Test Sends & Previews Before a WhatsApp Broadcast

A test send is a real WhatsApp broadcast delivered to a single phone number you choose, using the same template, the same header file, the same buttons and the same personalisation as the live campaign — so you see exactly what your customers will see, on a real phone, before anyone else gets it. It is not a simulation and it is not a mockup. It goes through Meta's servers and lands in a real chat.

That distinction is the whole point. The on-screen preview in the composer shows you the body text and nothing else. Only a test send exercises the full path: template rendering, media upload, button links, and variable resolution. Sixty seconds of testing is the cheapest insurance in broadcasting, because the alternative is a typo, a broken link or a raw {{1}} reaching a list of several thousand people who cannot unsee it.

Where the test send lives

In the broadcast composer, scroll to the section headed 5. Review & test. Its hint line reads: Check how many people will receive this, and send yourself a test first. That section contains three things — the recipient preview, the estimated sending time, and the test field itself.

1. Finish building the broadcast first: pick the template, upload the header file, fill in any button values, and choose your audience. A test on a half-built broadcast tests half a broadcast.
2. In the field labelled Send a test to one number, type a phone number in full international format, for example +923001234567.
3. Click Send test. You should see “✓ Test sent — check that number.” within a second or two.
4. Open WhatsApp on that phone and read the message there, not in the builder.

Two practical notes. The field takes one number at a time — there is no multi-recipient test list, so to check with three colleagues you send the test three times. And the test-send button is available to admin users, so if a team member cannot see it, that is a permissions matter rather than a bug.

Preview panel versus test send versus live send

These three things check different amounts of the campaign, and confusing them is how broken broadcasts get launched.

Read the middle column carefully. A test send covers almost everything except the audience — and the audience is checked by a separate control described below.

What the on-screen preview does and does not show

The panel labelled Preview in the composer renders the template's body text only, as plain text. It is a wording check, not a WhatsApp bubble mockup. It deliberately does not draw the header, the footer or the buttons, because the only rendering that counts is WhatsApp's own on the recipient's device.

Separately, when you upload a header file the composer shows a small local thumbnail so you can confirm you picked the right image or video off your computer. That thumbnail is proof you selected the correct file; it is not proof WhatsApp accepted and delivered it. Only the test send proves that.

Checking who will actually receive it

A test send tells you nothing about your audience, so use the button labelled Preview recipients in the same section. It returns a line such as “1,842 recipient(s) will receive this”, followed by a note about what was removed.

ConvoFly strips three groups automatically before that count is produced: duplicate numbers, contacts who have opted out, and numbers previously found not to be on WhatsApp. When any of those apply, the message spells it out — for example “37 opted-out contacts and 12 previously-undeliverable numbers (not on WhatsApp) skipped”. If your count is dramatically lower than expected, that line is usually the explanation.

Treat a wildly wrong count as a stop signal. A number that is 10× too big usually means a segment filter did not apply; a number close to zero usually means it applied twice. Fix the audience before you fix anything else.

How long the send will take

Under the recipient count, ConvoFly appends an estimate: “⏱ Estimated sending time: about 15 minutes (WhatsApp limits bulk sending speed).” The estimate assumes roughly 120 messages a minute, which is the pace bulk sending realistically runs at.

This matters for timing-sensitive campaigns. If you want a lunchtime offer to land at lunchtime for everyone, a list that takes 90 minutes to drain needs to start well before noon. It also matters for countdown offers: the last person on a long list sees measurably less time on the clock than the first. And it explains the most common false alarm in broadcasting — a send that is still running an hour later is usually normal pacing, not a fault. The same tier and pacing rules are covered in messaging limits and tier caps.

How personalisation behaves in a test

This is where a test send earns its keep, because different sources behave differently:

• Contact attributes. Variables mapped to a contact field are resolved against the test number's own contact record. If that number exists in your contacts with a first name stored, you see the real name. If it does not, you see the default value you configured. Either way you learn something: seeing the fallback is how you confirm the fallback reads naturally.
• Google Sheet columns. When the audience comes from a Google Sheet, a test shows either the column's fallback value or the column name in brackets, such as [FirstName]. That confirms the variable is wired to the right column even though no sheet row is being sent.
• CSV uploads. A test renders these as [CSV value #1] and similar, so you can confirm the ordering of your columns matches the numbering of the placeholders.

The failure this catches most often is an off-by-one: {{1}} and {{2}} swapped, producing “Hi Lahore, your order to Ayesha is on its way.” In a preview panel that is invisible. In a test send it is obvious. More on variable setup in personalising broadcasts with variables.

Tracked buttons in a test send

If you enabled Track clicks on this button, real recipients each get a unique tracking token in the URL. A test send has no real recipient to attribute a click to, so ConvoFly substitutes a placeholder value in the link. The button will render and be tappable, but treat the tracked link as a rendering check rather than an end-to-end click test, and confirm the destination you typed into the Destination URL field is the page you meant. Click reporting is covered in tracking clicks in WhatsApp broadcasts.

Does a test send cost money?

Yes. A test is a genuine template message, and Meta states that “you are only charged when a template message is delivered” (Meta, WhatsApp pricing). A delivered test is a delivered message. Rates vary by country and Meta publishes them in separate rate cards, but the order of magnitude is one message against a campaign of thousands — a rounding error against the cost of resending a broken campaign. Marketing templates are charged on delivery; there is no free test mode.

It also consumes one unit of your daily messaging allowance if the test number has not messaged you in the last 24 hours, since that allowance counts unique phone numbers reached outside a customer service window (Meta, messaging limits). On the starting tier of 250 that is 0.4% of your day. Test anyway.

A nine-point pre-launch checklist

1. Click Preview recipients and confirm the count is within range of what you expected.
2. Read the exclusions line — opted-out and not-on-WhatsApp numbers — and make sure the size of that deduction makes sense.
3. Note the estimated sending time and check the last message will still land at a sensible hour.
4. Send test to your own number.
5. Read it on a phone, in the WhatsApp app, at normal screen brightness.
6. Check the header image is the right file and is not cropped badly on a narrow screen.
7. Tap every button. Confirm each link opens the page you intended and any coupon copies correctly.
8. Check every personalised value, and specifically check what the fallback looks like for a contact with nothing stored.
9. If the template has a countdown, confirm the expiry time is still in the future when the last message will land.

Then, and only then, schedule or send.

What a test send cannot tell you

• Whether your audience is correct. That is what the recipient preview is for.
• Whether tracked clicks will attribute correctly, since the test uses a placeholder token.
• Whether you will hit your daily tier cap. Compare your recipient count against your tier before launching.
• Whether the message is a good idea. A test proves the message renders; it does not prove anyone wanted it. Frequency and relevance are covered in managing opt-outs, frequency and quality.

Build the habit into the routine rather than treating it as an optional extra. Every broadcast, one test, read on a phone. It is the single cheapest step in the whole process and the only one that catches mistakes while they are still free to fix. If you are sending your very first campaign, walk through it alongside how to send your first WhatsApp broadcast.