Are You Actually Ready To Pour Money Into Ads?

Most first ad campaigns die before launch, in places nobody checked: an offer only friends buy, a founder with no free hour, a CRM that cannot say which lead closed, a pixel that sends Meta half the story.

Ads do not fix those. Ads find them, at about $145 a lead for B2B on Facebook (Metadata, 2025 data).

This check finds them first, for free. Every check has a pass mark, the exact fix if you fail, and how long the fix takes. At the end: a score out of 40, one of three verdicts, and a 14-day plan to get ready. For B2B, SaaS, agency, consulting and high-ticket service firms doing $30K+ a month that sell on a call. Not for pre-revenue. No closed deals means nothing to amplify yet.

Why good businesses lose their first $10K on ads

Here is the post-mortem founders describe most often. A composite, not one client, but every line of it is common.

"We tried Facebook. Spent about $8K over six weeks. Got 60 leads. Maybe four were real. Two booked. One showed. Ads don't work for B2B."

Ads work fine for B2B. That campaign was dead before launch, for three reasons that have nothing to do with creative.

Reason 1: The platform never got enough signal to learn

Meta learns from the event you tell it to chase. Its learning-phase guidance is widely reported as roughly 50 optimisation events per ad set in a 7-day window. Below that, delivery is a guess.

Now do the maths most agencies skip.

What you optimise for	Typical cost per event	50 a week costs	Per month
Qualified booked call	$200 (Jeremy Haynes's reported scale number)	$10,000	about $43,000
Lead (Facebook, B2B average)	$145 (Metadata, 153 B2B advertisers, 2025 data)	$7,250	about $31,000
Lead (Facebook, all industries)	$27.66 (WordStream 2025)	$1,383	about $6,000
Quiz completion	your number, usually well under your CPL	cheaper	cheaper

Read the top row twice. At $200 a call, optimising one ad set for booked calls costs about $43,000 a month just to exit learning. Most $30K to $100K a month firms cannot fund that, and the agency that told them to "optimise for calls" did not mention it.

The fix is a signal ladder. Optimise on a higher-volume event (a qualified lead or a completed application). Then send the deeper events (booked, held, won) back to Meta from your CRM so delivery tilts toward buyers. That is a readiness question, not an ads question. It needs a CRM with stages and tracking that talks back. Gates 6 and 7.

Reason 2: Meta's own quality floor is 200 leads a month

Meta's Conversion Leads optimisation is the tool built for exactly this: it optimises for leads that reach a sales stage in your CRM, not form fills. Meta reports an average 15% lower cost per quality lead, and 21% in a test across 1,031 advertisers (via Click Here Digital and Jon Loomer).

The entry rules, straight from Meta's developer documentation:

200+ leads a month going in.
A CRM stage that leads reach within 28 days.
That stage converting between 1% and 40% of leads.
Data uploaded back to Meta at least daily.
Built on Meta Instant Forms (lead ads).

Notice the 200. Fifty events a week is about 215 a month. Two separate rules, one number. Roughly 200 leads a month is the line where Meta stops guessing about you. Your budget has to reach it, or your plan has to say how it gets there.

Reason 3: The test was too small to be a test

"Let's start small, $1,500, see what happens" sounds prudent. It is often the most expensive option on the table. A budget that never leaves learning produces noise. The owner reads the noise as "ads don't work for us" and walks away from a channel that would have worked.

Alex Hormozi's rule in $100M Leads (via the Margin Notes summary) is blunt: budget about 2x what one customer pays you in the first 30 days, per ad test. Kill a test at 1x with zero leads. Let it run to 2x if leads come but sales do not. Expect most tests to lose; the summary puts it at about 9 in 10. His order is track first, then lose on purpose, then scale what won. Skip the middle step and you never reach the third.

The real lesson

Every one of those three reasons is visible before you spend a dollar. So is everything else in this check.

Ads are an amplifier. Plug a good business into one and it gets louder. Plug in a leaky one and you hear the leaks from the next suburb.

The 10-minute quick score

No spreadsheet. Eight questions, one per gate. Each is the knockout check for that gate: fail it and the gate is red no matter how the other four checks go.

Answer honestly. Nobody is watching. That is the point.

#	Gate	Knockout question	Pass if
1	Offer	Can you say who you help, the result, the timeframe and the mechanism in one sentence, without "and"?	Yes, and a stranger repeats it back correctly
2	Proof	How many deals in the last 90 days came from people who did not know you, your team or a referrer?	3 or more
3	Budget	Can you commit a written ad budget for 90 days that would not change a payroll decision if every dollar lost?	Yes, and it is at least the floor in Gate 3
4	Capacity	If 7 qualified calls landed tomorrow, who takes them, in which hours?	A named person, with the hours already blocked
5	Speed to lead	When a form comes in at 2:14pm on a Tuesday, when does a human first reach out?	Under 5 minutes, every time
6	CRM	Is there a written, one-page definition of a "qualified call" that sales and marketing both signed?	Yes
7	Tracking	Can you name which ad produced your last three closed deals?	Yes, from data, not memory
8	Founder calendar	Is the founder's next 30 days free of a trip, launch or hire that takes more than 3 working days?	Yes

Read your quick score

8 of 8. You are rare. Run the full check anyway to find the ambers, then read the bridge at the end.
6 or 7, and you passed 1 to 4. Ready with fixes. The misses are in the plumbing (speed, CRM, tracking, calendar). All fixable inside 14 days.
Any fail on questions 1 to 4. Not yet. Not "never". Those four are the business itself: offer, proof, money, people. Ads cannot repair them, and spending before they are fixed is how the $8K post-mortem gets written.

A worked example

Priya runs a B2B compliance consultancy at $58K a month. Deal size $9,000, first 30 days cash $4,500 (half up front). Illustrative, built from the patterns above.

#	Answer	Result
1	"We help fintechs pass SOC 2 in 90 days with a done-with-you audit sprint."	Pass
2	11 deals in 90 days. 10 referrals, 1 from a webinar.	Fail (1 stranger win)
3	$6,000 a month committed for 90 days	Fail (her floor is $18,000, see Gate 3)
4	"Me, I guess." No blocks on the calendar.	Fail
5	"Same day, usually."	Fail
6	No written definition	Fail
7	"The webinar one came from LinkedIn, I think."	Fail
8	Hiring a delivery lead this month	Fail

One out of eight. Three knockouts in the top four. Not yet.

That stings. It also kept $18,000 in her account: three months of her $6,000 budget, which would have landed on a founder with no free hours and an offer that has only ever sold on someone else's word.

Her fix list is short. A budget that matches her floor, or a cheaper test plan (Gate 3). Two weeks of cold outreach to prove the offer works on strangers (Gate 2). Two blocked call windows a day, or a part-time closer (Gate 4). Then the plumbing. She re-checks on day 30.

Your honest test number, in 60 seconds

Most founders launch without ever seeing this number. Work it out now, on the back of anything.

Floor A = 200 x your cost per lead                            (Meta's 200-leads-a-month line)
Floor B = 2 x first-30-day cash per customer x tests a month  (Hormozi's 2x-per-test rule)
Floor C = $3,000 a month, held 60 days                        (common agency minimum)

Your honest monthly test budget is the highest of the three. No cost-per-lead data yet? Plan at $60 and stress-test at $145. The $60 is a planning assumption, not a benchmark: it sits between the $27.66 all-industry Facebook average and the $145 B2B figure.

Priya: Floor A = 200 x $60 = $12,000. Floor B = 2 x $4,500 x 2 tests = $18,000. Floor C = $3,000. Her number is $18,000 a month. She committed $6,000. The $12,000 gap is the whole story. A campaign funded at a third of its floor never tells you whether ads work. It only tells you it was underfunded.

Write your number and your gap down. Gate 3 shows the three ways to close it.

The full check shows exactly how to fix every red. Five checks per gate, pass marks, fixes, time per fix, and the scorecard that turns your 40 answers into a verdict and a 14-day plan.

Gates 1 and 2: Offer and proof

These two come first because they are the only gates ads can make worse. A weak offer with no budget loses slowly. A weak offer with a budget loses quickly and leaves a pixel trained on the wrong people.

Score 1 point per pass. Knockout checks are marked (K).

Gate 1: Offer

1.1 The one-sentence offer (K)

Pass: One sentence. Who, result, timeframe, mechanism. A stranger can repeat it after one read.
Fail looks like: "We're a full-service growth partner for ambitious brands."
Fix (30 minutes): Fill the template, then read it to someone outside your industry. If they cannot say it back, cut words until they can.

We help [specific buyer at a revenue stage] get [measurable result] in [timeframe] through [named mechanism], without [the thing they hate].

Example: "We help Series A SaaS teams cut churn by 20% in 90 days with a cancellation-flow rebuild, without hiring a retention team."

1.2 A written ICP with a revenue floor

Pass: One page. Industry, size, revenue floor, the trigger that makes them buy now, three disqualifiers.
Fail looks like: "Anyone who needs marketing."
Fix (45 minutes): List your five best clients by profit, not revenue. Write what they had in common the month before they signed. That is your ICP. The revenue floor becomes a question on your form.

1.3 Price or price range decided

Pass: You can state the range on a call without checking with anyone. Your form can ask a budget question that matches it.
Fix (1 hour): Write three tiers with a price and a scope line each, for example: Audit $2,500 (two weeks, written plan), Sprint $9,000 (90 days, done with you), Retainer $4,000 a month. Kill the tier you have not sold in six months. Then decide whether the page shows a "from" price. Showing it filters tyre-kickers; hiding it books more calls of mixed quality. Pick one on purpose and make the form's budget question match it.

1.4 Top five objections, answered in writing

Pass: A list of the five reasons people do not buy, each with a one-paragraph answer and a proof point.
Fix (1 hour): Pull your last ten lost deals. Write down the real reason for each, not the polite one. Group them. Answer the top five. These answers become your ad angles, your landing page FAQ and your pre-call emails.

1.5 Risk reversal on paper

Pass: A written guarantee, pilot, or exit clause you are willing to print in an ad.
Fix (30 minutes): Pick one: a paid pilot with a clear success metric, an out clause after 30 or 60 days, or a guarantee tied to an action you control. If you would not print it, it is not a risk reversal.

Gate 2: Proof

2.1 Three wins from strangers in the last 90 days (K)

Pass: 3+ closed deals from people with no prior relationship and no referrer.
Why it matters: Referral deals close on trust borrowed from the referrer. Cold traffic arrives with zero trust. If more than 80% of your recent wins are referrals, your offer is proven on warm trust only.
Fix (14 days): The Stranger Proof Sprint. Pick 50 accounts that match your ICP and have never heard of you. Send a short, specific note to five of them each weekday for two weeks.

Subject: [their company] and [specific problem]
Hi [name], noticed [specific, verifiable observation about their business]. We fixed the same thing for [similar company type] in [timeframe]: [one number]. Worth a 20-minute look at whether it applies to you? If not, no reply needed.

If 50 strangers produce zero conversations, the offer is the problem, and you just learned it for the cost of your time instead of a quarter of ad spend.

2.2 Three proof assets with written permission

Pass: Three case studies, video clips or screenshots, each with the client's written OK to use in ads.
Fix (1 week, mostly waiting): Send the permission request today. Ask for one specific result and offer to send the wording first. That is a much smaller yes than "can we have a testimonial?"

Hi [name], we're putting together a few short examples of client results. Would you be OK with us sharing [the specific result] with your name and logo? I'll send you the exact wording first. A one-word "yes" is plenty.

2.3 At least one proof asset with a number

Pass: A before-and-after number with a timeframe. "Pipeline went from $80K to $310K in four months."
Fail looks like: "They were amazing to work with!"
Fix (1 hour): Call your happiest client. Ask: what was the number before, what is it now, how long did it take. Write it down while you are on the phone.

2.4 Proof that matches the ICP

Pass: Your best proof is from the buyer type you plan to advertise to.
Fail looks like: Every case study is retail. The ads will target SaaS.
Fix (1 hour): List your last ten wins by industry and size. If fewer than three match the segment you plan to target, pick one of two paths. Aim the first 60 days of ads only at the segment your proof already covers, or run the Stranger Proof Sprint in the new segment first. Then write one line on the ad brief and hold the team to it:

First 60 days we target: __. The proof we show them: __, __, __.

2.5 Two clients would take a reference call

Pass: Two named clients who would take a 15-minute call from a prospect this month.
Fix (1 day): Ask four clients, expecting two yeses.

Hi [name], a [company type] is close to signing with us and asked to speak to a client. Fifteen minutes, any day this month, and I'll send you their questions first. Would you be up for it?

If nobody says yes, ask one of them why. That answer is worth more than the reference.

Gates 1 and 2 draw on a principle common to Sabri Suby's King Kong teaching and Hormozi's Core Four ordering: prove the offer converts before buying traffic for it. Paraphrased, credited as inspiration.

Gate 3: Budget and the honest test number

This gate gives you a number. Write it down. It is the most useful number in this whole document.

The three floors

Your honest minimum monthly test budget is the highest of three floors.

Floor A, the learning floor   = 200 leads x your expected cost per lead
Floor B, the Hormozi floor    = 2 x 30-day customer cash x number of tests per month
Floor C, the agency floor     = about $3,000 a month, held for at least 60 days

Floor A comes from Meta: 200 leads a month for Conversion Leads, roughly 50 events a week to exit learning.
Floor B is inspired by $100M Leads: about 2x a customer's first-30-day cash per ad test.
Floor C is the commonly recommended agency minimum for a first B2B Meta test ($2,500 to $5,000 a month, about $3,000 held for 60 days). Agencies with an interest in bigger budgets wrote it. Treat it as a floor, not advice.

Your cost per lead is somewhere in this range

Benchmark	CPL	What it is
Facebook, all industries	$27.66	WordStream 2025, 1,000+ campaigns. The floor.
Google, business services	$103.54	WordStream 2025
Instagram, B2B	$138	Metadata, 2025 data
Facebook, B2B	$145	Metadata, 2025 data. The ceiling for most.
LinkedIn, B2B	$202	Metadata
Google, B2B	$524	Metadata

There is no public benchmark for quiz-funnel CPL in B2B. If you have no data of your own, plan at $60 (an assumption) and stress-test every floor at $145.

Priya, worked

Floor A $12,000, Floor B $18,000, Floor C $3,000 (the 60-second version in chapter two). Her honest test number is $18,000 a month. She had committed $6,000.

So her real choices are three. Raise the budget. Lower the floor (Floor B drops if she runs one test at a time, or charges more up front). Or accept that $6,000 buys a slower, noisier test and set expectations to match. What she must not do is spend $6,000, expect an $18,000 result, and blame Facebook.

The five checks

3.1 The test budget clears your highest floor (K)

Pass: Written monthly budget at or above the highest floor, committed for 90 days.
Fix: Raise it, reduce test count, or raise 30-day cash (3.3). If none of those is possible, you are not yet.

3.2 Losing it would not change payroll

Pass: If the entire 90-day budget returned nothing, you would not cut a salary or miss a bill.
Why it matters: Scared money pulls tests at day 9, before they have data. Every decision gets worse.
Fix: Lower the budget to the "would not hurt" line and lengthen the test window, or wait a quarter.

3.3 You know your first-30-day cash

Pass: You can say what a new client pays you in the first 30 days, to the nearest $500.
Fix (1 hour): Look at your last ten deals. Then ask whether a setup fee, a paid audit or a half-up-front term would raise it. Hormozi's idea of client-financed acquisition: when 30-day cash covers acquisition cost, ads pay for themselves inside one card cycle. A $2,000 onboarding fee can be the difference between a test you can afford and one you cannot.

3.4 You know your maximum cost per held call

Pass: You have computed it.

value per held call   = deal value x close rate x gross margin
max cost per held     = value per held call / 3

Priya (assumed 20% close rate, 60% margin): $9,000 x 20% x 60% = $1,080 of gross profit per held call. At the 3:1 ratio Hormozi uses for lifetime gross profit to acquisition cost, she can pay up to $360 per held, qualified call. That number is her kill line. Any campaign above it after the test window gets cut. Below it gets scaled.

Fix (15 minutes): Use the formula with first-year deal value. If clients renew, use lifetime value instead and the line moves up. No close rate of your own yet? Use 19%, the new-logo win rate across 655,000 CRM opportunities (Ebsta and Pavilion, 2025). It is an opportunity win rate, not a per-call rate, so treat it as a cautious floor.

3.5 Accounts and payment in your own name

Pass: Meta Business Manager and ad account owned by your business, your card on it, admin access you control. Same for Google if used.
Fail looks like: "The last agency set it up on their account."
Fix (1 day): Create your own. Ad history, pixel data and audiences belong to whoever owns the account. Rent nothing you cannot take with you.

Gate 4: Sales capacity

Five to seven booked calls a day sounds like a dream. It is a schedule. Here is what it looks like on a calendar.

The capacity maths

booked per month = booked per day x 22 working days
held per day     = booked per day x show rate
closer hours     = held per day x (call length + 20 min prep and notes)
closers needed   = held per day / 4

The "4" is the midpoint of the 3 to 6 live calls a day that high-ticket closer job ads cluster around. A job-ad number, not a study. Use it as a planning line.

Booked per day	Show rate	Held per day	Hours a day (45-min call + 20)	Closers (held / 4)
5	70%	3.5	3.8	0.9, so 1 at the edge
7	70%	4.9	5.3	1.2, so 2 once follow-ups land
7	85%	6.0	6.4	1.5, so 2

Now add follow-up calls, proposals and the second meetings that every real B2B deal needs. A founder who also runs delivery does not have 5 spare hours a day. Nobody does.

The five checks

4.1 A named closer with hours blocked (K)

Pass: A named person, with recurring call blocks on the calendar for at least the "hours a day" figure in the table at 5 booked a day.
Fix (1 day): Block the hours now, before the calls exist. If the hours are not available, decide today: hire, contract a part-time closer, or start at 3 a day and grow.

4.2 A sales process written down

Pass: A one-page call structure: opener, diagnosis questions, the offer, the close, next step. Anyone on the team could run a call from it.
Fix (2 hours): Record your next three calls. Write down what you actually do. That is version one.

4.3 A no-show owner and a reschedule path

Pass: One person whose job is to text, call and rebook no-shows within 15 minutes of the missed start.
Fix (30 minutes): Assign it. Save the message.

Hi [name], it's [your name] from [company]. We had [time] booked and I didn't see you join. No stress. Here's a link to grab a new time this week: [link]. Or reply with a time that works and I'll send the invite.

4.4 A weekly pipeline review on the calendar

Pass: 30 minutes a week, same time, reviewing booked, held, won, lost and why.
Fix (5 minutes): Book it, recurring, now.

4.5 Delivery can absorb the wins

Pass: If you closed 50% more clients next month, delivery would hold quality. Hormozi's point on this is simple: ads can grow a business that has room to deliver. Without the room, ads grow refunds.
Fix (1 hour): Do the sum. Delivery hours per client per week x (current clients x 1.5) = hours you would need. Compare it with the delivery hours your team actually has. Example: 6 hours a client, 10 clients, so 15 clients at 1.5x = 90 hours a week. A team with 70 delivery hours needs that hire before the ads, not after. Then write the trigger:

When we reach __ clients, we hire __. The search starts at ____ clients.

Gate 5: Speed to lead and follow-up

The lead is warmest the second they submit. Every minute after, someone else's ad is in their feed.

What the data says

Calls in the first minute were 391% more likely to convert than slower calls, and conversion kept rising with up to six spaced call attempts (Velocify, 2016).
Of 2,241 US firms tested, only 37% answered a web lead within an hour, and 23% never answered at all (Harvard Business Review, 2011).
No-show rates by how the call was booked: instant booking at form fill about 8%, paid social lead form about 33% (Focus Digital, modelled estimates).
Same-day bookings no-show around 7%. Bookings two weeks out: published estimates of 42% to 67% (Focus Digital).

Translation: the fastest path from "interested" to "on the calendar" wins, and a long booking window is a quiet killer.

The five checks

5.1 First human touch under 5 minutes in business hours (K)

Pass: You submitted your own form three times this week, at random times. A human reached out inside 5 minutes each time.
Fix (1 day): Route new leads to a phone notification, not an inbox. Name who answers in each hour block. Then test again.

Opener, phone: "Hi [name], it's [your name] from [company]. You just filled in our [quiz or form] about [topic], so I thought I'd call while it's fresh. Have you got two minutes? I just want to make sure the call we book is actually useful to you."

5.2 Six attempts over three days

Pass: A written cadence: call plus SMS at minute 1, call at hour 1, call plus email day 1 afternoon, call day 2, SMS day 2, final call plus "closing the loop" email day 3.
Fix (1 hour): Build it as a workflow in your CRM so it does not depend on memory.

5.3 After-hours auto-reply with a booking link

Pass: A lead at 11pm gets an SMS and email within a minute, with a calendar link and a human promise for the morning.

Hi [name], thanks for filling in [quiz name]. It's [your name] at [company]. Our team is offline right now, but you can grab a time for tomorrow here: [link]. If it's easier, reply with a good time and I'll call you first thing.

5.4 Calendar inside the flow, booking window capped at 7 days

Pass: Qualified applicants see a calendar on the next screen, not in an email. No slots more than 7 days out.
Fix (2 hours): Embed the calendar after the application. Close any slot past day 7. Add same-day and next-day slots.

5.5 Reminders and deliverability that actually land

Pass: Reminders at 24 hours, 1 hour and 1 minute before the call (the cadence Chili Piper publishes with its under-5% outbound no-show result), plus a morning-of SMS. Your sending domain has SPF, DKIM and DMARC (at least p=none), one-click unsubscribe, and a spam rate under 0.30%. Those are Gmail's bulk-sender rules for 5,000+ messages a day, with 0.10% advised. Yahoo adds unsubscribes processed within 2 days.
Fix (1 day, needs your DNS login): Ask whoever manages your domain for the three records. Check them with a free DMARC checker. Send your reminders to a Gmail and an Outlook address and see where they land.

1-hour reminder SMS: "Hi [name], we're on at [time] today. Here's the link: [link]. I've read your answers and I've got two ideas for [their stated problem] ready. See you soon. [your name]"

Write the reminder about their problem, not your company. People turn up for their problem.

Gate 6: CRM and the qualified-call definition

Here is the fight in every company that buys ads.

Marketing: "We sent you 40 leads." Sales: "They were all garbage." Both are right, because nobody wrote down what "qualified" means.

This gate ends that fight before it starts.

The Qualified Call Definition, one page

Copy this. Fill it. Have sales and marketing sign it. Pin it.

Field	Your rule
Revenue floor	Monthly revenue of at least $____
Decision-maker	Founder, CEO, or the person who signs: ____
Problem	Has stated ____ as a current problem
Timeline	Wants it solved within ____ days
Budget	Can invest at least $____ (or "from" price accepted)
Disqualifiers	Any one of: __, __, ____
Counts as a "qualified call" when	Meets all rows above and the call is held

Show rate is a qualification problem first

The biggest show-rate swing in the sources behind this check is not about reminders. In a case Hormozi has described on his podcast, a business went from a 49% to a 70% show rate, and from a 27% to a 41% close rate, mainly by fixing targeting and cancelling roughly 75% of unqualified bookings. Reminders were in the mix. Qualification was the lever.

So the rule on paper matters more than the text message. A booking that fails your definition gets a polite redirect, not a slot on the closer's calendar.

Hi [name], thanks for booking. Looking at your answers, I don't think a call is the best use of your time yet. We work best with companies at [revenue floor] and up. Here's [free resource] that covers what we'd tell you on the call. When you get there, reply to this email and I'll get you in the same week.

The five checks

6.1 Written, signed qualified-call definition (K)

Pass: The table above, filled, signed by whoever sells and whoever markets.
Fix (1 hour): Fill it in a meeting with both people. Do not leave until it is signed.

6.2 Pipeline stages that match reality

Pass: At minimum: New lead, Booked, Held, Qualified, Proposal, Won, Lost (with reason), Disqualified.
Fix (2 hours): Rebuild the pipeline with those stages. Each stage gets an owner. The "Qualified" or "Held" stage is the one you will send back to Meta (Gate 7), and Meta needs it to happen within 28 days of the lead and convert 1% to 40% of leads.

6.3 Every call dispositioned within 24 hours

Pass: Every held call has a stage and a note by the next working day. Check the last 20.
Fix: Make it a rule with a name on it. Add a daily 5-minute "close the loop" block at the end of the closer's day.

6.4 "How did you hear about us?" on every form and call

Pass: A required free-text field on the form, and the question asked on the call.
Why: Platform attribution misses podcasts, word of mouth and the LinkedIn post they saw three weeks ago. Refine Labs compared self-reported answers with software attribution across 620 conversions and found the two told very different stories. You want both.
Fix (15 minutes): Add the field. Tag answers weekly.

6.5 List cleaned before it seeds anything

Pass: Your contact list has been cleaned in the last 6 months before you upload it as a custom audience or lookalike seed.
Why: Contact data decays at roughly 2% a month, often quoted as 20% to 30% a year. Vendor figures, no reachable primary study, but the direction is not in doubt. A list untouched for two years is more than a third wrong. Seed a lookalike from it and Meta goes looking for the wrong third.
Fix (1 hour): Remove bounces, competitors, staff, students and anyone below your revenue floor. Upload only buyers and qualified prospects as the seed.

Gate 7: Tracking and signal

Nobody books a call to discuss pixels. This gate is still the one that quietly sets your cost per lead before the first ad runs.

Two Meta-reported figures explain why. Pixel plus Conversions API runs about 13% lower cost per result than pixel alone (Meta's Blueprint figure, study not public). Conversion Leads adds the 15% to 21% lower cost per quality lead from earlier. A business without server-side tracking is paying a tax on every lead. Neither number is guaranteed for you. Both point the same way.

The signal ladder

Meta learns from whatever you tell it to chase. Pick the deepest event that still clears about 50 a week per ad set.

Quiz complete. High volume. Low quality. A starting signal.
Application or lead. The usual optimisation event. Conversion Leads needs 200+ a month.
Qualified booked call. The best signal, but at $200 a call, 50 a week is $10,000.
Held call or won deal. Best of all. Rarely enough volume to optimise on.

Optimise on step 2. Send steps 3 and 4 back from the CRM. That is how a small budget borrows the signal of a much bigger one: Meta gets the volume it needs from leads and the quality it needs from your pipeline.

The five checks

7.1 You can trace a closed deal back to an ad (K)

Pass: For your last three closed deals, you can show the source, campaign and ad from data.
Fix (1 day): UTM convention on every link (utm_source, utm_medium, utm_campaign, utm_content with the ad ID), captured into hidden fields on the form and saved on the CRM contact.

7.2 Pixel and Conversions API, deduplicated

Pass: Each key event fires from both browser and server with the same event_name and a shared event_id, so Meta counts it once. Events Manager shows both sources, no deduplication warnings.
Fix (1 to 3 days): Most CRMs and funnel tools have a native CAPI connection. Turn it on, pass event_id, then use the Test Events tool to watch an event arrive from both sides.

7.3 Event Match Quality checked

Pass: You have looked at the Event Match Quality score (out of 10) on each event in Events Manager and acted on the low ones. Meta does not publish a pass mark. Practitioners chase 6 and up. Only matched events count for delivery and attribution.
Fix (2 hours): Send email, phone, external_id, fbp and fbc where you have consent. Server web events also need action_source, event_source_url and client_user_agent.

7.4 CRM stages sent back to Meta, daily

Pass: The Qualified or Held stage from Gate 6 is sent back to Meta with the lead ID at least daily. This is the Conversion Leads plumbing.
Fix: Meta estimates 3 to 4 weeks for a developer build, or under a day with Zapier or a CRM partner integration. Then 2 to 4 weeks of learning. Start this on day 1, not day 30.

7.5 One sheet of truth, plus consent

Pass: One weekly sheet: spend, leads, booked, held, qualified, won, revenue. Plus a consent banner where your traffic needs one. Google enforced consent signals for EEA users from March 2024.
Fix (30 minutes): Make the sheet. Seven columns. Fill it every Monday. This was Hormozi's whole first phase: measurement first, one to two days, zero leads, entirely worth it.

Gate 8: The founder calendar

Nobody puts this on a checklist. It decides more first campaigns than creative does.

Weeks 1 to 4 of a new campaign are when the decisions get made. Which ad to kill. Which lead was weird. Whether the offer line is landing. Whether to answer the closer's question in an hour or next Thursday. If the founder is in Lisbon, mid-launch, or interviewing for a COO, those decisions wait. Waiting decisions cost money every day.

The five checks

8.1 No 3-day absence in the next 30 days (K)

Pass: No trip, conference, launch or hiring sprint that takes the founder out for more than 3 working days in the first 30 days of ads.
Fix: Move the launch date, not the trip. A two-week delay costs nothing. A campaign with nobody steering costs a month of spend.

8.2 Two call windows a day, protected

Pass: Two recurring windows (say 10:00 to 12:00 and 14:00 to 16:00) open for booked calls, every weekday, for the first 30 days. Even if someone else closes, the founder takes the first 20 calls to hear the market.
Fix (10 minutes): Block them. Mark them busy to everyone except the booking calendar.

8.3 A 20-minute daily ads check-in, weeks 1 to 4

Pass: 20 minutes every weekday with whoever runs the ads: yesterday's leads, quality notes, one decision.
Fix (5 minutes): Book it at the same time every day and run the same agenda. Never skip three in a row.

Yesterday: spend, leads, booked, cost per lead. (5 min)

Quality: read three leads aloud. Would we take that call? (10 min)

One decision: kill, keep or change one thing. Write it down with a name on it. (5 min)

8.4 The first 20 call recordings get watched

Pass: A plan to listen to the first 20 calls (or their summaries), tagging the objection that killed each lost one.
Why: The first 20 calls are the cheapest market research you will ever buy. They tell you which ad angle, which qualifier, which page line to change.
Fix (30 minutes to set up): A sheet with five columns: call date, ad that sourced it, held or no-show, outcome, and the reason in the prospect's own words. After 20 rows, the most common reason for "no" becomes the next ad angle and the next FAQ line. The most common reason for "yes" becomes the headline.

8.5 A written "stop and fix" rule

Pass: You decided in advance what would make you pause: cost per held call above your kill line (Gate 3) for 14 days, show rate under 50% for a week, or a closer at capacity.
Fix (15 minutes): Write the three triggers and the action each one sets off. Decisions made in advance are calm. Decisions made on a Friday night after a bad week are not.

My stop and fix rules:

If cost per held, qualified call is above $__ for 14 days, we pause the worst ad set and change __.

If show rate is below __% for 7 days, we tighten the qualifier on __.

If the closer's calendar is over __% full, we pause new spend until __.

Your score, and the 14-day plan

Count your passes. One point each, 40 in total.

The scorecard

Gate	Checks passed (0 to 5)	Knockout passed?
1. Offer	___	1.1 Y / N
2. Proof	___	2.1 Y / N
3. Budget	___	3.1 Y / N
4. Capacity	___	4.1 Y / N
5. Speed to lead	___	5.1 Y / N
6. CRM	___	6.1 Y / N
7. Tracking	___	7.1 Y / N
8. Founder calendar	___	8.1 Y / N
Total	___ / 40

Gate light:

Green: 5 of 5, or 4 of 5 with the knockout passed.
Amber: 3 of 5 with the knockout passed.
Red: the knockout failed, or 2 or fewer passed.

The three verdicts

Verdict	Rule	What it means
Ready	34+ and no red gates	Launch. The ambers get fixed inside the first two weeks while the build runs.
Ready with fixes	26 to 33, and Gates 1 to 4 are not red	The business is sound. The plumbing is not. Every miss in Gates 5 to 8 can be fixed in 14 days.
Not yet	Any red on Gates 1 to 4, or under 26	Offer, proof, money or people. Fix those first. Re-check in 30 days.

"Not yet" is not a fail grade. It is a refund on money you have not spent.

Priya's full score

Gate	Passed	Knockout	Light
1. Offer	4	Y	Green
2. Proof	2	N	Red
3. Budget	3	N	Red
4. Capacity	1	N	Red
5. Speed to lead	2	N	Red
6. CRM	1	N	Red
7. Tracking	1	N	Red
8. Founder calendar	3	N	Red
Total	17 / 40		Not yet

Her plan: the Stranger Proof Sprint (Gate 2), half up front plus a $1,500 onboarding fee to lift 30-day cash (Gate 3), a part-time closer at 15 hours a week (Gate 4), and the hiring sprint moved to month two (Gate 8). The plumbing comes in weeks 3 and 4. Re-check on day 30.

The 14-day readiness plan

For "Ready with fixes" and anyone fixing ambers. Each day has an owner: You (founder), Team (whoever sells or runs ops), Tech (whoever manages the domain, CRM and site).

Day	Do	Owner	Clears
1	Write the one-sentence offer and ICP. Send three proof permission requests. Start the 50-account Stranger Sprint if Gate 2 is red.	You	1.1, 1.2, 2.2
2	Pull the last 90 days: leads, booked, held, won, deal value, 30-day cash. Compute the three budget floors and your max cost per held call.	You	3.1, 3.3, 3.4
3	Confirm ad accounts, Business Manager, pixel and payment are in your name. Request DNS access.	Tech	3.5
4	Fill and sign the Qualified Call Definition. Write the disqualify message.	You + Team	6.1
5	Rebuild pipeline stages. Add "How did you hear about us?" as a required field.	Team	6.2, 6.4
6	Block call windows, the daily ads check-in, the weekly pipeline review. Move anything that breaks 8.1.	You	4.1, 4.4, 8.1 to 8.3
7	Write the one-page call structure from three recorded calls. Assign the no-show owner.	You + Team	4.2, 4.3
8	Set up SPF, DKIM, DMARC. Check one-click unsubscribe. Send test emails to Gmail and Outlook.	Tech	5.5
9	Build the speed-to-lead workflow: phone alert, six attempts in three days, after-hours SMS.	Team	5.2, 5.3
10	Embed the calendar after the application. Cap the window at 7 days. Set 24h, 1h, 1min and morning-of reminders.	Tech	5.4, 5.5
11	UTMs and hidden fields. Pixel plus CAPI with shared `event_id`. Test Events tool check.	Tech	7.1, 7.2
12	Event Match Quality review. Start the CRM-to-Meta stage upload (Conversion Leads plumbing). Clean the contact list.	Tech	7.3, 7.4, 6.5
13	Secret-shop your own funnel three times at random hours. Time the first human touch. Fix what broke.	You	5.1
14	Make the one sheet of truth. Write your stop-and-fix rules. Re-run this check.	You	7.5, 8.5

Day 15 is launch day if you scored 34+. If not, you know exactly which rows are left.

Do this today

Ninety minutes. No agency. No spend.

Run the quick score in chapter two. Circle every knockout you failed. (10 min)
Compute your honest test number. The 60-second version in chapter two, using your real 30-day cash and a $60 CPL if you have no data. Write the highest floor, and the gap to what you planned to spend, on a sticky note. (15 min)
Submit your own form from your phone right now. Start a timer. Stop it when a human contacts you. That number is your speed to lead. (5 min, then wait)
Fill the Qualified Call Definition with whoever sells. Both sign it. (30 min)
Open your calendar and block two call windows a day for the next 30 days. If you cannot find them, Gate 4 just failed, and you found out for free. (10 min)

Or have it built

You can do all of this yourself. Every check above has the fix and the time it takes. Plenty of founders will, and should.

Or insaneX builds the whole machine, and the 14 days of plumbing come with it.

We'll Build You A B2B Booking Machine That Lands 5-7 Qualified Sales Calls On Your Calendar Every Single Day.

Meta and Google ads, a quiz funnel that qualifies before the calendar, the calendar inside the flow, CRM follow-up by email and SMS, reminders, and your qualified and held stages sent back to Meta so it chases buyers, not form fills. Live in 14 days. For B2B firms doing $30K+ a month.

$54.3M+ revenue generated for clients. 9x average ROAS. 85+ SaaS and B2B brands scaled. $7M+ ad spend tested.

Bring your score. On the call we go through your reds and ambers line by line, tell you which ones we fix inside the build and which ones are yours. If your verdict is "not yet", we will say so, and tell you what to fix first.

Book a strategy call

Sources and inspiration

Grades: A primary or publisher's own data, B reputable secondary citing a named primary, C summary, vendor or job-ad data. Directional only.

Platform rules

Meta, Conversions API for CRM (Conversion Leads): 200+ leads a month, stage within 28 days, 1% to 40% stage rate, daily upload, time to value. A. https://developers.facebook.com/docs/marketing-api/conversions-api/conversion-leads-integration
Meta, Conversions API best practices: deduplication, required fields, Event Match Quality. A. https://developers.facebook.com/docs/marketing-api/conversions-api/best-practices
Meta Business Help Center, learning phase (about 50 events a week, widely reported). Verify in a browser. https://www.facebook.com/business/help/112167992830700
Conversion Leads 15% and 21% lower cost per quality lead (Meta figures via secondaries). B. https://www.clickheredigital.com/insights/meta-shift-to-lead-optimization and https://www.jonloomer.com/?p=44212
Pixel plus CAPI 13% lower cost per result (Meta Blueprint figure). B/C. https://funnel.io/destinations/meta-conversions-api
Google, email sender guidelines. A. https://support.google.com/a/answer/81126
Yahoo, sender best practices. A. https://senders.yahooinc.com/best-practices/

Costs and budgets

WordStream 2025 Facebook benchmarks via Search Engine Land. B. https://searchengineland.com/facebook-ad-costs-jump-beat-google-461690
WordStream 2025 Google benchmarks via Search Engine Land. B/C. https://searchengineland.com/google-ads-costs-keep-rising-but-conversion-rates-improved-in-2025-451167
Metadata.io B2B advertising benchmarks. A for its own data. https://metadata.io/b2b-advertising-benchmarks
GrowthSpree, Meta budget for B2B SaaS 2026. C. https://www.growthspreeofficial.com/blogs/meta-ads-budget-b2b-saas-2026.md
Stackmatix, Meta budget for early-stage B2B SaaS. C. https://www.stackmatix.com/blog/meta-ads-budget-b2b-saas-startups
Jeremy Haynes, call funnels at scale (his reported $200 per qualified call). C. https://recapio.com/digest/how-to-generate-40-50-qualified-sales-call-daily-1m-mo-call-funnels-by-jeremy-haynes
Ebsta and Pavilion, 2025 GTM benchmarks (19% new-logo win rate, 655,000 opportunities). B. https://benchmarks.ebsta.com/hubfs/V3%202025%20Benchmark%20Report/gtm_benchmarks_digital_report.pdf

Speed, shows and sales

Velocify, speed and spaced follow-up (2016). A. https://prnewswire.com/news-releases/velocify-research-shows-time-of-day-has-minimal-impact-on-sales-effectiveness-consider-quick-and-strategic-follow-up-instead-300275320.html
Harvard Business Review, "The Short Life of Online Sales Leads" (2011). A. https://hbr.org/2011/03/the-short-life-of-online-sales-leads
Focus Digital, no-show rate by booking channel (8% instant, 33% paid social form) and lead time (about 7% same day, 42% to 67% two weeks out). C, modelled, sample undisclosed. https://focus-digital.co/average-no-show-rate-for-booked-sales-meetings-by-booking-channel/
Chili Piper, improving show rates (24h, 1h, 1min reminders; under 5% outbound no-shows; March 2020). A, vendor. https://www.chilipiper.com/blog/improve-your-show-rates
Bridge Group 2024 SaaS AE report summary (5.7-month ramp). B/C. https://daffodil-greyhound-ashp.squarespace.com/blog/5-essential-learnings-from-the-2024-saas-ae-report-bridge-group
High-ticket closer job ads (3 to 6 calls a day). C. https://remotive.com/remote/jobs/sales/high-ticket-closer-5999236

Data and attribution

Refine Labs, the attribution mirage. A for its own data. https://www.refinelabs.com/blog/attribution-mirage
Contact data decay estimates (about 2% a month, 20% to 30% a year; no reachable primary). C. https://landbase.com/blog/data-decay-rate-statistics

insaneX figures

$54.3M+ revenue generated for clients, 9x average ROAS, 85+ SaaS and B2B brands scaled, $7M+ ad spend tested: insaneX's own delivery records, marked [P] (pending). Each prints only once its How insaneX Counts ledger row (definition, period, as-of date) is filled. https://insanex.io/resources/how-we-count
Priya is an illustrative composite built from the patterns above, not a client. The opening post-mortem is a composite too.

Frameworks credited as inspiration

Alex Hormozi, $100M Leads: test budget at about 2x 30-day cash, about 9 in 10 tests lose, track then test then scale, LTGP to CAC of 3:1, client-financed acquisition. Via Margin Notes summaries. C. https://margin-notes-600.pages.dev/books/100m-leads/ch-11 and https://margin-notes-600.pages.dev/frameworks/three-phases-of-scaling-paid-ads
Alex Hormozi, The Game podcast, show-rate case (Ep 912) and fixing the business before ads (Ep 964), via Shortform summaries. C. https://www.shortform.com/podcast/episode/the-game-w-alex-hormozi-2025-06-20-episode-summary-throwback-i-doubled-a-business-in-60-days-to-show-its-not-luck-ep-912 and https://www.shortform.com/podcast/episode/the-game-w-alex-hormozi-2026-04-23-episode-summary-before-you-run-ads-fix-this-one-thing-first-ep-964
Alex Hormozi, friction and targeting on lead quality. C. https://rosetta.to/u/alexhormozi/learn-paid-ads-in-30-minutes
Sabri Suby, King Kong and Sell Like Crazy: proving the offer before scaling traffic, honest scorecard verdicts, attributing to revenue rather than leads. Principles paraphrased from course notes; no text reproduced.

Are You Actually Ready To Pour Money Into Ads?

What is inside

Why good businesses lose their first $10K on ads

Reason 1: The platform never got enough signal to learn

Reason 2: Meta's own quality floor is 200 leads a month

Reason 3: The test was too small to be a test

The real lesson

The 10-minute quick score

Read your quick score

A worked example

Your honest test number, in 60 seconds

We'll Build You A B2B Booking Machine That Lands 5-7 Qualified Sales Calls On Your Calendar Every Single Day.