SKU in Shopify, and builds the Klaviyo flow that pre-empts the "where is my order" ticket in the first place.
Victor is not a dashboard and not a support widget. He is the employee who ties the inbox to the P&L — which is exactly the connection the operational-efficiency breakdown argues the owner should stop carrying by hand.
If you want to see what an AI employee does with your own store, ad, and supplier data, meet Victor and connect your store.
FAQs
What is the best AI support agent for a Shopify store?
For routine, high-volume support, the best choice is an outcome-priced agent that is native to Shopify and bills per resolved conversation — Gorgias and Zendesk both price this way, per Gorgias and Zendesk. If you want support answered in the context of your orders, ads, and supplier costs, the right tool is an AI employee that works across those tools, not an inbox-only chatbot.
How much does an AI support agent cost per ticket?
Expect roughly $0.90 per resolved conversation on Gorgias and near $1.50 on Zendesk, per Gorgias's explainer and eesel's reporting. You pay only when the AI fully closes the ticket; escalated conversations route to a human and are not billed as resolutions.
Is an AI support agent cheaper than a virtual assistant?
Against a fully-loaded US VA at $28–$65/hour, per CallForce, AI is far cheaper on routine volume. Against a $6–$10/hour offshore VA, per DDIY, the dollar gap on a few hundred tickets is small — the real wins are instant response and no management overhead.
Who is liable if the AI gives a customer the wrong answer?
You are. A tribunal held Air Canada responsible for its chatbot's misinformation and dismissed the "separate legal entity" defense, per CBC. That is why agents grounded in your live store data — rather than a model guessing policy from memory — matter, and why review stays part of the job.
What is the difference between a chatbot and an AI employee?
A chatbot converses and resolves requests on one surface. An AI employee takes multi-step actions across several tools toward a goal — the dividing line in every analyst definition, per the McKinsey definition via Solo.io. Gartner calls rebranding the former as the latter "agent washing," per its 2027 forecast.
Can one AI handle both support and the rest of my store?
Yes — that is what the AI-employee layer is for. Victor drafts customer-support email and also works in Shopify, Meta Ads, Google Ads, and Klaviyo, so the same system that answers a ticket can reprice a SKU or build a flow. The AI employees for ecommerce guide maps which parts of the operation each layer covers.