Read a week of your shop's chats and you will notice something: most of them are the same handful of questions. "Do you have it?" "How much is delivery?" "Will it fit?" These common customer questions in an online store are not a nuisance. They are the last step before a purchase, and every one left unanswered is a sale at risk. This article sorts them into nine groups, shows where the answer to each should come from, and marks which ones a bot can handle and which need a person.
Why the same questions keep coming
Customers ask when the page did not answer them, or when they do not trust what it says. A product page can show "in stock", but the customer wants to hear it for the size they need. A delivery page can list carriers, but the customer wants to know if the parcel arrives before Saturday.
The questions also line up with the reasons people abandon a purchase. In Baymard Institute's research on US shoppers, the top reasons are extra costs that are too high (40%), slow delivery (20%), an unsatisfactory returns policy (13%) and not seeing the total cost up front (12%). Each reason is a question someone could have asked in chat. A fast, correct answer removes the doubt before it turns into a closed tab. Why speed matters so much is covered in our pillar article on the cost of slow replies.
Common customer questions online stores get, in nine groups
1. Availability: "Do you have it?"
"Is the DX-450 in stock?" "Do you have it in size 42?" "When will it be back?" This is usually the largest group. The answer lives in your product feed: availability and, if you export it, stock quantity. It is only as good as the feed is fresh, so check how often your export updates. More in what a YML product feed is.
2. Price and discounts: "How much?"
"What is the price?" "Is there a discount?" "Why is it cheaper on another site?" Current price and old price come from the feed. Promo codes and personal discounts usually do not, and a customer asking for a special price should go to a manager.
3. Product details: "What exactly is it?"
"What is the battery capacity?" "Is it cotton?" "What are the dimensions?" These answers sit in the product description and parameters. If a parameter is missing from your feed, nobody can answer it automatically. That makes this group a good audit: every repeated spec question is a field your product page lacks.
4. Fit and compatibility: "Will it work for me?"
"Will this charger fit my drill?" "Which size do I need if I wear M?" "Does this case fit the 15 Pro?" Answers come from parameters, compatibility lists and size charts. A good answer also suggests the right accessory or an alternative.
5. Choosing and comparing: "Which one should I take?"
"What is the difference between these two?" "What do you recommend for a beginner under $100?" Comparisons are where guessing is most dangerous. If the facts are in the catalog, an assistant can lay them side by side. If they are not, the honest answer is "let me pass this to a manager", not a confident guess that ends in a return.
6. Delivery: "When and how much?"
"How much is delivery to Lviv?" "Do you ship abroad?" "Can I pick it up today?" In Ukraine the question is often about Nova Poshta or Ukrposhta, in other markets about couriers and pickup points. The answer belongs on your delivery page. Write down carriers, costs, free-delivery thresholds and usual timing in plain text.
7. Payment: "How can I pay?"
"Can I pay on delivery?" "Do you accept cards?" "Is there an installment plan?" Again, this belongs on a written payment page. Customers ask because the options are hidden at checkout.
8. Returns and warranty: "What if it doesn't fit?"
"Can I return it?" "How long is the warranty?" "Who pays for return shipping?" General terms belong on your returns page and can be answered automatically. A specific dispute ("I want my money back for this order") is a different conversation and needs a person.
9. Orders already placed: "Where is my order?"
"Where is my parcel?" "Can I change the address?" "I got the wrong item." These need access to your order system and, often, a decision. They should go straight to a manager.
Which questions a bot should answer and which go to a person
| Question group | Where the answer lives | Who should answer |
|---|---|---|
| Availability | Product feed | Assistant |
| Price | Product feed | Assistant; special prices go to a manager |
| Product details | Feed parameters, description | Assistant |
| Fit and compatibility | Parameters, size charts | Assistant, if the data exists |
| Choosing and comparing | Catalog data | Assistant if the facts are there, otherwise a manager |
| Delivery | Delivery page | Assistant |
| Payment | Payment page | Assistant |
| Returns and warranty | Returns page | Assistant for terms, a manager for disputes |
| Orders already placed | Order system | Manager |
Roughly, the first eight groups are about information that already exists. The last one, plus complaints and wholesale, is about decisions. We explain the handover rules in detail in when a chatbot should hand over to a human.
How to prepare your answers
You do not need a script for each question. You need the right information in the right place.
- Clean up the feed. Make sure availability, price, old price, category and key parameters are exported. Missing parameters cause most "I don't know" answers.
- Write the policy pages. Delivery, payment, returns, warranty and contacts, each on its own page, in plain language. Include numbers: costs, thresholds, days.
- Add a FAQ page for the questions that do not fit elsewhere, such as gift wrapping or invoices for companies.
- Decide what goes to a person. Write down who handles complaints, wholesale and order problems, and how fast.
- Review chats monthly. Every question the bot could not answer is a gap in your feed or your pages.
How Vendigit handles these questions
Vendigit's assistant answers groups 1 to 8 from two sources. Product questions come from your product feed: it searches the catalog, suggests alternatives and accessories, and compares products. Prices, names and links on product cards are filled in straight from the feed, and the AI is instructed not to state a price or stock level it didn't find. Delivery, payment, returns, warranty and contacts come from your own site: after you verify your domain, Vendigit reads up to 20 policy pages and re-reads them weekly. You can also add documents by hand.
If something is not written anywhere, the assistant says the shop does not publish it, instead of inventing it. Order questions, complaints, refund disputes, wholesale and special prices go to your team: the people you choose get an email, and the dialog shows up in your dashboard. The assistant replies in Ukrainian, Russian or English, in the language the customer writes in.
FAQ
What is the most common customer question in an online store?
In most shops it is availability: "Do you have it?", often with a size or colour. Price and delivery follow close behind.
Should I answer common questions with a FAQ page or a chatbot?
Both. The FAQ page is a source of truth. A chatbot that reads it brings the right answer to the customer at the moment they ask, without making them search.
What if my delivery terms change often?
Keep them on one page of your site and update it there. Vendigit re-reads policy pages every week, and you can trigger a new collection from the dashboard at any time.
Can the assistant answer questions about a specific order?
No. Vendigit does not connect to your order system. Questions about placed orders are handed to a manager.