Switch on a WhatsApp auto reply and your numbers get flattering fast. Every message answered in under a second. Response rate: 100%. Then you check orders at the end of the week and nothing has moved.
That gap is the honest place to start. Automatic replies are excellent at one job — confirming a message arrived — and useless at another: getting a buyer to a decision. Most sellers hire a tool for the first job and quietly expect the second.
Here is what each type of automated message is actually for, the exact moments where a canned reply costs you a customer who had their card out, and how to build a WhatsApp setup where the automation closes something instead of just being polite.
What a WhatsApp auto reply really is
The WhatsApp Business app gives you three things, and they are narrower than most people assume:
- Greeting message — fires once when someone messages you for the first time, or after 14 days of silence.
- Away message — fires outside the business hours you set, or always if you switch it on permanently.
- Quick replies — saved snippets you pull up with a shortcut like /shipping. A human still has to send them.
Notice what's missing. None of these read the customer's message. A greeting message says the same 40 words to someone asking "do you have this in size 40?" and someone asking "is this shop still open?" The tool has no idea which is which, because it was never designed to know.
That's fine. Trouble starts when a seller treats those three features as customer service and stops there.
The five questions your inbox is actually full of
Look at a week of first messages in almost any product business and the same short list dominates: price, availability in a specific size or colour, shipping cost, delivery time, and "is this real / does it come with a warranty?" In most seller inboxes, that handful covers something like eight out of ten opening messages.
Now compare that to a typical greeting message:
"Hi! Thanks for contacting us. Our team will get back to you as soon as possible. Working hours: 9–18, Saturday to Thursday."
Nobody asked any of that. The message answers a question the buyer didn't have while ignoring the one they did. It's the conversational equivalent of a hold tone.
Where canned replies quietly cost you sales
Three moments do most of the damage.
1. The away message that broadcasts your absence
"We're currently closed and will reply tomorrow morning" is honest, and it's also permission to go shopping elsewhere. Evening is when people browse. A buyer who reads that at 22:40 has four hours and three competitor chats ahead of them. We looked at how long buyers actually wait before moving on in this breakdown of DM response windows, and the tolerance is far shorter than most shops assume.
2. The greeting that arrives before the question
People often send their question first and your greeting lands on top of it. Now the buyer's real message is sitting above a generic block of text, and when a human finally opens the chat, the thread reads like it's already been handled. Messages get skipped this way every single day.
3. The quick reply that's technically correct
A customer asks: "I need the beige one by Thursday, is that possible?" Your saved snippet fires: "Shipping is 2–4 business days nationwide. Orders over $50 ship free." Every word true, and the buyer still doesn't know whether beige exists or whether Thursday works. They now have to ask again — and roughly a third of people simply don't.
Canned text scales your tone. It doesn't scale your answers.
What the arithmetic looks like
Take a mid-sized shop getting 90 WhatsApp messages a day. Say 70 are genuine first-contact questions, and 40% of those land outside working hours — evenings, Fridays, the hours when people actually browse. That's 28 conversations a day meeting an away message.
Assume a modest 5% of first-contact chats would have become orders if someone had answered. At an average basket of $60, those 28 nightly conversations are worth around $84 a day, or about $2,500 a month in orders that never get placed. Not lost to a competitor's better price. Lost to a sentence about business hours.
Run that arithmetic on your own inbox before you decide auto-replies are "good enough for now." Most sellers are surprised by which side of the ledger the automation is sitting on.
Auto reply vs AI assistant: the difference buyers feel
An auto reply is a trigger and a fixed string. An AI assistant reads the message, checks what you actually sell, and answers the specific thing that was asked.
Same customer, same 22:40 message — "hi, do you have the beige tote in stock? how much?"
| Auto reply | AI assistant |
|---|---|
| "Thanks for your message! We'll respond during business hours." | "Yes, the beige tote is in stock — $45, and it ships within 24 hours. Want me to hold one for you?" |
The first message costs you nothing and earns you nothing. The second either closes or produces a real objection you can work with, at 22:40, while the buyer is still in the mood.
The practical test is simple: if your automation can't answer "is it available and how much?", it isn't handling customers. It's handling appearances.
Keep the auto-replies. Give them a smaller job.
None of this means switching everything off. Greeting and away messages have one legitimate use: setting expectations, precisely. The rule is that a message should either answer something or promise something specific. Vague reassurance does neither.
Rewrite them like this:
- Weak: "We'll reply as soon as possible." → Better: "We reply to every message before 11:00 tomorrow. If you need it faster, send 'urgent' and we'll call you."
- Weak: "Thanks for contacting us!" → Better: "Send the product name or a screenshot and we'll confirm the price and stock in one message."
- Weak: "Our working hours are 9–18." → Better: "Orders placed before 15:00 go out the same day."
Each of those does work. The last one has closed sales on its own.
A layered setup that works
The setups that hold up under volume have three layers, not one:
- Instant answer layer. An assistant that knows your catalogue handles price, stock, sizing, shipping and returns — around the clock, in your wording, without a person touching it.
- Escalation layer. Anything involving a discount, a complaint, a custom order or an angry customer goes to a human, with the conversation already summarised. Automation should hand over cleanly, not bluff.
- Expectation layer. Your greeting and away messages, now doing the narrow job they're good at — telling people exactly when a human is coming, if one is needed at all.
If you're building this across more than one channel, the same logic applies to Instagram and Telegram. Our guide to WhatsApp automation marketing covers the campaign side of it.
What to measure once it's live
Response rate stops being a useful number the moment you automate, because it goes to 100% and stays there. Watch these instead:
- First-answer resolution — what share of buyers get their actual question answered in the first reply, with no follow-up needed.
- Handoff rate — how many chats need a human. Falling over time means the assistant is learning your catalogue properly.
- Night-and-weekend conversion — the clearest signal of all. If sales outside working hours are still near zero after a month, your automation is decorating your inbox, not serving it.
- Abandoned threads — conversations where the customer stops replying after your automated message. This is the number that exposes bad canned replies.
Track them for four weeks before deciding anything. Auto-reply performance is easy to feel good about and hard to be right about.
The short version
Automatic replies are a courtesy, not a salesperson. Use greeting and away messages to make specific promises, keep quick replies for the parts a human is already handling, and put a real assistant in front of the five questions that make up most of your inbox. That's the difference between an inbox that looks managed and one that produces orders on a Friday night.
If you'd rather not maintain a pile of canned text that ages badly, Vardast connects to WhatsApp, Instagram and Telegram and answers customers from your own product information, in your own tone, handing the conversation to you when it should. Sellers usually notice it first in the messages that used to sit unanswered until morning.