Industries: Fashion and apparel

Customer service for Shopify fashion and apparel stores

Good customer service for a Shopify fashion store means answering sizing, exchange, return, order status and restock questions quickly and from written policies. Fyncall's AI agent can answer fit questions from your own size guide, look up orders, check stock by variant and start return requests, while exchanges, refunds and anything sensitive go to your team.

Last updated: 19 November 2026 · By

Fyncall wrote this page, and it describes our own product, so weigh it accordingly. Shopify and Meta rules link to their own pages, checked 19 November 2026. Where policies or regulation come up, this is not legal advice.

What do customers ask Shopify fashion stores?

This page covers five groups of fashion support questions: sizing and fit, exchanges, returns, order status and restocks. Each one needs a different source of truth, so it helps to sort them before you write a single reply.

Question typeTypical messageWhere the answer lives
Sizing and fit"I'm usually a 10, which size should I get in this dress?"Your size guide and product descriptions
Exchanges"Can I swap this for a medium?"Your exchange policy, the order and current stock
Returns"How do I send this back?"Your returns policy and the order's dates
Order status"Where is my order?"Shopify fulfilment and tracking
Restocks"When will the black one be back in a small?"Current stock, plus whatever you have decided to tell customers about restocks

We have not included figures for how often each one comes up, because it differs by store. Tag a sample of your own recent conversations first. Our guide to reducing Shopify tickets shows how to do that and match each reason to a fix.

How do you answer sizing questions well?

Answer with measurements, not opinions. A good sizing reply names the size, explains why using the garment's measurements or fit notes, and says what to do if it does not fit.

That means writing down what your team already knows. A useful fit guide covers:

  • Body and garment measurements for each size, in the units your customers use.
  • How each style fits: true to size, relaxed, cropped, or "size down if between sizes".
  • Fabric notes that change fit, such as stretch or shrinkage after washing.
  • Model height and the size they wear, if you photograph on models.
  • What happens if the size is wrong, linking to your exchange policy.

Fyncall's AI agent answers from your knowledge base and your synced Shopify data. The Shopify sync brings in product titles, descriptions, variants, tags and stock. It does not list size charts or metafields among the synced fields, so put your fit guide in a document and link it to each agent. Files can be PDF, Word, text or Markdown, up to 10 MB each. The docs recommend one topic per file and clear headings, so a separate fit guide for each product family (tops, denim, footwear) tends to retrieve more cleanly than one long file. The Free plan holds 5 knowledge base files and Pro has no limit.

A Fyncall agent cannot see how a garment looks on a customer, and it should not guess. If a question goes beyond what your guide says, such as alterations or a fit complaint, it is a good moment to hand over.

How should a fashion store handle exchanges?

Treat an exchange as a return plus a new item. Confirm the order is eligible, check the replacement size is in stock, then process it, usually in Shopify admin.

Shopify describes an exchange as a way to "send a customer an alternative item as part of the return, such as a different size or color", managed from the Orders page in your Shopify admin. The same page says that with self-serve returns, customers can request returns of fulfilled items from their order status page.

Fyncall's documentation does not describe an exchange tool, so here is how the work splits:

  1. The AI agent looks up the order, checks it against your returns policy, and checks whether the new size is in stock.
  2. It can start a return request, or hand over to your team with a summary.
  3. A person on your team creates the exchange in Shopify admin.
  4. If the customer wants a different item rather than a straight swap, the person can use commerce actions after taking over. On WhatsApp, for example, they can send a tappable list of variants and prices, create an order and send a checkout link. Only in-stock variants are shown by default.

Write your exchange rules in exact terms: how many days after delivery, what condition the item must be in, how many exchanges per order and who pays postage. Our FAQ automation guide has examples of vague policy wording and clearer rewrites.

What should you automate, and what should stay human?

Automate questions with a factual answer in your store data or your written policies. Keep anything that needs judgement, money or empathy with a person.

SituationSuggested handlingWhy
Size advice from your guideAIThe answer is in a document you control
Is it in stock in my size?AIStock by variant comes from Shopify
Where is my order?AIOrder status and tracking are synced
Change my address before it shipsAI, with the customer's confirmationThe agent asks the customer to confirm before changing an order
Start a return within policyAIEligibility is checked against your policy
ExchangesPerson, with AI gathering the detailsNo documented exchange tool; done in Shopify admin
RefundsPersonMoney leaves your store, so keep a person on the decision; the AI can check eligibility and status
Damaged, faulty or wrong itemPersonNeeds photos, judgement and often a goodwill decision
Complaints or legal mentionsPersonA sensitive topic trigger can hand these over

You can switch tools off for each agent in guardrails, one by one or a whole group at a time, such as Order Modifications. A store that does not want AI changing live orders can block the Order Modifications group and leave lookups on. The support automation guide goes through this task by task.

How do you answer restock questions?

Tell the customer what you know now, and only promise a date if you have one. The agent can check current stock for each size and colour, and anything about future stock has to come from what you have written down.

Fyncall syncs inventory from Shopify, so the agent can say whether a size is available right now. Its docs do not describe back-in-stock alerts or restock dates. If you want the AI to answer "when will it be back?", add a short document saying what you are willing to share, for example "we do not give restock dates" or a list of styles you reorder regularly. Then add a Must Follow rule such as "Never promise a restock date unless it is in the restock document." Keep that document current, because an outdated restock note becomes a wrong answer.

How do you automate order status questions?

Let the agent look up the order and share the live status and tracking, instead of pasting a generic reply. It finds the order by order number, email or phone, and gives an estimated delivery date and the tracking link.

Because orders, fulfilments and tracking sync from Shopify, the agent works from the same data your team sees. When tracking is stuck or a parcel is marked delivered but missing, that is a handover case. See what WISMO means, our guide to automating order status questions, and ready-made WISMO reply templates.

How do you set guardrails for a fashion store's AI?

Write the rules your best team member already follows, then give them to the agent as guardrails. Fyncall has three kinds: Never Discuss, Never Say and Must Follow, plus a required tone and an optional maximum reply length.

Examples that suit apparel:

  • Must Follow: "Only recommend a size using the measurements in the fit guide."
  • Must Follow: "Do not promise free return postage unless the returns policy says so."
  • Never Discuss: unreleased collections, or other brands' sizing.
  • Never Say: words that do not match your brand voice.
  • Tone: casual on Instagram and more formal on email, since each channel has its own agent.

Before you go live, use the agent's Test sandbox, described in the AI agents overview, to try your awkward real questions: a customer between sizes, an exchange outside the window, a "where is my order" with no order number. Save these as scenarios and run them again after you change a rule.

When should the AI hand over to a person?

Hand over when the customer asks for a person, when they are frustrated, when the AI is not getting anywhere, and on sensitive topics such as complaints. Fyncall lets you set each of these in AI Escalation settings.

You set the frustration threshold and the number of failed attempts before handover. A VIP setting puts high-value customers first when another trigger fires; it does not cause a handover on its own. Outside your working hours the AI can carry on as normal, suggest a ticket, or open one automatically. When a person takes over, they see the full conversation and a short summary, and can release it back to the AI when they are done. Read more about human handoff and our handoff guide.

Each channel also has its own reply mode: Auto-Reply, Draft & Review or Off. Draft & Review, where your team approves each AI reply, is a sensible way to start.

How do Instagram and WhatsApp fit into fashion support?

Shoppers who find a product on Instagram may ask about it there, so DMs and comments become support channels. Fyncall brings Instagram DMs, comments, story replies and mentions, and WhatsApp, into the same inbox as email and web chat.

On Instagram, flows can start from a DM, a story reply, a story mention, or a comment on all posts and reels or one specific item. A flow on one reel could DM sizing details to people who comment "size?". Instagram flows reply by DM, and Instagram automations are listed under Pro. Fyncall's inbox applies a seven-day reply window to Instagram DMs, measured from the customer's last message, with no template fallback once it closes. See Instagram DM automation and our guide to Meta's Instagram DM rules.

On WhatsApp, Meta says a customer service window of 24 hours starts when a customer messages or calls you, resets each time they do, and that outside it you can only reach them with a template message. That matters for exchanges that run over several days: prepare approved templates, such as "your exchange has been sent", before you need them. WhatsApp connects through Meta's Cloud API or Twilio. See WhatsApp customer support and our WhatsApp message templates.

The agent detects the customer's language and replies in it, and it understands voice messages, which helps when a WhatsApp customer describes a fit problem out loud.

What does Fyncall cost for a fashion store?

Fyncall has a Free plan and a Pro plan at $34 a month or $370 a year. Free includes up to 2 team members and 25 AI messages a month; Pro includes unlimited team members and 500 AI messages a month.

An AI message is one reply the AI sends to a customer, and human replies are unlimited on both plans. Pro adds unlimited AI workflows, Instagram automations, unlimited knowledge base files, AI Copilot and ticketing with SLAs. The pricing page has the full list.

Fyncall is the wrong fit if you need phone, SMS or Facebook Messenger support, or if your store is not on Shopify. It is also a newer product, without a long public track record.

Common questions

Can the AI recommend a size?

Yes, if you give it a fit guide. It answers from the documents you upload and your product descriptions. Without a guide, it has nothing reliable to base a recommendation on, and you should block the topic or hand it over.

Can the AI process an exchange or a refund?

Not exchanges: Fyncall's documentation does not describe an exchange tool, so your team creates exchanges in Shopify admin. For refunds, we suggest keeping the decision with a person. The AI can check eligibility, start a return request and check where a refund has got to, and your team issues the refund. Our return and refund reply templates can help them answer consistently.

Can it tell customers when an item is back in stock?

It can say whether a size is in stock now. Fyncall does not document back-in-stock alerts, so restock dates only come from information you add to the knowledge base.

Do I need saved replies as well as AI?

They help. Saved replies are on both plans, and your team can use them for answers you want worded the same way every time. See our Shopify canned responses.

Is there a version of this page for other categories?

Yes. See our pages for beauty and supplements stores.

Ready to try Fyncall?

Install it tonight, let the AI draft your first replies, and switch it on one channel at a time as you start to trust it.

  • Free plan, no credit card
  • 14-day free trial on Pro
  • Unlimited human replies on every plan