A seller I was watching last winter did what almost everyone does eventually. She had roughly 800 numbers saved from a year of orders, built one broadcast list, wrote one message about a weekend sale, and sent it at eleven at night.
By morning she had six orders and about forty people who had blocked her.
She counted the six. I would have counted the forty.
Broadcasting on WhatsApp is the cheapest reach an online seller will ever get, and it is also the fastest way to ruin a contact list that took a year to build. What separates those two outcomes is not the size of the discount. It is who receives the message, how often, and whether it reads like it was written by a person who remembers what you bought last time.
Broadcast lists, groups and API templates are three different tools
Sellers use the word "broadcast" for all three, then wonder why the results look nothing alike.
- Broadcast lists live inside the WhatsApp Business app. Up to 256 contacts per list, and each message lands in a normal one-to-one chat, so replies come back privately. There is one rule that quietly kills half of these campaigns: the recipient must have your number saved in their phone, or the message never arrives. No error, no warning, no delivery. It simply does not show up.
- Groups are a room, not a channel. Everyone sees everyone. One customer complains about shipping in front of 200 others, and half the members mute the thing within a day. A group can work for a small community of repeat buyers who already like each other. It is a bad way to announce a sale.
- API templates, on the official WhatsApp Business Platform, are pre-approved message formats you can send to anyone who opted in, with no saved-number requirement, priced per conversation. This is the only one of the three that scales past a few hundred people. If you keep bumping into the 256 ceiling, that is your cue to read about when an online store should move to the WhatsApp Business API.
Choose the wrong one and the campaign is decided before your copy matters at all.
Opt-in is a delivery mechanism, not a legal checkbox
Most sellers treat permission as paperwork they can skip. It is not paperwork. In the app, permission shows up physically as a saved contact. On the API, it shows up as your quality rating. Both of them decide whether your next message is seen by anyone.
The easiest opt-in is a question you were going to ask anyway:
- "Want me to message you here when this comes back in your size?"
- "Should I send you the new drop before it goes on the page?"
- Right after an order ships: "Save this number so my updates actually reach you."
That last line does more for your reach than any caption you will write this month. Ask for it in the minutes after someone pays, while they still like you. Nobody saves a stranger's number two weeks later.
Your list of 800 is really four lists
Sending one message to everybody is the most expensive habit in this whole channel. Split by behaviour, not by age or city:
- Bought in the last 60 days. Warm, know your name, highest reply rate you will ever see.
- Bought once, more than six months ago. They need reminding who you are. A short "we have your size again" beats a discount.
- Asked about something and never bought. The most neglected group on any seller's phone. They already told you exactly what they want.
- Opted in, never bought anything. Handle gently, send rarely, judge them on replies rather than orders for a while.
Now run the arithmetic on that 800-contact list. If group three is 120 people who each asked about one specific item, a message naming that item will beat a generic sale blast so badly that comparing them stops being fair. One product, one person's real question, answered a month later than they expected but exactly when it matters.
You do not need a fancy tool for this. A note in each contact's name field works. "Sara boots 38" is a segment.
Frequency: the line between welcome and muted
Here is a rule I will defend: one promotional broadcast per week, maximum, to any single contact. Transactional messages such as order shipped, tracking code, back in stock in your size do not count against that budget, because people genuinely want those.
Twice a week and reply rates start sliding. Daily, and you are training people to swipe you away without reading, which is worse than being blocked. Being blocked at least stops the bleeding. Being ignored costs you money on every send and tells you nothing.
A rhythm that holds up over months looks roughly like this:
- One merchandising message a week, to one segment, not the whole list.
- Restock alerts sent only to the people who asked about that exact item.
- A short seasonal or campaign message a few times a year, to everyone.
- Order and shipping updates whenever they happen, no limit.
What a broadcast that gets replies actually looks like
The bad version, which you have received yourself:
"MEGA SALE 30% OFF EVERYTHING LINK IN BIO"
There is nothing to reply to. Nothing addressed to me. It could have been sent to anyone, and it was.
The better version:
"Hi Sara, the black boots you asked about in March are back and size 38 is in. Want me to hold a pair until tomorrow evening?"
Same product, same discount available the moment she asks about price. The difference is that it ends in a question she can answer with one word while standing in a queue. That is the whole trick, and it breaks down into four parts:
- One product. Not the catalogue, not the whole season.
- One reason you are writing today. Restock, a size back, last few pieces, price going up on Saturday.
- One question at the end that can be answered with "yes", a size, or a number.
- A way out. "Tell me to stop and I will take you off this list." It costs you a handful of contacts and saves you a report.
How you actually get blocked and reported
Blocks are rarely about the offer. They come from a short, boring list of habits:
- Sending to numbers that never opted in: bought lists, scraped numbers, everyone who once commented on a post.
- Sending identical text to hundreds of people so it reads like a printed flyer.
- Sending at 11pm, or at 8am on a public holiday.
- Offering no way to unsubscribe, which makes blocking the only exit available.
- Following up three times on a message nobody answered. Once is a follow-up. Three times is pressure, and people remember pressure.
On the API side this is not a matter of feelings. Blocks and reports pull your quality rating down, and a low rating means throttled template sends or a restricted number. You can lose the channel outright. In the app, heavy blocking degrades your delivery quietly, with no dashboard to warn you. The punishment for spam always arrives late and lands hard.
Measure replies, not sends
"We reached 800 people" is not a result. Delivered is not a result either. Five numbers are worth your attention:
- Reply rate: replies divided by delivered. This is the health metric for the whole channel.
- Opt-out and block rate. If it climbs campaign over campaign, you are overdrawing an account you cannot top up.
- Orders within 72 hours of the send, not clicks, not views.
- Revenue per 1,000 messages sent, which is the only honest way to compare a small warm segment against a big cold one.
- Cost per order, if you are paying per conversation on the API.
A campaign to 120 warm buyers that returns 14 replies and 5 orders beats a blast to 800 that returns 6 orders and 40 blocks, and it is not close. For the wider picture across all your chat channels, the six DM sales metrics worth tracking are the ones I would put on a wall.
Broadcasts do not fail at send. They fail forty minutes later
This is the part nobody plans for. A good broadcast to a warm list produces a wall of replies inside the first hour, and almost all of them are questions. Is it available in 39. How much is shipping. Can I pay when it arrives. Do you have it in brown. Is the sale still on tomorrow.
If those replies sit unanswered for six hours, you did not run a campaign, you made an announcement, and the sale went to whoever answered first. Everything after the send button is a speed problem.
Two things fix it. First, set up your WhatsApp catalogue so it can take orders, which answers price and variant questions before anyone has to type them. Second, have something replying instantly while you are asleep, at the dentist, or packing the last twelve boxes.
That second part is why Vardast exists. It sits on your WhatsApp, Instagram, Telegram and website, answers the flood of size, price and shipping questions in your own voice using your real product information, and hands the conversation over to you the moment it becomes something a human should handle. You send the broadcast. It makes sure nobody is left waiting.
Start smaller than you want to
If you are building this out this week, do not start with 800. Take your 100 warmest contacts, write one message naming one product, end it with one question, and send it at ten in the morning. Count the replies. Do it again next week with what you learned, and only then widen the list.
A list of 800 people who read you is worth more than a list of 8,000 who muted you. That is the entire discipline, and it takes about three campaigns to feel it.