AI assistant for retail store owners: stock questions, collection and opening hours

Saturday, half past eleven. There's a queue of four at the till, someone wants a gift wrapped, a customer is holding up two jumpers asking which looks better, and the shop phone is ringing behind the counter. You know what it is before you pick up. "Have you got the green one in a size 12?" "Are you open on the bank holiday?" "I ordered something for collection, is it ready?" Meanwhile three Instagram messages and a WhatsApp have come in asking much the same thing.

An AI assistant for retail store enquiries takes those questions off the counter. It answers the phone, WhatsApp, Instagram and your website chat from the information you give it: your products, your hours, your returns policy. When something needs you, it writes it down and flags it, and you deal with it when the queue has gone.

This page is about independent shops: boutiques, grocers, florists, bookshops, gift shops, sports shops, tailors and pharmacies. It covers the questions shops actually get, what the AI can be trusted with and what it should pass to you, how accurate stock answers can be, the consumer law it mustn't get wrong, and what it costs, with the assumptions spelled out.

Where an AI assistant for retail store owners fits

Most shop enquiries fall into a small number of patterns, and most of them don't need you personally. They need a correct answer, quickly, at a moment when you're busy with a paying customer.

Klugon covers the channels shoppers use:

• Phone. Your existing number forwards to Klugon, for every call or only when you don't answer. The AI talks to the caller in an AI-generated voice and each call is saved to your inbox. More on how that works on the AI phone receptionist page.
• WhatsApp. Shoppers who'd never ring will happily message. The WhatsApp AI chatbot answers on your own WhatsApp Business number.
• Instagram and Facebook Messenger. For boutiques and gift shops, Instagram DMs are often the busiest channel of all.
• Your website. A chat window, added by pasting one line of code. Plus a public Klugon page with your products and a QR code for the shop window, so people walking past after hours can ask a question.

Everything arrives in one inbox. You don't need to check four apps at the end of the day.

What the AI does in each case is the same. It looks things up in your catalogue and your knowledge base, gives an answer, and either completes the job or hands it to you. It isn't a till, a stock system or a web shop. It sits in front of those and deals with the questions.

The calls and messages shops get

Every shop is different, but the same handful of questions turns up in all of them.

"Have you got it in stock?"

The most common question, and the hardest one to answer well. "Do you have the Le Creuset casserole in blue?" "Have you got the new Richard Osman in?" "Size 9 trail shoes, wide fit?" "Is there any sourdough left?"

The answer depends on what's on the shelf right now. The AI can only answer from what Klugon knows, and Klugon's catalogue is the one you keep, not your till's. The stock section below goes through what that means and how shops handle it.

Opening hours and holiday hours

"Are you open Sunday?" "What time do you close tonight?" "Are you open on Boxing Day?" These are easy, as long as you keep holiday hours current. Most shops change their hours around Christmas, bank holidays and the summer, and the AI needs those changes before customers start asking.

Click-and-collect and reserve-and-collect

"Can you put one aside for me?" "I'll pick it up tomorrow, can you hold it?" "My order says ready for collection, when can I come?" Independent shops often do this informally: someone rings, you put the item behind the counter with a sticky note. The AI can take that request properly, with name, item and collection time, and flag it. The section on collection below explains the options, including one that uses Klugon's booking slots.

Returns, refunds and exchanges

"I bought this as a present, can I exchange it?" "It's faulty, can I get my money back?" "I've lost the receipt." These are where the AI has to be most careful, because consumer law gives customers rights your goodwill policy can't take away. It can explain your policy and the basics of the law accurately. It shouldn't decide a disputed refund.

Gift vouchers, gift wrap and special orders

"Do you sell gift cards?" "Can you gift wrap it?" "Can you order in a book you haven't got?" "Do you do engraving?" "Can I get it in a different colour?" Straightforward questions, mostly about your services. Special orders usually need you to check with a supplier, so the AI takes the details and passes them on.

Appointments some shops take

Plenty of retailers book time as well as selling things: alterations and fittings at a tailor, wedding consultations at a florist, bike or racket servicing at a sports shop, gait analysis at a running shop, personal shopping at a boutique, consultations at a pharmacy. Klugon's booking engine handles these well, because they're appointments with a length and often a named person.

Delivery, parking and the practical stuff

"Do you deliver locally?" "Where can I park?" "Do you take American Express?" "Is there step-free access?" "Can I bring my dog in?" Quick answers, asked over and over.

What doesn't need answering

Sales reps, recruitment agencies, energy brokers and the people selling you card machines. Tell the AI to take a message.

When the questions come

Shop enquiries have their own rhythm. Weekday mornings bring practical questions: are you open, is my order ready. Lunchtime and the hour after work bring stock questions from people deciding whether a trip is worth it. Saturdays are busy in the shop and on the phone at the same time, which is the problem. Evenings and Sundays bring messages, from people browsing on their phones after the shop is shut.

The retail year is even more uneven:

• November and December: Black Friday, then Christmas, when stock questions, gift-wrap requests and "will it arrive in time" messages pile up.
• Boxing Day and January: returns, exchanges and sales questions.
• February: Valentine's Day, the peak of the year for florists.
• March: Mother's Day, for florists, gift shops and cafes in shops.
• Late August and September: back to school, for sports shops, shoe shops and uniform stockists.
• Spring and early summer: weddings, for florists, tailors and boutiques; gardening and outdoor gear for others.

If you know your peak, update what the AI knows a few weeks before it, not on the day.

What the AI handles well

Given accurate information, these are the jobs it's good at:

• Product questions from your catalogue: price, description, sizes and colours you've listed, and whether you've marked it as in stock.
• Suggesting related products, such as a matching belt or a vase to go with the flowers, if you switch that on.
• Opening hours, holiday hours, location, parking, accessibility and payment methods.
• Your policies: returns, exchanges, gift receipts, delivery areas, gift wrap, vouchers, special orders.
• Booking appointments with the right person for the right length: fittings, consultations, servicing.
• Taking reservation and collection requests and flagging them for you.
• Capturing leads for special orders and larger jobs, with contact details and notes.
• Answering in other languages. On the phone that's English, German, French, Spanish, Italian, Portuguese, Dutch, Japanese and Arabic. In a town with lots of visitors, that's useful.
• Answering when the shop is shut. Shoppers browse in the evening. A quick answer at 21:00 can be the reason someone comes in on Saturday instead of ordering online.

It also gives you a record of what people ask. After a few weeks, the insights page shows the most asked questions and the ones the AI couldn't answer. For a shop, that's free market research: if twenty people asked about a brand you don't stock, you might want to stock it.

What it hands to a person

Some conversations need you, and you should decide which before you go live.

• Faulty goods and disputed refunds. The AI can explain your policy and that customers have legal rights. It shouldn't refuse a refund or argue about whether something is faulty.
• Complaints. Take the details, apologise, flag it.
• Stock it can't confirm. If the catalogue doesn't say, or stock is the kind that changes by the hour, it should say it'll check and flag it rather than guessing.
• Special orders and price quotes that depend on a supplier.
• Questions about medicines or health at a pharmacy. Those go to a pharmacist, always.
• Age-restricted products. The AI can say you sell them and that ID will be checked; it shouldn't take reservations for them without your rules.
• Large or trade orders, such as a school ordering team kit or an office ordering Christmas hampers.
• Anything about a payment that went wrong.

What "hand over" actually means

Klugon can't put a caller through to you. There is no live transfer. In messages and website chat, when the AI hands a conversation over, it marks it as needing a person, shows a "Needs a person" banner in the inbox and, if you've connected them, posts it to Slack or sends a Zapier event that you can route to wherever you'll see it. On a phone call, the reliable route is a note: the AI saves the caller's name, number and what they need as a lead, which appears on your leads list and posts to Slack or Zapier in the same way. It doesn't email you or send a push notification.

So on the phone the AI says something like "I've passed that to the team and someone will get back to you today." You see it in Slack, the inbox or the leads list and ring back. On WhatsApp, Instagram and Messenger, you can reply directly from the Klugon inbox within Meta's 24-hour reply window. For website chats and phone calls, you reply by phone or email.

The AI doesn't stop after it flags a conversation. If the customer keeps messaging, it keeps answering. Write your instructions so it doesn't make promises on your behalf, for example "never say an item has been put aside; say the request has been passed to the team".

Stock: how accurate the answers can be

This is the section to read carefully, because it decides whether the AI helps or annoys your customers.

Klugon has its own product catalogue. You add products by hand, with a name, description, price, image and a stock number. When a customer asks about a product, the AI searches that catalogue and tells them what you've entered, including whether it's marked as in stock.

There's no import and no sync with your till or web shop. If you sell the last one at 10:15 and don't change the number in Klugon, the AI will still say it's available at 10:30.

How shops deal with that depends on what they sell:

• Slow-moving, high-value stock (a boutique with one of each, a bike shop, a jeweller): keep the key lines in the catalogue and update stock when something sells. It's a minute's work and these are the questions worth answering accurately.
• Fast-moving stock (a grocer, a newsagent, a bakery counter): don't try to keep live stock in Klugon. Put your ranges into the knowledge base instead ("we always carry oat milk, gluten-free bread on Tuesdays and Fridays") and tell the AI to say "we usually stock it; I'll ask the team to check today" for anything specific.
• Large catalogues (a bookshop with thousands of titles): don't enter every SKU. Enter your featured and frequently asked-about lines, and have the AI take special order requests for the rest.

Write the honesty into the instructions: "If you're not sure an item is in the shop today, say so and offer to have someone check." A customer who's told "I'll check" and gets a message back is happy. A customer who drives across town for something that sold yesterday is not.

If you use Shopify POS, Square, Lightspeed or EPOS Now

Most independent shops run their till on Shopify POS, Square, Lightspeed, EPOS Now or similar, and many have a web shop on Shopify, WooCommerce or Wix. Klugon doesn't connect to any of them. There's no product sync and no order sync.

What you can do is add the Klugon chat widget to your Shopify, WooCommerce or other website by pasting one script tag, so shoppers can ask questions on your product pages. The Shopify chatbot page shows where to paste it. Your web shop stays the place where people buy; the AI answers questions and points people to the right product page using your links library.

Click-and-collect, reservations and orders

Shops ask most often about taking orders, so it's worth being exact.

We don't recommend building your shop's ordering around payments in chat, and the AI doesn't take payments on phone calls. In practice, shops use one of these set-ups:

• Send them to your web shop. The AI answers the question, then shares the product page link from your links library. Customers order and pay through the shop you already have, and your click-and-collect works as it does now.
• Reservation requests as messages. The customer asks you to hold an item. The AI takes their name, number, the item and when they'll come in, saves it as a lead and flags it. You check the stock, put it behind the counter and reply. This is the informal "I'll keep one for you" turned into a written list.
• Collection slots as bookings. If you get a lot of collections, create a bookable service called something like "Collection, 15 minutes" with your collection hours. The AI books a slot, and in chat or on WhatsApp it can add a note to the booking saying what's being collected. You get a list of collections by time in the dashboard. Customers pay at collection.

None of these send an automatic confirmation or a reminder afterwards. The customer gets their answer or booking reference in the conversation itself, and that's it. If you want to message them when it's ready, you reply from the inbox on WhatsApp, Instagram or Messenger, or you ring them.

Grocery store WhatsApp ordering

Small grocers, delis and farm shops often already take orders on WhatsApp, usually from regulars who send a list. The AI can take that list, answer questions about what you carry, and flag the order for whoever is picking. It won't price and total a basket against your till, so the person picking confirms the total. The grocery stores page covers this in detail, including how to handle substitutions and delivery slots.

The boutique AI sales assistant

Boutiques get questions that are half product, half advice: "Would this work for a wedding?" "Does it come up small?" "What would you wear it with?" The AI can answer from what you've written about fit, fabric and styling, and suggest related pieces from your catalogue. It works best when your product descriptions say what you'd say in person. The clothing boutiques page shows how to write them so the AI sounds like you, and our AI sales assistant page explains the product-suggestion settings.

Sample conversations

Two examples of what conversations look like once a shop is set up. Both are samples written for this page. The shops and customers are made up, and the AI voice on a phone call is AI-generated.

Sample 1: a stock question on Instagram

Customer: Hi! Do you still have the rust linen dress from your post on Sunday? Size 12?

AI: Hi! That's the Amalfi linen midi. We have it in rust in sizes 8, 10 and 14 at the moment. Size 12 is showing as sold out, I'm sorry.

Customer: Oh no. Does it come up big or small? Wondering if the 14 would work

AI: Our notes say it's cut generously through the body with a tie waist, so many customers size down. If you're usually a 12, the 10 might be closer than the 14. The best thing is to try both. We're open tomorrow 10:00 to 17:30.

Customer: Could you put a 10 aside for me? I can come Friday lunchtime

AI: I'll pass that to the team now with your request: Amalfi linen midi, rust, size 10, collecting Friday lunchtime. They'll confirm in this chat once it's set aside. Can I take a name?

Customer: Jess Hall

AI: Thanks, Jess. One more thing: if it's not right when you try it on, you can exchange it within 28 days with the receipt. Anything else I can help with?

Customer: No that's great thanks!

What happened behind the scenes: the AI searched the catalogue, read the stock figures and the fit notes, and saved a lead with the reservation details. It flagged the conversation and posted it to the shop's Slack. It didn't say the dress had been put aside, because its instructions say only the team confirms that. At 9:40 the next morning the owner checks the rail, bags the dress and replies to Jess from the Klugon inbox. The 28-day exchange is the shop's own goodwill policy, which it uploaded.

Sample 2: a phone call to a sports shop

AI: Hello, Fellside Sports. How can I help?

Caller: Hi, are you open tomorrow? And do you do that gait analysis thing?

AI: Yes, we're open tomorrow, Sunday, from 10:00 to 16:00. We do offer gait analysis. It takes about 30 minutes and it's free with a pair of running shoes. Would you like to book a time?

Caller: Yeah, tomorrow if possible.

AI: Tomorrow I have 11:00 with Sam or 14:30 with Ellie. Which suits you?

Caller: 11:00.

AI: Can I take your name and a mobile number?

Caller: Dan Moore, 07700 900233.

AI: Thanks, Dan. You're booked for gait analysis at 11:00 tomorrow with Sam. Your reference is FS-2290. Wear or bring the socks you run in, and your current trainers if you have them.

Caller: Great. Oh, I bought some trail shoes last month and the sole's coming away. Can I bring them back?

AI: I'm sorry to hear that. I can't decide on a faulty item over the phone, but I'll pass it to the team so they can look at them when you're in tomorrow. Please bring the shoes and your receipt or the card you paid with.

Caller: OK, thanks.

AI: I've passed a note to the team. See you at 11:00 tomorrow. Goodbye.

What happened behind the scenes: a 30-minute appointment booked with Sam in a free slot, posted to Slack, and a note saved for the team as a lead, "possible faulty trail shoes, bought last month, customer in at 11:00 Sunday". The AI didn't promise a refund or a replacement, because the instructions say faulty goods go to a person. Under the Consumer Rights Act the customer may well be entitled to one, and that's a conversation the manager has with the shoes in hand.

Consumer rights the AI must get right

An AI that answers returns questions is giving out statements about your obligations. Get the information right at the start, and write the hand-over rules so it never contradicts the law. This section covers UK law; if you trade elsewhere, check your own country's rules.

Faulty goods under the Consumer Rights Act 2015

The Consumer Rights Act 2015 says goods must be of satisfactory quality, fit for purpose and as described. If they're not:

• The customer has a short-term right to reject faulty goods and get a full refund, normally within 30 days. For perishable goods such as flowers or food, the period can be shorter, matching how long the goods could reasonably be expected to last.
• After that, they're generally entitled to a repair or replacement, and if that fails, a price reduction or a final right to reject.
• Once you agree a refund is due, it has to be paid without undue delay and within 14 days.

Your own returns policy for change of mind ("exchange within 28 days with a receipt") is separate. That's goodwill, and you can set it how you like, as long as it doesn't suggest customers have fewer rights than the law gives them. Upload both, clearly labelled, so the AI can tell the difference. For practical, plain-English guidance, the Chartered Trading Standards Institute's Business Companion site is good.

The rule for the AI is simple: it can explain your policy and say that customers have legal rights for faulty goods, but a claim that something is faulty always goes to a person.

Orders taken remotely: the 14-day cancellation right

When a sale is made at a distance, without the customer in the shop, the Consumer Contracts Regulations 2013 apply. That includes orders by phone, WhatsApp or online, including many click-and-collect orders.

• Customers can usually cancel within 14 days of receiving the goods, without giving a reason.
• You have to give certain information before the order: who you are, the total price, delivery costs, and the cancellation rights.
• If you don't tell them about the right to cancel, the cancellation period can extend up to 12 months.

There are exceptions that matter to particular shops: goods made to the customer's specification or clearly personalised (a tailored suit, an engraved gift), goods that go off quickly (flowers, fresh food) and sealed goods unsealed after delivery for health or hygiene reasons. The UK government's guide to online and distance selling for businesses sets out what you need to tell customers.

If the AI takes a reservation request that you then turn into a sale by phone, you're making a distance sale. Make sure you, or the information the AI gives, covers what you're required to say.

Prices and reviews under the 2024 Act

The consumer provisions of the Digital Markets, Competition and Consumers Act 2024 came into force on 6 April 2025. Two parts are relevant to a shop's AI assistant:

• Drip pricing. Mandatory charges have to be included in the headline price. If you have a compulsory fee, such as a minimum delivery charge on all local orders, the AI's price information should include it from the start.
• Fake reviews. Writing, commissioning or publishing fake reviews is banned. Don't put invented customer quotes into the AI's knowledge for it to repeat.

Medicines and age-restricted products

Pharmacies have extra professional rules. The AI can answer about opening hours, services and whether you offer a particular consultation, but any question about symptoms, doses, interactions or whether a medicine is suitable goes to a pharmacist. The pharmacies page covers this in detail.

For age-restricted products (alcohol, knives, some solvents, fireworks, lottery), the AI can say what you stock and that ID is checked at the till. It can't verify age, and you should decide whether it takes reservations for these at all.

Customer data

Every conversation includes personal details: names, phone numbers, sometimes addresses. The ICO's guidance for organisations covers what you need to tell customers. Klugon is operated by a German company under the GDPR. Its sub-processors are listed in the privacy policy, with international transfers covered by standard contractual clauses. Calls aren't recorded, each call is saved as a conversation in your inbox, you can export data as CSV and full deletion is by email request. Lead capture records marketing consent only when the customer gives it explicitly, which matters if you plan to email them about your sale later.

Setup checklist for a shop

Most of the work is gathering information you already have in your head.

Gather first

• Opening hours for each day, and your holiday hours and closures for the next few months.
• Your returns and exchange policy, written down, with a clear line between goodwill and legal rights.
• Delivery, collection and gift-wrap policies, with any charges.
• Your products: at least the lines people ask about most, with price, description, sizes and colours, fit or care notes, and a stock figure where you'll keep it current.
• Links to your web shop's main pages: bestsellers, gift cards, delivery info.
• Any appointments you take, with how long each lasts and who does them.
• The list of things that always go to you: faulty goods, complaints, special orders, medicines, trade orders.
• Who checks the inbox, and when.

Set up in Klugon

1. Sign up and start the 7-day free trial on the product you need first. For most shops that's the AI social agent (WhatsApp, Instagram, Messenger) or the website chatbot, with the phone added once you're confident. A card is needed for the trial and is charged when it ends unless you cancel.
2. Fill in the business profile, address, opening hours and time zone.
3. Add products to the catalogue, with images and stock numbers.
4. Upload policies and range information to the knowledge base as PDF, Word, text or CSV, or paste them in. The website scan can read your homepage and up to seven key pages to get you started.
5. Add your links: web shop pages, map, returns policy, gift cards.
6. If you take appointments, add them as services with lengths and the staff who do them. Add a "Collection" service if you'll use collection slots.
7. Write the assistant's instructions: tone, how to talk about stock, hand-over rules, and whether to suggest related products.
8. Test it in the browser with the questions customers actually ask.
9. Connect Slack or Zapier so flagged conversations and new bookings reach you.

Connect the channels

• Instagram and Messenger. Connect your Facebook Page and professional Instagram account. Instagram needs Meta app review before members of the public can message the assistant, so start this early.
• WhatsApp. Connect your WhatsApp Business number through Meta's Cloud API, with a Meta app, phone number ID and access token. Meta's WhatsApp Business site explains the account side.
• Website. Paste the widget script into your site. It works on Shopify, WooCommerce, Wix, Squarespace or a hand-built site.
• QR code. Print the QR code for your public Klugon page and stick it in the window with "Questions when we're closed? Scan here."
• Phone. Numbers are added once the trial has ended; until then you test in the browser. Buy a new number, or keep your existing one by forwarding it to a Klugon number with your phone provider. Forwarding only when you don't answer is a sensible place to start.

Tools you probably already use, and how they fit

What it costs

Klugon bills in US dollars. There are three separate products, and shops often need two of them.

A phone number is $9 a month, a connected WhatsApp number is $19 a month, and yearly billing takes 20% off. On the social and chat products, new conversations stop at 110% of your monthly allowance, and conversations already under way carry on. Full details are on the pricing page.

These worked examples use made-up assumptions. Put in your own.

Example 1: a boutique that lives on Instagram

Assumptions:

• About 6 Instagram and WhatsApp conversations a day, so roughly 180 a month.
• About 2 website chats a day, so 60 a month.
• The phone is answered in the shop, with no AI on calls.

That fits the social Starter (300 conversations) and the chatbot Starter (100 conversations).

Cost: $79 social Starter + $19 WhatsApp number + $49 chatbot Starter = $147 a month.

Example 2: a grocer taking WhatsApp orders

Assumptions:

• About 15 WhatsApp conversations a day from regulars, so around 450 a month.
• Calls forwarded to the AI only when nobody answers, about 6 a day at an average of 1.5 minutes: 6 × 1.5 × 30 = 270 minutes.

450 conversations is over the social Starter limit. New conversations would stop at 330, so this shop needs social Business (1,500 conversations). The calls fit voice Starter.

Cost: $249 social Business + $19 WhatsApp number + $99 voice Starter + $9 phone number = $376 a month.

Set against paid cover

From April 2026 the UK National Living Wage for workers aged 21 and over is £12.71 an hour. Say you took on someone for 20 hours a week mainly to deal with the phone and messages on your busiest days. That's 20 × £12.71 = £254.20 a week, or about £1,101 a month, before employer's costs.

A person on the shop floor also sells, restocks, wraps and spots shoplifters, so that isn't a straight swap. The fair comparison is whether the AI takes enough questions away from the counter that you and your existing staff can serve the customers who are actually there, and whether it catches enquiries that currently go nowhere because they come in after you've locked up.

What an unanswered question costs

Suppose 10 stock and hours questions a week go unanswered or get answered too late, 1 in 5 of those people would have come in and bought, and the average sale is £40. That's 2 × £40 = £80 a week, or about £350 a month. If those numbers look like yours, the AI covers its cost. If you already answer nearly everything, it may not, and that's a fine reason not to use it. Our guide on how to never miss a customer call covers the cheaper options too.

Measuring whether it helps

After the first month, look at:

1. Conversations by channel. Where questions actually come from. Many shops are surprised how many arrive on Instagram after closing time.
2. Busiest hours. Whether the AI is taking pressure off Saturday or mainly answering in the evening.
3. Handled without a handover. If the AI is flagging most conversations, your catalogue or policies have gaps.
4. Questions without a good answer. The list to work through each week. It shows what customers want to know and you haven't told the AI.
5. Most asked questions. Useful beyond the AI: if everyone asks whether you stock a brand, that's a buying decision.
6. Leads and bookings. Reservation requests, special orders and appointments the AI took.

And the simplest test: is the phone behind the till ringing less on a Saturday, and are you answering fewer DMs at 22:00?

Retail shop types we cover in detail

Every type of shop has its own questions, peaks and rules. We've written a separate guide for each:

• Grocery stores: WhatsApp order lists from regulars, substitutions, delivery slots, and what "in stock" means when stock changes by the hour.
• Florists: same-day delivery cut-offs, Valentine's and Mother's Day peaks, wedding and funeral consultations, and collecting card messages correctly for a florist order chatbot.
• Pharmacies: opening hours, prescription collection questions and consultation bookings, with a firm line that anything clinical goes to a pharmacist.
• Clothing boutiques: Instagram DMs, sizing and fit questions, reserve-and-try, and writing product notes so a boutique AI sales assistant sounds like you.
• Bookshops: "have you got this title" for a stock of thousands, special orders, author events and book club bookings.
• Tailors: alterations and fittings by appointment, turnaround questions, and why bespoke work is exempt from the usual cancellation right.
• Gift shops: gift ideas by budget, wrapping and personalisation, and the November-to-December crush.
• Sports shops: sizing, gait analysis and fitting appointments, bike and racket servicing, team kit orders and back-to-school.

Frequently asked questions

Can an AI assistant check stock for my shop?

It can answer from the stock figures you keep in Klugon's product catalogue. It doesn't connect to your till or web shop, so those figures are only as current as you keep them. For fast-moving stock, tell it to say "I'll ask the team to check" instead.

Does it connect to Shopify, Square or my EPOS?

No. There's no product or order sync with Shopify, Square, Lightspeed or any EPOS. You can add the Klugon chat widget to a Shopify or other website by pasting one script tag.

Can customers pay through the AI?

Don't plan on it. The AI doesn't take payments on phone calls, and we don't recommend building your shop's ordering on in-chat payment today. Send customers to your web shop, or let them pay when they collect.

Can it reserve items for click-and-collect?

It can take a reservation request with the customer's name, the item and their collection time, and flag it for you. You confirm it once the item is behind the counter. If you handle a lot of collections, you can also take collection slots as bookings.

Will it send customers a confirmation or a reminder?

No. The customer gets their answer or booking reference in the conversation. Klugon doesn't send separate texts, emails or WhatsApp messages, and there are no automatic reminders or follow-ups.

How does it handle returns and refunds?

It explains your returns policy and that customers have legal rights for faulty goods. Claims that something is faulty, refund disputes and complaints are flagged for you to deal with.

Can it transfer a call to me in the shop?

No. It can't put a call through. On a call it saves the caller's details and a note for you, which appear on your leads list and in Slack or Zapier if connected, and tells the caller someone will get back to them.

Can I keep my shop's phone number?

Yes, using call forwarding. Klugon gives you a number for $9 a month and you forward your existing line to it, either for all calls or only when you don't answer. Number porting isn't supported.

Is it worth it for a small shop?

It depends how many questions you miss or answer late. If messages pile up after closing and the Saturday phone pulls you off the till, probably. If you already answer nearly everything, probably not. The 7-day trial lets you test it in the browser first.

Which channels should a shop start with?

Start where your questions already arrive. For most boutiques and gift shops that's Instagram and the website. For grocers it's usually WhatsApp. Add the phone once the information is right.

Klugon's AI receptionist for small business works best in a shop when it's honest about what it knows: your hours, your policies, the lines you keep up to date. Give it that, decide what comes to you, and let it deal with the questions while you deal with the customers in front of you.