Chatbot for Customer Support After the Sale: Answer ‘Where Is My Order?’, ‘When Is My Refund?’ and ‘Is It Fixed Yet?’ Without Order Lookups (Free Status Answer Builder)
Chatbot for customer support setups usually stop at pre-sale questions like prices and opening hours, but most support chats arrive after the sale: where is my parcel, when will my refund land, is my phone repaired yet. A small business bot usually can’t see your order system, so it can’t look these up. It can still answer most of them well by explaining your timeline honestly and handing over only the cases that are actually late. This guide shows how, with a free tool that turns your process stages into a ready bot answer and a day-by-day script.
⚡ In short
- Most after-sale chats are status questions. Customers want to know where things stand and when the next step happens.
- Without an order lookup, the bot answers from your timeline: what normally happens on which working day.
- Every status answer needs the stage, the next step with its day, and the day a person checks by hand.
- Only customers past that check-by-hand day go to your team, as a ticket with their order number.

The problem
What a chatbot for customer support can and can’t do after the sale
Picture the inbox of a small online shop the week after a sale. Hardly anyone asks what you sell. They ask: “order kab aayega?”, “refund not received”, “it says shipped 3 days ago, where is it?”, “is my booking confirmed?”. These are status questions. The customer has already paid, and now they’re anxious about the gap between paying and receiving.
Status questions are tricky for a website bot. A big retailer’s bot connects to the order database and says “your parcel is in Nagpur, arriving Thursday”. Most small businesses don’t have that connection, and building one is a developer project. LiveChatBot’s bot answers from the knowledge and saved replies you give it; it doesn’t read your courier dashboard or your bank.
So many owners decide the bot can’t help after the sale and route every “where is my order” chat to a person. That works until the busy week, when your team spends the day typing “it usually takes 5 to 7 days” fifty times.
There’s a middle path. Most status questions aren’t really “look up my order”; they’re “is this normal?”. A customer who ordered two days ago and hasn’t got the parcel isn’t facing a problem; they just don’t know your timeline. If the bot explains it clearly (“orders ship within 2 working days and arrive within 6”), the customer relaxes and the chat ends. Only the customer on day 11 has a real problem, and that one deserves a person.
The rest of this guide shows how to write answers that separate “normal” from “late” without any lookup.
The five families
The status questions small businesses get most
Almost every after-sale chat falls into one of these. Each has a timeline you can write down.
Delivery
“Where is my order?”, “not received yet”, “shipped but not here”. Stages: packed, shipped, out for delivery, delivered.
Refund
“Refund kab milega?”, “money not credited”. Stages: return received, refund approved, sent to bank or UPI, shows in account.
Repair or service
“Is my laptop ready?”, “when will the technician come?”. Stages: received, diagnosed, parts ordered, ready for pickup.
Booking
“Is my appointment confirmed?”, “didn’t get the confirmation”. Stages: request received, confirmed, reminder sent.
Replacement or warranty
“I asked for a replacement, what next?”. Stages: claim received, approved, replacement dispatched.
Documents
“When will I get my invoice or certificate?”. Stages: request received, prepared, emailed.
Step 1
Map each process in stages, with honest working days
Pick the process customers chase you about most. For most shops that’s delivery; for service businesses it’s repairs or bookings. Write its stages in order, from the moment it starts to the moment the customer has what they wanted.
Then put a number next to each stage: the working day it usually happens, counted from day 0. Not the best case, not the worst case: the day by which most cases (say nine out of ten) have reached that stage. If you aren’t sure, look at your last twenty orders or refunds and check the dates.
An illustrative refund for a small clothing shop might look like this:
- Day 0: we approve the refund
- Day 2: the refund is sent to your bank or UPI
- Day 7: the money shows in your account
Three things to watch:
Use working days and say so. Customers count Sunday; your courier and bank don’t. A refund that takes “7 days” over a long weekend feels late on calendar day 8.
Include stages you don’t control. Banks and couriers take their own time. Customers don’t care whose step it is; they care when it ends. “Your bank may take up to 5 working days after we send it” is honest and stops a lot of anger.
Keep it to three to five stages. Customers don’t need your internal workflow. They need the milestones they can see or check.
Finally, choose a check-by-hand day: the working day after which a case is no longer normal and a person must look at it. Usually that’s one to three days after the last stage. That one number decides which chats the bot answers and which go to your team.
Free tool
Turn your stages into a status answer
Enter a process, its stages with working days, and your check-by-hand day. The tool writes the answer for your bot, shows what a customer hears on every day, and flags timelines that will cause trouble.
Name the process customers chase you about (refund, delivery, repair, booking, replacement).
List its stages, one per line, as stage | working day it usually happens. Start with day 0.
Check it by hand after working day (the day a person takes over):
Runs only in your browser; nothing is sent or saved.
Step 2
Write status answers with three parts
A good status answer works for the customer on day 1 and the customer on day 9, without knowing which one is asking. It does that by giving the whole timeline and letting the customer place themselves on it. Every good answer in a chatbot for customer support has the same three parts.
1. Where it should be by now
List the stages with their days. "Day 0: we approve the refund. Day 2: it's sent to your bank or UPI. Day 7: it shows in your account." Customers instantly see which stage they're at.
2. What happens next, and by when
The list already does this, but say the finish line in plain words too: "Most are complete by day 7." Anxious customers read the last line first.
3. When a person takes over, and what to send
"If it has been more than 9 working days, share your order number and our team will check it personally." This is the most important sentence. It tells the early customer to wait without sounding dismissive, and tells the late customer exactly how to get help.
What to leave out
Don't promise exact dates ("you will get it on Friday"); the bot doesn't know when this order started. Don't blame the courier or bank. Don't add "sorry for the inconvenience" to a case that is on time; it suggests something has gone wrong when nothing has.
Write one answer per process, in the words customers use. If people say "parcel" and "courier", use those words, not "shipment" and "logistics partner". Our guide to writing answers your bot can use covers wording in more detail.
Before and after
Status answers, rewritten
Step 3
Hand over the late cases as tickets, not chats
The customer past your check-by-hand day has a real problem, and the bot's job changes from explaining to collecting. It needs three things from them: the order, booking or job number, a contact (email or phone) so you can reply later, and a one-line description of what they see ("courier says delivered but I didn't get it").
Turn that into a ticket rather than leaving it as a chat. A late refund rarely gets solved in one sitting: someone has to check the bank statement or call the courier, which can take a day. In LiveChatBot, a chat can become a ticket with a status (open, pending, solved) and an owner, and the customer gets a ticket page they can open without logging in to see progress. See turning a chat into a ticket and how a ticketing system fits a small team.
If the customer writes when nobody is online, your opening hours and away form collect their email or phone, and offline messages can become tickets too, so late cases asked at midnight still land in the morning queue. Set this up with opening hours.
Make the handover sentence specific: "I've passed this to our team with your order number. You'll get an update by email within one working day." Vague promises ("someone will get back to you") send the customer straight back to chat, and to WhatsApp, and to your Instagram comments. Our handover guide has more wording.
If you use the REST API or Zapier, the ticket.created webhook can also post late cases into a sheet or your team's channel, so nobody has to watch the inbox to notice them.
Step 4
Questions with conditions: warranty, returns and replacements
Some after-sale questions aren't about time; they're about eligibility. "Can I return this?", "is it under warranty?", "can I get a replacement for a broken part?". The bot can't inspect the product, but it can walk the customer through the conditions you already apply.
Write these as short if-then answers, in the order you'd check them yourself:
- Time limit first. "Returns are accepted within 7 days of delivery." Most ineligible cases stop here.
- Condition second. "The item must be unused, with tags." "Warranty covers manufacturing faults, not accidental damage."
- Proof third. "Keep your invoice or order number ready."
- Then the next step. "If all three apply, share your order number and a photo, and we'll start the return."
The bot doesn't decide whether a claim is approved; a person does. What the bot does is stop the obviously ineligible requests early, politely, and make sure the eligible ones arrive with everything your team needs.
Keep the wording identical to your published policy. If your returns page says 7 days and the bot says 10, the customer will quote the bot. When you change a policy, update the bot the same day; our guide on updating your bot after a business change shows how to find every copy of an old fact.
For anything involving money disputes, damaged deliveries or angry customers, skip the conditions and hand over straight away. A calm intake for complaints is covered in our complaints guide.
Setup plan
A week to cover your after-sale questions
Choose the status question you get most. Count last month's chats about it, so you can compare later.
Check the dates on your last 20 cases. Write the stages and the day most cases reach each one.
Use the tool above. Pick a check-by-hand day one to three days after the usual finish.
Add the answer to the AI Agent's knowledge and save it as a saved reply for the keyword bot and your team.
Opening hours, away form and tickets switched on. Decide who owns late cases.
Ask the bot the question ten ways, including Hinglish and typos. Check it never promises an exact date.
Switch it on, then start on the second most common process.
Step 5
Measure the right thing: late cases, not chat counts
After a few weeks, two numbers tell you whether this is working.
Status chats that ended at the bot. These are customers who read the timeline and didn't need anyone. If most "where is my order" chats now end there, your team has those hours back. Our weekly review routine shows how to tag chats by outcome.
Late cases per week. This is the number of tickets created after the check-by-hand day. It's not really a chat metric; it's a measure of your process. If late refunds jump in one week, the problem is likely your bank, courier or a stuck step in your own workflow, not the bot. A rising count is an early warning you'd otherwise only get from bad reviews.
Watch for two warning signs as well:
- Many customers ask again a day or two later. Your days are too optimistic. Move the stages later; a promise you keep beats a fast promise you break.
- Your team keeps finding nothing wrong in late tickets. Your check-by-hand day is too early. Add a day of grace.
You can also prevent some status chats entirely. A proactive message in Flows, shown only on your order-tracking or returns page, can share the timeline before anyone asks. See how to create a chat flow and proactive live chat. And if you're deciding which other support jobs to hand to a bot, our rulebook for who answers what and staffing calculator are good next reads.
Checklist
After-sale support bot
Getting started
Set up your after-sale answers this week
Start with the one process your customers chase most. Write its stages, find the real working days, run them through the builder above and load the answer into your bot. A chatbot for customer support that handles just that one question well can take a noticeable share of after-sale chats off your team in a busy week.
LiveChatBot's free plan includes the AI Agent with a small free monthly allowance, a free keyword bot that answers from saved replies, the team inbox, tickets with a customer ticket page, opening hours with an away form, Flows and the Android app, for two websites with 100 visitors per website each month. There are no per-seat charges. See pricing and start at livechatbot.in.
Frequently asked questions
Can a chatbot for customer support track orders without a connection to my store?
Not order by order. Without a lookup it can't see a specific parcel. It can explain your usual timeline so customers know whether their case is normal, and collect the order number for anyone who is past your check-by-hand day.
Should the bot give calendar days or working days?
Working days, and say so in the answer. Couriers and banks work in working days, and customers who count weekends will think a normal case is late.
What if my delivery times differ by city?
Write one line per zone in the same answer, for example metro cities and the rest of India, each with its own days. Keep it to two or three zones so the answer stays short.
When should the bot hand over straight away?
Damaged or missing items, money disputes, angry customers and anything about a specific account. Those need a person, so the bot should collect details and create a ticket rather than explain timelines.
Does the status answer builder save my data?
No. It runs entirely in your browser and nothing is sent or saved.
Keep reading
Related guides
Chatbot Feedback Examples: 40 Copy-Ready Ways to Ask, by Moment, and What the Bot Says Back (Free Question Checker)
Chatbot Features That Only Work in Pairs: A Dependency Map So Nothing You Switch On Breaks a Chat (Free Broken-Pair Finder)
Chatbot Dialogue Examples for the Hard Moments: 12 Annotated Exchanges With Angry, Confused, Off-Topic and Silent Visitors (Free Reply Drill)
Answer after-sale questions calmly
Give your bot honest timelines, turn late cases into tickets, and let your team focus on real problems. Start free on livechatbot.in.