Shopify customer support automation works best in layers. Let Shopify’s own notifications and order status page answer “where is my order?”, let saved replies or an AI agent answer return policy and shipping questions, and keep a person in charge of refunds, damaged items and complaints. Order changes should run only after the customer confirms.
This guide is for Shopify store owners and support leads at small teams. It maps every common support task to the cheapest tool that can handle it, rates the risk, and ends with a 30-day plan. If you have already decided to use AI and want the detailed rollout, read our guide to Shopify AI customer service, which covers guardrails, testing and measurement.
What can you automate in Shopify customer support?
Shopify customer support automation can handle most repetitive answers and a few low-risk order actions. The questions that repeat every day (order status, tracking, shipping times, return policy) follow fixed rules and use data Shopify already holds, so software can answer them. Decisions that cost money or need judgement (refunds, damaged items, angry customers) should stay with a person.
Automation in a Shopify store comes in six layers, from cheapest to most capable:
- Shopify notifications. Order confirmation and shipping emails that go out without anyone touching them.
- The order status page. The page at the end of checkout where customers can follow their shipment.
- Shopify Flow. Shopify’s own automation app: a trigger, some conditions, then actions such as tagging an order.
- Saved replies. Pre-written answers your team sends in two clicks.
- Helpdesk rules. Tags, routing and priorities applied to incoming conversations.
- An AI agent. Software that reads the question, looks up the order and replies in the customer’s own words.
Start at the top. Every question answered by a notification is one that never reaches your inbox, and it costs nothing.
Which support tasks does Shopify already automate for you?
Shopify already sends customer notifications and hosts an order status page. According to Shopify’s notification settings, most notifications are sent automatically, and you choose which events trigger them. The shipping notifications include a shipping confirmation when you fulfil an order, a shipping update when tracking details change, and out for delivery and delivered messages when the carrier reports those events.
The order status page shows shipping updates once you add tracking numbers, so customers can check progress without contacting you. Shopify Flow is a free app on the Basic, Grow, Advanced and Plus plans. Its actions include adding order tags, adding customer tags and sending an internal email, which is enough to flag risky orders for your team. Shopify Inbox is a free storefront chat app with automated responses for FAQs, greetings and order updates. It is a chat tool rather than a full helpdesk.
How do you find which questions to automate first?
Count your own conversations before you buy or switch anything. Your store’s mix of questions is the only number that matters, and it differs by product, price and shipping setup.
- Export your last 100 customer conversations from every channel: email, chat, WhatsApp and Instagram DMs.
- Read each one and give it a single reason tag: order status, tracking, shipping times, return policy, address change, cancellation, refund, damaged item, product question, complaint or other.
- Count the tags and sort them from most to least common.
- Mark each reason as “rule-based” (the answer follows a written policy or Shopify data) or “judgement” (a person has to decide).
- Automate the most common rule-based reasons first. Leave judgement reasons with your team.
A hundred conversations is usually enough to see a pattern in a small store, and it can be done in an afternoon. If one reason dominates, start there. If your counts are spread evenly, start with the reason that has the simplest, most stable answer. Our guide to reducing Shopify tickets shows how to match each top reason to a fix and measure the change afterwards.
While you read, note any answer your team gives differently each time. Those are policy gaps, and automation will repeat them faster. Write the policy down first. “Items can be returned within 30 days of delivery, unused, with tags attached” can be applied by software; “we’re flexible on returns” cannot.
Why should order status (WISMO) be the first automation?
Order status questions, often called WISMO (“where is my order?”), should come first because the answer already exists in Shopify. The order, its fulfilment status and the tracking number are all in your admin. Nobody needs to make a decision, so the risk of a wrong answer is low.
WISMO automation has two levels. The free level is to make sure every fulfilled order has a tracking number, the shipping confirmation and shipping update notifications are switched on, and the customer can find the order status page easily. That alone can stop some customers from asking.
The second level handles the customers who ask anyway, often on WhatsApp or Instagram rather than replying to the email. An AI agent connected to Shopify can find the order by order number, email or phone, and reply with the status, the tracking link and the estimated delivery date. Because the answer comes from live order data, it stays accurate as the shipment moves.
Order status questions are also a good first test of an AI agent, because you can check every answer against the Shopify admin in seconds. For the notifications, tracking links and WISMO replies in more detail, see our post on Shopify order status automation.
Which order actions are safe to automate, and which need a person?
Read-only tasks such as order lookups and policy answers are safe to automate. Changes to an order are safe only when the order has not shipped and the customer confirms the change. Anything involving money back or a judgement about fault should keep a person.
| Support task | Cheapest layer that automates it | Risk level | Keep a person? |
|---|---|---|---|
| Order lookup | Order status page; AI agent for chat and messaging | Low | No |
| Tracking link | Shipping confirmation and shipping update notifications | Low | No |
| Shipping times | Published shipping page, saved reply or AI agent | Low | No, once the policy is written |
| Return policy | Published policy, saved reply or AI agent | Low | No, for the policy itself |
| Address change, unfulfilled order | AI agent, after the customer confirms | Medium | Review at first, then no |
| Cancel an unfulfilled order | AI agent, after the customer confirms | Medium | Review at first, then decide |
| Refund request | AI agent checks eligibility; a person issues the refund | High | Yes |
| Damaged item | Tag and route, saved reply asking for photos | High | Yes |
| Complaint | Escalation to a person | High | Yes |
Two rules sit behind this table. First, an action that changes an order should never run on a guess: the agent states the change and waits for a clear yes. Second, the order must not have shipped. Once a parcel is with the carrier, an address change or cancellation is no longer a simple edit.
Refunds are the clearest line. Software can check whether an order is inside the return window and whether a refund is already in progress. Deciding to send money back belongs to your team, because it depends on context the order data does not show.
Damaged items and complaints need a person for a different reason. The customer wants to be heard, and a fast but generic reply can make things worse. Automation can still help here by tagging the conversation, asking for a photo of the damage and moving it to the top of the queue.
How do tagging, routing and saved replies fit in?
Tagging, routing and saved replies are the non-AI middle layer. They do not answer customers on their own, but they cut the time each conversation takes and make sure the right person sees the urgent ones first.
- Tags record why a customer got in touch. Shopify Flow can tag orders and customers in Shopify (for example, a high-value order), and a helpdesk can tag conversations by reason.
- Routing sends tagged conversations to the right place, such as damaged-item reports to the person who handles claims.
- Saved replies cover answers that need a human touch but follow a pattern, such as asking for photos of a damaged item or explaining your return steps.
Saved replies are also a good source for an AI agent later. If your team already sends the same return-policy reply ten times a week, that text is a tested answer you can add to the agent’s knowledge base. Detailed triage and SLA set-up are covered in our guide to automating a Shopify helpdesk; keep this layer simple at the start: a handful of reason tags, one rule for urgent conversations, and saved replies for your top five questions.
Does automation work on WhatsApp and Instagram as well as email?
Yes, but Shopify’s built-in layers mostly cover email and the storefront. Shopify notifications go out by email and the order status page lives on your store. Customers who message you on WhatsApp or Instagram need a tool that connects to those channels and to your Shopify orders.
Each messaging channel has its own reply window. According to Fyncall’s channel docs, WhatsApp enforces a 24-hour customer service window that starts from the customer’s last message; after it closes, only pre-approved template messages can be sent. Instagram allows replies for 7 days from the customer’s last message, with no template fallback. An automation plan should account for both, so a customer who asks on Saturday night still gets a reply in time.
The practical test is simple. Pick the channel where most of your audit’s order status questions arrived, and make sure your automation answers there, not only on email.
What does Shopify customer support automation cost?
The Shopify layers cost nothing extra: notifications and the order status page come with your store, and Shopify Flow and Shopify Inbox are free apps. Cost enters with helpdesks and AI agents, which charge in different units:
- Per ticket. You pay for each conversation the helpdesk handles. Gorgias prices its helpdesk plans by tickets per month.
- Per resolution. You pay each time the AI handles or resolves a request. Gorgias charges per automated interaction for its AI Agent once the interactions included in your plan are used, and each of those interactions also counts as a helpdesk ticket.
- Per seat. You pay for each agent who logs in. Zendesk prices its plans per agent.
- Flat. You pay one plan price with a set allowance included.
(Checked Nov 2026; plans change, so confirm on each vendor’s site.) Per-ticket and per-resolution pricing rise with your volume, so a busy sale week costs more. Per-seat pricing rises with your team size. To compare tools on your own numbers, use the conversation count from your audit and our side-by-side pricing comparison.
Fyncall uses flat pricing. The Free plan is $0 and includes 25 AI messages a month. Pro is $34 a month and includes 500 AI messages a month and unlimited team members. An AI message is one reply the AI sends to a customer; replies from your team never count. There are no per-ticket or per-resolution fees. Remember costs outside the subscription too: WhatsApp messages are charged by Meta or Twilio, not by your helpdesk.
What does a 30-day automation plan look like?
A 30-day plan moves from free fixes to AI, one week at a time, so each step is tested before the next.
- Week 1: audit. Export the last 100 conversations, tag each by reason and count them. Write down your shipping times, return window and cancellation rules as plain sentences.
- Week 2: notifications and tracking links. Check that shipping confirmation and shipping update notifications are on, that every fulfilled order has a tracking number, and that the emails point to the order status page. Set up saved replies for your top five questions. Add one or two Shopify Flow workflows if you need to tag risky orders.
- Week 3: AI on one channel in Draft & Review. Connect an AI agent to the channel with the most order status questions. In Draft & Review mode, the AI writes each reply and your team reads, edits and sends it. Start with read-only tasks: order lookups, tracking and policy answers.
- Week 4: widen. Once the drafts are consistently right, switch that channel to Auto-Reply so the AI sends replies on its own, then repeat Week 3 for the next channel. Turn on order changes only when you are happy with how the agent asks for confirmation.
Keep refunds, damaged items and complaints with your team throughout. For the full rollout checklist, including what to test and how to measure results honestly, see our Shopify AI customer service guide.
Where does Fyncall fit?
Fyncall covers the helpdesk and AI layers of this map for Shopify stores. It is an AI helpdesk with one shared inbox for WhatsApp, Instagram, email and web chat, Shopify order details beside every conversation, and an AI agent for each channel you connect. You can read more about the Fyncall helpdesk for Shopify and how it connects to Shopify.
The Fyncall AI support agent can take these Shopify actions:
- Look up orders and their status by order number, email or phone.
- Share the tracking link and estimated delivery date.
- Update a shipping address, add an order note or cancel an order that has not shipped, only after the customer confirms.
- Check refund eligibility and refund status. Issuing the refund stays with your team.
Each of these tools can be switched off. In the agent’s Guardrails tab, Tool Access Control lets you turn off a single tool or a whole category, so you can keep order lookups on while blocking order changes. Reply modes are Auto-Reply, Draft & Review and Off, set for all channels with an override per channel. AI Escalation hands a conversation to your team when a customer asks for a person, shows frustration, when the AI fails to resolve the issue after several attempts, or when sensitive topics such as legal mentions and complaints come up. Your team gets the conversation with a summary, and the AI keeps helping the customer while they wait. Our post on Shopify AI agent handoff covers the triggers and rules in more detail.
Guardrails (including switching individual tools off), escalation settings, Draft & Review and saved replies are all on the Free plan. AI Copilot, which drafts replies and looks up orders for your team, and ticketing with SLAs are on Pro.
Fyncall has limits worth knowing before you choose. It covers WhatsApp, Instagram, email and web chat, but not phone calls, SMS or Facebook Messenger. It works with Shopify only. If phone or SMS support matters to your customers, a tool that offers those channels will suit you better.
Common questions
Can Shopify automate customer service without a third-party app?
Partly. Shopify sends order and shipping notifications automatically and hosts an order status page with tracking, and Shopify’s free Flow app can tag orders and customers. Answering questions in chat, WhatsApp or Instagram needs a chat app such as Shopify Inbox or a helpdesk.
What is WISMO automation?
WISMO stands for “where is my order?”. WISMO automation means answering order status questions without a person, first through shipping notifications and the order status page, then with a chat tool or AI agent that looks up the order and sends the tracking link.
Should an AI agent issue refunds?
We recommend it does not. An AI agent can check whether an order is eligible for a refund and whether one is already in progress, which saves your team a lookup. The decision to send money back should stay with a person. Fyncall’s AI checks eligibility and status only.
How many conversations do I need before automating?
There is no minimum. Even a store with a few conversations a day benefits from switching on notifications and writing clear policies. An AI agent becomes worth testing once the same rule-based questions take up a noticeable share of your team’s week, which your 100-conversation audit will show.


