how i cook $100M+ meta ads offers with sonnet 5

the number one variable in your meta ads account is not the creative. it's not the targeting. it's not the budget.

it's the offer.

the offer determines your cost per booked call, your show rate, your close rate, and your cost to acquire a customer. a $50 CPL with a weak offer produces worse economics than a $120 CPL with an offer so compelling that 45% of booked calls close on the first conversation.

most operators build offers by sitting in a room and guessing what sounds good. "done-for-you meta ads management" or "growth consulting for service businesses" or "full-stack marketing solution." they pick a price point based on what competitors charge, add a guarantee they're not sure they believe in, and start running ads against it.

we've generated $10M+ for service businesses. the offer engineering process we run internally before touching a single ad account uses sonnet 5 to build offers from the ICP up, not from the operator's imagination down.

here's the full workflow.

why the offer is the multiplier (not the creative)

a strong creative with a weak offer generates curiosity and no conversions. people click, they read, they land on the offer page, and they leave because the thing being sold doesn't connect to the specific problem they're experiencing.

a mediocre creative with a strong offer generates conversions despite itself. the offer is so aligned with the ICP's pain that even an average-looking ad produces booked calls because the prospect recognizes themselves in the description of the outcome.

we've tested this head-to-head across dozens of accounts. same audience, same budget, same creative batch. different offers. the offer that named the ICP's specific gap outperformed the generic offer by 3-4x on booked call volume. not on clicks. on booked calls. the downstream metric where money happens.

the creative gets the click. the offer gets the booking.

what most operators get wrong about offer building

three mistakes that show up in nearly every service business we audit:

mistake 1: building the offer from the service out.

the operator starts with what they do (meta ads management, funnel building, creative production) and describes it as the offer. "we'll run your meta ads for you." this describes a service. it's not an offer.

an offer is a bridge between the prospect's current state and their desired outcome. it answers: what specific result will i get, in what timeframe, with what level of risk?

mistake 2: pricing based on competitors.

the operator looks at what other agencies charge and prices somewhere in the range. this produces a commodity offer competing on price against every other commodity offer in the market.

pricing should be based on the value of the outcome to the ICP, not on the cost of delivery or the competitive landscape. if the outcome is worth $500k to the client and the price is $5k, the decision is easy. if the outcome is vaguely described and the price is $5k, the decision is hard regardless of how low the price is.

mistake 3: no belief architecture.

before a prospect will book a call, they need to hold three beliefs:

• belief 1: "the problem i'm experiencing is real, and it's costing me more than i think." (problem belief)
• belief 2: "the mechanism this person uses to solve the problem is credible and different from what i've tried." (mechanism belief)
• belief 3: "this specific person/company is the right one to implement this mechanism for me." (vehicle belief)

most offers only address belief 3 (why us) and skip beliefs 1 and 2 entirely. the prospect arrives on the call without understanding why the problem is urgent or why the mechanism is different. the closer has to do all three jobs in 30 minutes. close rates collapse.

if you want a look at how we structure offer engineering for our clients before touching the ad account, DM me "META" and i'll walk you through it.

the sonnet 5 offer building workflow

this is the exact process we run internally. four phases, each building on the output of the last.

phase 1: ICP deep extraction

the prompt:

here is everything we know about this client's ideal customer profile:

[paste: client's ICP description, past customer data, testimonial 
screenshots, intake form responses, sales call recordings or notes]

extract the following in exhaustive detail:

1. the 5-7 specific pain points this ICP experiences daily.
   for each one, describe the emotional texture: what does 
   this feel like at 2am when they're worried about it?

2. the 3-4 desired outcomes they want most. rank by urgency,
   not by importance. what do they want RIGHT NOW vs what 
   they want eventually?

3. the gap between their current state and desired outcome.
   what specific things are missing or broken in their 
   current approach?

4. the failed solutions they've tried before. what did they 
   spend money on that didn't work? what's the residual 
   skepticism from those failures?

5. the buying triggers: what specific event or realization 
   would cause them to search for a solution TODAY rather 
   than "someday"?

the output is the raw material the offer gets built from. without this depth, the offer is generic. with it, the offer describes the ICP's situation so precisely that the prospect feels like you're reading their mind.

phase 2: gap analysis and offer structure generation

the prompt:

using the ICP deep extraction above, generate 5 different 
offer structures for this service business.

for each offer structure:

1. name the offer in language the ICP would use 
   (not industry jargon, not marketing speak)

2. describe the specific outcome delivered, the timeframe, 
   and the risk reversal (guarantee, performance clause, 
   or satisfaction mechanism)

3. map the offer to the ICP's top 3 pain points. for each 
   pain point, explain exactly how this offer resolves it.

4. identify which of the 3 beliefs (problem, mechanism, 
   vehicle) this offer is strongest at building, and which 
   one it's weakest at.

5. estimate conversion probability relative to the other 4 
   offers. explain your reasoning.

rank the 5 offers by estimated conversion probability from 
highest to lowest.

sonnet 5 generates the structures. the operator reviews, picks the top 2, and moves to pressure testing.

phase 3: belief architecture pressure test

the prompt:

take the top 2 offer structures from the previous output.

for each one, pressure-test against the 3 belief requirements:

BELIEF 1 (problem belief): "the problem is real and costing 
me more than i think"
- does this offer make the cost of inaction clear?
- does it surface a pain the ICP is currently ignoring?
- if not, rewrite the offer positioning to build this belief.

BELIEF 2 (mechanism belief): "the mechanism is credible and 
different from what i've tried"
- does this offer explain WHY the approach works, not just 
  WHAT it does?
- does it differentiate from the failed solutions the ICP 
  has already tried?
- if not, rewrite the mechanism description.

BELIEF 3 (vehicle belief): "this specific person/company is 
the right one to implement this"
- does this offer include specific proof (case studies, 
  metrics, timeframes)?
- does it address the residual skepticism from the ICP's 
  failed past attempts?
- if not, add the proof elements.

output the revised offers with all three beliefs addressed.

phase 4: CTA and funnel integration

the prompt:

take the winning offer from the pressure test.

generate:

1. the exact CTA line for the ad (what action the prospect 
   takes and what they get when they take it)

2. the landing page headline that bridges the ad to the offer

3. the 4-6 application questions that qualify the prospect 
   against the ICP criteria extracted in phase 1

4. the confirmation page copy that reinforces the decision 
   and increases show rate

5. the pre-call email sequence (3 emails) that builds all 
   3 beliefs before the prospect arrives on the call

output everything in the client's voice and tone.

the output from phase 4 feeds directly into the funnel build. the offer, the landing page, the application, the confirmation sequence, and the pre-call emails are all built from the same ICP data, using the same language, targeting the same beliefs.

what the workflow produces (real example)

one client came to us selling "done-for-you google ads management for ecommerce." pricing: $2,500/mo retainer.

after running the sonnet 5 workflow:

• ICP extraction revealed the primary pain wasn't "bad ads." it was "i've hired 3 agencies and none of them lasted more than 90 days because they couldn't show me what they were doing or why"
• the gap analysis showed the strongest offer structure wasn't "better ads." it was "transparent scaling system with a weekly dashboard you'll understand in 30 seconds"
• the belief pressure test identified that mechanism belief was the weakest: the ICP didn't believe any agency could actually do what they promised because of failed past experiences
• the revised offer positioned the mechanism as a documented, auditable system rather than a black-box service. the guarantee was restructured around transparency milestones, not revenue outcomes

the restructured offer increased booked call volume by 2.8x and close rate went from 18% to 31%. not because the service changed. because the offer matched the ICP's actual psychology instead of the operator's assumption about what sounded good.

the offer is the variable you're ignoring

most operators spend 90% of their optimization time on creative testing and campaign structure. those matter. they're second-order variables.

the offer is the first-order variable. get it wrong and every creative, every campaign structure, every budget allocation fights an uphill battle. get it right and average creative with a resonant offer outperforms exceptional creative with a generic one.

the sonnet 5 workflow takes 2-3 hours to run completely. the output feeds every downstream decision: funnel copy, creative angles, ad hooks, sales scripts, objection handling. one session of offer engineering touches every metric in the account.

if you want us to audit your service business and show you how we'd run this exact offer engineering workflow on your ICP and restructure the offer before touching the ad account, DM me "META" and we'll build it.

($10M+ generated for service-based businesses through meta ads funnels, fully done-for-you)

zack