New Offer Meta Ads Teardown: How a Cold Launch Turned 129 Calls Into 24 Clients

A brand-new agency offer with no warm list and no track record on paid traffic turned 129 booked sales calls into 24 clients on cold Meta traffic. Of those 129, 97 showed up, 71 were qualified and 24 bought at a $7,000 average deal: $168,000 in revenue on $35,076 in ad spend.

This is the fourth teardown in the series, after the B2B SaaS, ecommerce agency and PR firm teardowns. The client stays anonymous, and the screenshots are in the $168K case study. It's the only one of my case studies with every stage tracked, so this teardown is mostly about where the 129 calls went.

Can a brand-new offer work on cold Meta traffic?

Yes, if the offer and the funnel are built for strangers. This one launched at the start of 2026 with no warm list, no track record on paid traffic and no proof that cold prospects would pay, and it still booked 129 calls at $271.91 each and closed 33.8% of the qualified ones.

The difference between a referral and a cold prospect is time. Someone who found you through your content may have followed you for three or four months before booking, so they arrive knowing what you sell. Inbound leads wait until they're ready to buy from you. Outbound leads need help to get ready. A stranger who clicked your ad trusts you less than they ever will again, so the funnel has to do in 48 to 72 hours what an organic audience does over a 90 to 120 day sales cycle.

One caveat: the case study doesn't record how many clients the offer had signed by hand before launch, and I wouldn't advertise an offer that has never sold. That test is in how to test a new high-ticket offer before building the funnel.

What did the funnel look like for an offer nobody had heard of?

Four pieces between the ad and the closer: Meta ads written to one avatar at a time, a VSL landing page with an application form, a booking page and confirmation page, and a pre-call sequence with a setter call within 5 minutes of booking. Each one did a job that a reputation would normally do.

Writing to one avatar lets a stranger recognize their own situation and lets the wrong person see it isn't for them. The page answers what's in it for me, why you and how you do it above the fold, and the VSL covers the prospect's problems before the service, because a cold buyer first wants to know whether you can identify their problem.

Here's how the calls moved through it, against the floors I use on every call funnel:

Stage	This account, from January 1, 2026	My floor	What I read from it
Booked calls	129 at $271.91 each	Set by what a call is worth	Cheap against about $1,300 per booked call
Showed up	97 of 129, 75.2%	80% of booked calls	Just under floor, strong for a cold start
Qualified	71 of 97, 73.2%	80% of shows	The stage that needs work
Closed	24 of 71, 33.8%	30% of qualified calls	Above floor: prospects arrived pre-sold
Revenue	$168,000 at a $7,000 average deal, 4.79x	Ad spend follows closer capacity	About $1,460 in ad spend per client

In the other three teardowns, call quality and close rate were missing. Here every row is filled in, so you don't have to guess where the leak is. "The leads are bad" is not a number.

Why is $272 per booked call cheap for a new offer?

Because each booked call was worth about $1,300 in revenue. The account paid $271.91 to book a call, about $362 per call where the prospect showed up and about $1,460 in ad spend per closed client, against a $7,000 average deal. Each client covered their own acquisition cost the day they signed.

I've called $272 an ugly number before, and on a dashboard it is. But a cheap lead who never books, never shows and never buys is wasted ad spend and wasted closer capacity, and it teaches Meta to find you more people just like them. Cost per lead is a number you can win while going out of business.

On a new offer you don't have your own benchmarks yet, and someone else's number was built by their close rate and deal size. So work it backwards: show rate times call quality times close rate times average deal is what a booked call is worth to you. The method is in how to tell whether your cost per booked call is actually bad.

The price point is also what makes a cold launch survivable: at a $5,000 deal size, one close pays for a month of clicks, and at $500 none of this arithmetic survives.

Why did 75.2% of cold prospects show up for an offer they'd never heard of?

Because the gap between booking and the call was filled. A setter called within 5 minutes of each booking, email and SMS reminders kept the call in front of the prospect, and retargeting meant they kept seeing the business. With no reputation behind the offer, that sequence is where the trust got built.

The sale is made between the booking and the call. A cold prospect arrives with whatever your funnel gave them in the meantime, and if that was only a calendar invite, they forget why they booked. The case study records the setter call, the reminders and the retargeting, not the counts. The version I typically run over the 72 hours before the call:

A confirmation page with 5 to 20 breakout videos of 1 to 3 minutes each, covering the common questions and what happens on the call.
6 to 8 emails, plus SMS reminders.
A selfie video from the closer that references what the prospect wrote in their application. On its own it moves show rate by double digits.
Retargeting at a 15 to 20 frequency across short-form and long-form content.
A confirmation message an hour or two before the call, with the meeting link in it.

Across past client accounts, that build has increased show rates by up to 34% and lifted close rates from 7% to 15%, and it costs almost nothing in ad spend.

75.2% is just under my 80% floor. People will tell you 80% isn't possible on cold traffic. That's fluff: the show rate is bought by what you install between the booking and the call. To push this account toward 80%, I'd check the closer's selfie video and the day-of confirmation first, because they're the cheapest to add. The full build is in how to fix a 50% show rate in two weeks.

Why were only 73.2% of the calls qualified?

Because the application let through some people who couldn't buy. Of the 97 prospects who showed, 71 were qualified, under my 80% floor, so the fix sits before the calendar: tighter application questions and ads that say out loud who the offer isn't for. Not a better closer, and not new targeting settings.

An unqualified call takes the same closer slot as a qualified one and produces nothing. At my benchmark of a 30% close rate and a $7,000 average deal, a qualified call is worth about $2,100, so a slot spent on someone who can't buy is expensive even when the booking was cheap.

On a new offer this is the stage I'd expect to need work, because your first message is a guess about who the buyer is, and the calls tell you whether the guess was right. My application rules:

Three to six questions, because every extra question costs you volume and buys you quality.
Four that are non-negotiable: current state, what they've already tried, budget reality and timeline.
Never budget first, and never name and email first.
Route the wrong answers to a "sorry, we can't work with you" page, and don't fire an event to Meta for them.

The ad is the other half. If nobody is disqualified by your ad, your closers are doing the disqualifying at your most expensive hourly rate, so say who the offer isn't for and mean it. The full question list is in the application questions that filter without killing volume.

For people who want it but can't afford it, have more than one way to buy: a done-with-you version recovers some of the ad spend you paid to get them on a call. The case study doesn't record whether this offer had one, so that's what I'd add, not what happened.

What would I watch next in this account?

Three things before raising spend: call quality after the application changes, the event the account reports back to Meta, and closer capacity, which I plan at roughly 80 calls per closer per cycle. Each one decides whether more budget buys more clients or just more calls for the same closer to triage.

Call quality, weekly. If a tighter application moves it toward 80% while booked calls hold, the same spend produces more closes without touching the ads. Read it weekly, never daily, because daily numbers on a cold funnel are noise. And until it's back near floor, the budget stays where it is: I never scale over a leak I can see.

The optimization event. Once your CRM checks every application against your qualification bar, send the qualified booked call back to Meta through the Conversions API instead of the form fill, so Meta looks for people who book calls rather than people who tap buttons. You need roughly 15 to 30 of those events a week; below that, optimize one step earlier.

Closer capacity. Ad spend follows closer capacity. Past roughly 80 calls per closer per cycle, prospects wait days for a slot and the show rate falls for reasons that have nothing to do with the ads. The math is in how many sales calls one closer can handle.

What can you copy from this teardown if you're launching a new offer?

Sell the offer by hand before you buy traffic, then write each ad to one avatar, put a VSL page and an application in front of the calendar, fill the days before every call, track all four stages from booked to closed, and fix the first stage under its floor before you raise spend.

Get 3 to 5 documented client results by hand first, because cold traffic needs proof.
Write the offer and the ads to one avatar, and say who it isn't for.
Put a VSL page and an application before the calendar. Application-first books about 20% of leads against roughly 60% for a bare calendar, but the people who get through show up at around 80% and close at about a third.
Fill the 72 hours before each call, starting with a setter call within 3 to 5 minutes of booking.
Track every stage weekly against a floor, so you know which one to fix.
Give a new account at least $1,400 over 14 days before you judge it, on cost per application and cost per booked call, never cost per lead.

If you want me and my team to build this for your offer, apply here and on the call I'll give you my honest opinion on whether we can help. If you'd rather build it yourself with guidance, the coaching waitlist is open. The guide below has the full scorecard with worksheets.

FAQ

Can you launch a brand-new offer with Meta ads?

Yes, if the offer is built for cold traffic and the funnel does the educating before the call. This brand-new agency offer had no warm list and no track record on paid traffic, and it made $168,000 on $35,076 in ad spend. I still wouldn't advertise an offer that has never sold to anyone, so sign a few clients by hand first.

How much does it cost to get a client with a new offer on Meta ads?

In this account, about $1,460 in ad spend per closed client: $35,076 across 24 deals at a $7,000 average. Each booked call cost $271.91, and each call where the prospect showed up cost about $362.

What show rate should you expect from cold traffic?

My floor is 80% of booked calls, and this account hit 75.2% on a completely cold start. A low show rate usually points at what happens between the booking and the call, not at the leads.

Should you fix the ads or the sales process first?

Neither, until you know which stage is under its floor. In this account the close rate was above floor while call quality was under it, so the next fix was the application and the ad message, not the closer.