Before You Press Send: Audit Your WhatsApp Broadcast List Like an Operator, Not a Marketer

Most WhatsApp broadcast advice starts at the message: a better offer, a better template, a better hour. It skips the question that decides whether a broadcast is defensible at all: who is on the list, and how do you know they agreed to be there?

I know this from the receiving end. A shop I bought from once sends offers for two years. A number I never gave to anyone wishes me a happy festival. A business I asked to stop sends a "final reminder" that is not final. None of those senders set out to be careless. They never audited a list before pressing send. The list grew, the consent blurred, and the send button asked no questions. So let us do the audit first, on one real list. If you do it honestly, the tool question mostly answers itself.

Sort first, send second: the three buckets

Put every number into exactly one of three buckets. No fourth bucket, no "probably fine" pile. A number you cannot place with evidence goes in bucket three, and bucket three does not get sent to. That rule alone will shrink most lists, and the shrinkage is the point.

Bucket one is opted in, with proof you could show someone. Bucket two is opted out, including everyone who ever said stop, in any wording, on any channel you control. Bucket three is never asked: imported, scraped, bought, shared by a partner, copied from an old spreadsheet, or collected for a different purpose and never given a WhatsApp choice. No "but they are a customer" override. Being a customer is not agreeing to marketing messages, and a purchase receipt is not an opt-in.

People who use a disposable inbox, like the free tool on this site's homepage, understand this instinctively: permission is specific and revocable. A broadcast list deserves the same respect, in the other direction. Your customers gave you a number for a reason. The audit asks whether marketing was that reason.

Bucket one: the opted-in list you can actually prove

A real opt-in has four parts: who, when, where, and what for. Who gave the number — its owner, not a friend, a form-filler, or a list broker. When — a date you can point to. Where — the form, chat, counter card, or checkout box that actually said WhatsApp marketing. What for — the kind of message agreed to, usually offers and updates, not every message you might ever want to send.

If a number fails any of those four, it is not in bucket one yet. It may belong there later, after you ask properly and the person says yes. It does not belong there now because you hope they would say yes.

My opinion, unsoftened: 400 proven opt-ins beat 4,000 maybes on every measure that matters. The proven list gets read, because its people expect the message. The maybe list gets reported, blocked, and ignored, and every report teaches Meta something about your number. A smaller, provable list is not a sacrifice for compliance. It is the list that actually works.

WhatsApp's rules reinforce this. Broadcasts go through approved message templates, which Meta reviews before they can be sent, and they go to contacts who have opted in. A tool can organise opt-ins, import them by CSV or Google Sheets, group them with labels and custom fields, and respect opt-outs automatically — WhatsMax's landing page says all of that. But no tool can manufacture consent for a number that never gave it.

Bucket two: the opted-out list that must stay out

Opt-outs rarely arrive tidily. One person replies "stop". Another writes "please remove me". A third writes "not interested" and never returns. A fourth unsubscribes on email and assumes that covers WhatsApp too. A fifth tells a staff member in the shop, and nobody writes it down.

Catch all five. Search the inbox for stop-words, not just "stop". Check contact notes. Ask staff who take messages in person. Then make the exclusion automatic, not a matter of memory. WhatsMax's landing page says opt-outs are respected automatically, and that an opted-out number is recorded and excluded from future campaigns. That is the right shape: a system rule, not a human memory. But the system can only exclude what it knows, so the audit must get every known opt-out into the system before the send.

Watch one subtler trap: the person who opted out of one thing but not another. Someone who stopped daily offers but still wants delivery updates is opted out of offers, not of everything. Someone who stopped everything is out of everything. Groups and labels must carry that distinction, or a careful opt-out gets undone by a careless segment. When in doubt, treat the broader opt-out as true. You can always ask again, properly. You cannot unsend a message to someone who asked you to stop.

Bucket three: the never-asked list, and why it stays unsent

Bucket three is where most broadcast trouble lives, and where most arguments happen. "But they bought from us." "But they are in our group." "But a partner shared the list." "But everyone does it." None of those is an opt-in.

Take the partner list, because it sounds most reasonable. A partner gives you 2,000 numbers of "people interested in this kind of product". Those people were interested in the partner's product, on the partner's terms, and they never met you. Sending them your offer is a cold message wearing a borrowed introduction. The honest move is to ask the partner to send one message on your behalf, inviting people to opt in to you directly. Those who say yes become bucket one. The rest were never yours to message.

The same logic covers scraped numbers, purchased lists, and numbers harvested from groups — and the quietest version, a list collected for one purpose and reused for another. A customer gave a number for delivery updates; now it receives weekly offers. That customer agreed to be told when their order arrived, not to weekly offers. My rule is simple, and expensive in the short term: do not send to bucket three, and do not keep it in the same audience as bucket one, where one wrong click can include it. Keep it separate, labelled unasked, and work it only by asking, through a channel where you do have permission. Some people will say yes. Those are the only ones who ever belonged on a broadcast list.

The template check: what Meta sees before your customer does

Once the list is sorted, audit the message, starting not with wording but with the template. Broadcasts do not go out as free text. They go out as approved templates, in marketing, utility, or authentication categories, which Meta reviews. You write the template, map its variables — the landing page gives the example of {{1}} mapped to contact fields, so each customer gets their own name or offer — submit it from the console, and track its approval status.

Audit it the way Meta will, then the way a customer will. Meta asks about category and policy: is this marketing dressed as utility, is the wording misleading, does it promise what it cannot deliver? The customer asks simpler questions: does this know who I am, does it say one thing clearly, and can I tell what happens if I reply? A template that passes Meta and confuses the customer is still a bad message. One that delights the customer and misstates its category will not be sent at all.

Check variables last, because they embarrass broadcasts most. A variable mapped to the wrong field turns a personal message into a public mistake: the wrong name, the wrong offer, a blank where a price should be. Send the template to yourself and one colleague first, with real contact data, not test data. If the test looks wrong, the broadcast will look wrong to hundreds of people at once, and WhatsApp has no recall button.

The 24-hour window check

The second message check is the clock. WhatsApp allows free-form replies only for 24 hours after the customer's last message. After that, you need an approved template to reopen the chat. The broadcast itself is a template, so it can go out — but the conversation it starts lives inside the window. If customers reply on day two and your team answers on day four, the team needs a template, not a casual reply, and an inbox that hides the countdown sets staff up to reply late by accident.

WhatsMax's landing page makes a point of the 24-hour service window being shown on every conversation, and says its drips and follow-ups are window-aware: free-form inside 24 hours, templates after. Check that in the demo before trusting it: open a conversation, find the window state, and confirm a closed window offers a template rather than a free-text box that will fail.

There is a human version too. If the broadcast goes out Friday evening and nobody returns until Monday, who answers the replies? A broadcast that starts conversations nobody staffs is worse than no broadcast. Send when someone can reply, or set a bot for routine questions until a human returns — and be honest about which questions a bot can handle. "What time do you close" is a bot question. "My order is wrong" is not.

Pacing, quiet hours and the per-recipient log

A broadcast is not one message; it is hundreds, and how they leave matters. WhatsMax's landing page describes campaign sending as atomic, with per-number pacing, quiet hours, and a pause control on a running campaign, plus sent, delivered, and read percentages with a per-recipient log behind them. Each prevents a different failure: pacing prevents the burst that looks like spam, quiet hours prevent the 2 a.m. offer that turns an opt-in into an opt-out, and pause stops a wrong price or wrong audience after 200 sends instead of 2,000.

The per-recipient log is the check I care about most, because percentages flatter everyone. A campaign reporting 92 percent delivered sounds excellent until you ask who the 8 percent were. The log answers that, and it answers the more important question: who opted out after receiving it. That group is your real feedback. If opt-outs spike after a broadcast, the problem is the list, the message, or the frequency, and the log tells you which. Read it before the next send, not after the third.

What WhatsMax changes, and what it cannot change

This has been an audit, not a review, deliberately. A review that starts with features will always find features. An audit that starts with a list finds the truth, and then the tool can be judged against it.

WhatsMax, listed today on the shop at $44 one-time for a Regular License with an instant ZIP download, is a multi-vendor WhatsApp marketing platform: campaigns, a shared team inbox, bots and a visual flow builder, a product catalog, contacts with a kanban board, drip sequences, and billing with reseller support, run from a business WhatsApp number. Its live demo runs at mavora.tempomailusa.com, where a free plan is included with no credit card to start, according to the demo's own landing page. Do not confuse that $44 script price with the demo's subscription plans, which are what a platform owner charges their vendors: Free at $0, Growth at $29 a month with 10,000 messages a month, and Scale at $99 a month with unlimited contacts, campaigns, and messages. Meta's per-conversation charges are separate and set by Meta, as the landing page says. Its headline claims, quoted as its claims and not my findings, are 28 paid add-ons built into the core at no extra cost, 100 percent of webhooks verified by signature before anything is processed, and payments through eight gateways with every payment verified before a plan activates.

What that changes for an audit is organisation and evidence. A shared inbox puts opt-outs and history in one place instead of on a passed-around phone. A template manager tracks approval instead of guessing it. Groups, labels, and a kanban board keep the three buckets as separate audiences instead of one spreadsheet where a sort error includes the wrong people. A per-recipient log means the next audit starts with data instead of feelings. Three guides ship with the product, the landing page says — an Admin Handbook, a Vendor Guide, and a Staff Quick Guide — suggesting the maker has watched three different people get confused, which is a better sign than a feature count.

What it cannot change is the list itself. It cannot turn a never-asked number into an opted-in one, make a dishonest template honest, staff your inbox on a Sunday, or undo a broadcast sent to bucket three. If your bucket one is small, WhatsMax will not fix that alone. It will help you run the small list well, and grow it properly, by asking. That is a real benefit, and a limit worth stating plainly: a tool that promises to fix a bad list is selling you a bigger problem.

The audit sheet I would run

For one list, before one send, in order: export the audience and count it, by source and period. Sort into the three buckets, with proof for bucket one and a reason for every number in bucket two. Move bucket three out entirely, labelled unasked, not "cold leads". Check every opt-out channel and confirm each exclusion in the system. Check the template — category, wording, variables, and a test send to real data. Check the window — who replies, when, and with what. Set pacing and quiet hours, confirm pause, and decide in advance what would make you pause. Send, then read the per-recipient log before planning the next one.

Run it once, properly, and the first send is smaller than planned while the second is better than the first, because it goes to people who actually wanted it. That is not disappointing. That is the audit working.

Who should not send at all

If you cannot prove opt-in for most of the list, do not send to it. Ask first, through a channel where you do have permission, and send to those who say yes. If customers gave a number for delivery and nothing else, do not borrow it for offers. If nobody will staff the replies, do not start the conversations. If the offer is not ready, the price is not final, or the template is not approved, do not send a placeholder and fix it later. A broadcast is not a draft; it arrives, all at once, in people's personal chats.

And if you are one person answering your own business WhatsApp, and the WhatsApp Business app already does what you need, do not buy anything. A shared inbox, template manager, and per-recipient log are for a team and a list that have outgrown a phone. Try the demo's free plan first and see whether the audit sheet actually needs the machinery. The earlier post in this series toured what a day on WhatsMax feels like for a shop: the day shows the work, this post shows the check before the work starts.

Frequently asked questions

What is a WhatsApp broadcast opt-in audit?

A pre-send check of one list: sort every number into opted-in with proof, opted-out, or never-asked, verify the template and the 24-hour window, set pacing and quiet hours, and confirm you can read the per-recipient log afterwards. Only the proven opt-in bucket gets sent to.

Does buying WhatsMax give me permission to message a list?

No. WhatsMax is a platform for campaigns, inbox, bots, and billing. Permission comes from the contact, not the tool. A purchased, scraped, or partner-shared list does not become opted-in by being imported.

How do I prove an opt-in?

Record who gave the number, when, where, and what for: the form, chat, or counter card where the number's owner agreed to WhatsApp marketing. If any of the four is missing, treat the number as unproven and ask again before sending.

What counts as an opt-out?

Any clear request to stop, in any wording, on any channel: "stop", "remove me", "not interested", an unsubscribe on another channel the customer reasonably meant to cover WhatsApp, or a verbal request to staff. Record it and exclude that number from future campaigns.

Why do broadcasts need approved templates?

WhatsApp requires broadcasts to use message templates, which Meta reviews before they can be sent, in marketing, utility, or authentication categories, with variables mapped to contact fields. Replies inside the 24-hour window after a customer's last message need no template; after it closes, a template reopens the chat.

What does the $44 WhatsMax price actually buy?

Per today's shop listing, $44 one-time for a Regular License buys the script itself, with an instant ZIP download, to run your own platform on your own server. The demo's Free, $29 Growth, and $99 Scale plans are what a platform owner charges vendors, and Meta's per-conversation charges are separate and set by Meta.

A broadcast list is not an asset you own. It is a set of permissions people lent you, and permissions can be withdrawn, forgotten, or misunderstood. Audit the list, send only to the bucket you can prove, and read the log after. Do that, and the tool — WhatsMax at $44 for the script, or any other platform — becomes what it should be: a way to run a good list well, not a way to send a bad list faster. If you only ever need a throwaway inbox for sign-ups, the free tool on this site's homepage will still be here, and the blog index has the rest of this series, including the shop-day tour of WhatsMax and a closing-time test for a POS script that judges a tool by its record, not its promise.