Click to WhatsApp ads: what agencies must build
Click to WhatsApp ads send a paid click straight into a WhatsApp thread with your business, and the moment that first message lands, a 24 hour timer starts running. Everything that decides whether the spend pays back happens inside that timer: who answers, how fast, on which client account, and whether the reply costs you nothing or opens a billing line.
This one is written for the person buying those ads on behalf of several clients, or running several closers against one inbound flow: agency owner, head of sales, ops manager. Below are the rules Meta publishes about the clock and the billing, what they force you to build before launch, and the routing model that turns a click into a booked call. You will not find a cost per lead benchmark here, because the public ones are not verifiable and yours would be different anyway.
TL;DR
- Meta's Marketing API documentation describes the format plainly: ads that click to WhatsApp "send people who click on your ads directly into conversations with your business in WhatsApp". Buying the ad is the easy half.
- A clock starts on contact. "When a WhatsApp user messages you or calls you, a 24-hour timer called a customer service window starts." Once it closes, "you can only send pre-approved template messages."
- Meta's pricing documentation describes a second window on top of it. When someone reaches you through a click to WhatsApp ad and you respond inside 24 hours with any type of message, that message is free and opens a Free Entry Point window of 72 hours in which any type of message is free.
- So reply speed is not a service standard on this channel, it is a cost decision. The same shift that misses the window also converts a free thread into a paid template thread.
- Multi account is where it breaks. Each client needs its own Page, its own linked number, a named owner per shift, and its own reporting line.
Table of contents
- What click to WhatsApp ads actually are
- The clock that starts on the first message
- The free window most agencies throw away
- Why response time becomes a cost line
- What breaks across several client accounts
- Routing, from ad click to booked call
- What to build before the first dollar is spent
- The numbers to instrument, and the ones not to copy
- Where SetScale fits
- FAQ
- Conclusion
What click to WhatsApp ads actually are
The format is documented, so start from the documentation rather than from a case study. Meta's Marketing API guide for click to WhatsApp states that these ads "send people who click on your ads directly into conversations with your business in WhatsApp" (page consulted 7 August 2026).
On creative, the same page says the format "support ads with an image, a video, a carousel, or a slideshow" and that you "can also include call prompts in your ad". That is a wider creative surface than most teams assume, and it means your existing static and video assets are usable without a rebuild.
The setup requirements are the part that decides your agency architecture. The guide lists an ad account with Meta, a Facebook page with a WhatsApp number linked to it, assets uploaded to Meta beforehand, a Page access token from someone authorised to advertise on that Page, and the ads_management, pages_manage_ads, pages_read_engagement and pages_show_list permissions.
Read that list again with ten client accounts in mind. A Page with a linked WhatsApp number is a per client object, not a per agency one, and so is the access that lets you advertise on it. Whoever owns those objects owns the channel.
Verdict: the ad is a commodity. The account structure underneath it is the thing you are actually buying.
The clock that starts on the first message
The moment the click becomes a message, you are inside a documented timer. Meta's Cloud API messaging guide is explicit: "When a WhatsApp user messages you or calls you, a 24-hour timer called a customer service window starts" (page consulted 7 August 2026).
While that window is open, the same page says you can send the service message types it lists. When it shuts, the rule changes completely: "When the window closes, you can only send pre-approved template messages."
There is one piece of good news in the mechanics. "If the user messages or calls you again before the timer expires, the timer resets to 24 hours." A live conversation keeps itself alive. A conversation you let go cold does not come back on its own terms.
This is not a WhatsApp quirk to file under compliance. It is the operating envelope of the channel, and it is the same envelope your response time commitments have to live inside on every other inbound source.
Verdict: you do not get to negotiate this clock, you can only staff for it.
The free window most agencies throw away
Here is the fact that changes the media maths, and it is the one least often repeated correctly. Meta's WhatsApp pricing documentation describes a Free Entry Point mechanism for exactly this traffic (page consulted 7 August 2026).
When a user reaches you through a click to WhatsApp ad, a 24 hour customer service window opens as normal. If you respond within those 24 hours using any type of message, that message is free and a Free Entry Point window opens on top. That second window runs for 72 hours, and during it you can send any type of message to the user at no charge.
Set that against the default billing rule on the same page: you are only charged when a template message is delivered. Non template messages inside an open customer service window are free, and utility templates delivered inside an open customer service window are free too. The page also notes that conversation based pricing is deprecated and was replaced by per message pricing on 1 July 2025, which matters if your internal cost model was written before that date.
So the sequence an agency wants is simple to state and hard to staff. Ad click, inbound message, human or automated reply inside 24 hours, then a 72 hour runway in which the follow up, the qualification and the booking link all cost nothing.
Miss the reply and you get the opposite: no free runway, and a re engagement that has to travel as a pre approved template with a delivery charge attached.
Verdict: on this channel, the first reply is a purchasing decision disguised as a customer service task.
Why response time becomes a cost line
Most teams treat first response time as a quality metric that lives on a client review slide. On click to WhatsApp traffic it stops being that and becomes a line in the cost of the campaign, because the documented windows attach a price to lateness.
| Window | What opens it | How long | What it allows |
|---|---|---|---|
| Customer service window | The user messages or calls you | 24 hours, reset by each new user message or call | The service message types listed in the Cloud API guide |
| Free entry point window | A reply within 24 hours to a user who arrived through a click to WhatsApp ad | 72 hours | Any type of message, at no charge |
| Outside both windows | Nothing, the timers have run out | Until the user writes again | Pre approved template messages only, and templates are the billable object |
Two consequences follow, and neither is about typing faster.
The first is that coverage beats speed. A median first reply of four minutes on weekday afternoons is worthless if Friday evening leads sit until Monday, because those are the ones that fall out of both windows. The gap that costs you money is the calendar gap, not the seconds.
The second is that the free runway is only free if someone uses it. A 72 hour window in which nobody follows up, qualifies or sends a booking link is a window you paid to open and then let expire.
Verdict: put first response time on the media report next to spend, not on the service report next to satisfaction.
What breaks across several client accounts
Running this for one brand is an inbox problem. Running it for ten is an ownership problem, and the failure modes are predictable.
- Object sprawl. Each client brings a Page and a linked WhatsApp number. There is no single agency inbox that legitimately collapses them, and pretending otherwise creates the exact handoff gaps that miss 24 hour windows.
- Ambiguous ownership. When a conversation belongs to everyone on the team, it belongs to nobody at 7pm. Every client account needs a named owner per shift, not a rota that exists only in someone's head.
- Invisible per client failure. Agency wide averages hide the account that is quietly missing every window. The same trap applies here as in running many client accounts on social: report per account or you will discover the problem when the client does.
- Access that leaves with a person. Page access tokens and advertising permissions granted to an individual become an incident the week that individual is on holiday.
- Reporting that cannot be handed over. If your setting sits inside a service you resell under your own brand, the per account view is the deliverable, not a nice to have.
Verdict: the constraint is not volume, it is that every one of these objects is per client and none of them is per agency.
Routing, from ad click to booked call
A workable route has five steps, and the first three all sit inside the first 24 hours.
- Capture with attribution intact. The conversation has to arrive tagged with the client account and the campaign that paid for it, before anyone touches it. Attribution reconstructed later is attribution you will argue about later.
- Reply inside the window, always. This is the step that opens the free 72 hour runway. It is also the step that should never depend on whether a particular human is awake.
- Qualify against the client's actual criteria. Budget, timing, fit, whatever that client sells against. A reply that is fast and useless still burns the runway.
- Hand off to a closer with the context attached. The handoff between setter and closer is where most booked calls die, and a thread that arrives without its qualification notes forces the closer to start over.
- Book, confirm and instrument. The booking lands in a calendar that someone owns, and the outcome flows back to the campaign that produced it, which is the only way a closer's calendar stays full for a reason you can repeat next month.
Verdict: steps 1 and 2 are infrastructure, steps 3 to 5 are where judgement earns its keep. Teams usually automate the wrong one first.
What to build before the first dollar is spent
A short build list, in the order that stops the expensive mistakes.
- One Page with a linked WhatsApp number per client, owned by an account that survives staff turnover.
- A named owner per client account per shift, including evenings and weekends, written down where the client can see it.
- A first reply path that fires inside 24 hours without a human bottleneck, plus an escalation path when it needs a person.
- A qualification script per client, because the criteria that matter are theirs and not yours.
- A per account report showing conversations, reply times inside and outside the windows, booked calls and shows.
- A cost model that separates media spend from template spend, and that reflects per message pricing rather than the deprecated conversation model.
If you are still deciding whether to run this in house, outsource it, or buy software for it, the full cost comparison of those three routes is the piece to read before committing budget, and it is worth checking what an all in one platform really adds to the bill once usage based charges are counted.
SetScale is being built for exactly this shape of problem, several client accounts and several closers under one reporting line. The product is not open yet, and the only thing you can do today is Join the waitlist.
Verdict: none of this list is optional at ten accounts, and all of it is cheap to build before launch and expensive to retrofit after.
The numbers to instrument, and the ones not to copy
Instrument the things the windows make expensive. Ignore the benchmark you read on a blog, including this one.
| Metric | Why it matters here | Cut it by |
|---|---|---|
| Median and 90th percentile first reply time | The median flatters you, the 90th percentile is where missed windows hide | Client account, day of week, hour |
| Share of ad conversations answered inside 24 hours | Directly determines whether the free runway opened at all | Client account, campaign |
| Share of conversations that used the 72 hour runway | A runway opened and unused is spend with nothing attached | Client account |
| Template messages delivered | The billable object under per message pricing | Client account, template category |
| Booked calls and show rate | The only outcomes the client is actually paying for | Client account, closer |
There is no credible public number for what a click to WhatsApp lead should cost you, and any figure quoted without a country, a vertical, a date and a methodology is decoration. Your first four weeks of your own data beat every one of them, which is the same discipline that applies to the famous response time statistics circulating in sales content.
Verdict: measure per client account or you are measuring nothing.
Where SetScale fits
SetScale is an AI setter built for teams and agencies rather than for a single operator. The angle is the one this article keeps running into: several client accounts, several closers, and a reporting line per seat instead of one shared inbox that hides which account is failing. A white label option is part of the intended direction for agencies that resell setting under their own brand.
Being straight about the state of things: the product is not open yet. There is no dashboard to show you, no client roster to point at, and no performance claim to make. A free trial of 7 days with a card required is planned for opening. Until then the only action available is the waitlist, and the infrastructure view of the problem is the piece that explains the wider thesis, alongside the rest of the blog.
Verdict: if the ideas above match a problem you have this quarter, the waitlist is the honest next step, and nothing more is being claimed.
FAQ
Do click to WhatsApp ads cost more than ordinary lead ads? Meta's pricing documentation covers the messaging side, not your media auction. What it does document is that a reply inside 24 hours to a user arriving from a click to WhatsApp ad is free and opens a 72 hour window in which any type of message is free. The messaging cost you can control is largely a function of whether you answer in time.
What happens if nobody answers within 24 hours? The customer service window closes. From that point Meta's Cloud API guide states you can only send pre approved template messages, and the pricing page states that you are charged when a template message is delivered. The conversation is recoverable, but on different terms and with a cost attached.
Does an automated first reply count for the free window? The pricing documentation describes responding within 24 hours "using any type of message". It does not distinguish between a message sent by a human and one sent by your system. What it does require is that the message actually goes out inside the window, which is an operations question rather than a tooling question.
Can we run every client through one WhatsApp number? The Marketing API setup expects a Page with a WhatsApp number linked to it, and each client typically brings their own Page. Collapsing several clients behind one number costs you per account attribution and per account reporting, which are usually the things the client is paying you to produce.
Should we buy this channel before our response process works? No. The documented windows mean that a weak reply process converts directly into higher messaging cost and lost conversations. Build the coverage and the ownership first, then buy the clicks.
Conclusion
Click to WhatsApp ads are not a creative problem. Meta publishes the two facts that decide the outcome: a 24 hour customer service window that starts when the user contacts you and allows only pre approved templates once it closes, and a Free Entry Point window of 72 free hours that opens only if you reply inside the first 24. Everything else is staffing, ownership and reporting.
Every claim above is checkable at the source, and it should be checked before it lands in a client contract: the click to WhatsApp guide, the Cloud API messaging guide and the WhatsApp pricing page. These windows belong to Meta and the terms can move.
If you are building the team version of this, with several client accounts under one reporting line, Join the waitlist.