← Back to Funnel HQ · VSL script → · Nurture →

The VSL call funnel: framework.

The operating spine for the free AI Search audit funnel, built from Jeremy's SOPs and written against our actual buyer. Every SOP note in the vault applies its framework to the Workshop, a $2,500 seat sold one-to-many. This funnel sells a conversation, one-to-one, and the mechanics are genuinely different. This page is the translation.

Start here

Why the webinar framework does not port over

Both funnels are ours and both work, but they run on opposite engines. Copying webinar mechanics into a call funnel is the single most common way this gets built wrong.

 Webinar funnel ($7 / $97)VSL call funnel (free audit)
What is soldA seat. The content is the product.A conversation. The plan is the product, the retainer is the outcome.
The filterPrice. Paying $7 or $97 is the qualification.The application. There is no price to filter on, so the eight questions do all the work.
ScarcityReal cohort dates. The session happens once.Senior calendar time. Real, but weak, and easy to fake by accident.
Volume shapeMany low-value yeses. Optimize for registrations.Few high-value yeses. Optimize for qualified show-ups, not leads.
Failure modeThey register and do not attend.They apply and do not book, or book and do not show. Two separate leaks.
Governing SOPsSOP-04 webinars, SOP-30 challenge funnel.SOP-03 VSL, SOP-32 setter pre-call, SOP-28 confirmation, SOP-18 sales ops.
The practical consequence: in the webinar funnel a lead is worth what they paid, so automation carries the follow-up. In this funnel a booked call can be worth a five-figure retainer, which means manual human contact is economically correct here and is not in the webinar. SOP-32 and SOP-18 both hang on that single fact.
SOP-35 · Messaging Mastery

The CPO spine, written for our buyer

Circumstances, Problems, Outcomes. The vault's version of this is written for a business owner who wants to learn SEO. Ours does not want to learn SEO, they want to stop losing ground while paying someone else to handle it. Everything in the funnel has to trace back to a line in this block.

Circumstances how they narrate it

"Search has always worked for us. Traffic has gone flat and nobody can tell me why. I keep hearing about AI search. I do not know if it is hype or if I am already behind."

They blame: the algorithm, their agency, the market getting noisier. Notice that all three are external. Do not argue with the narrative in the ad, meet it.

Problems the actual one

They have no measurement of whether the answer engines recommend them. So they cannot tell the difference between SEO working and SEO being quietly replaced underneath them.

By the time it shows up as a traffic drop in a monthly report, the citation ground in their category has already been taken by whoever was measured first. The report is a trailing indicator of a decision already made.

Outcomes

Grand: be the business the answer engines name in their category. A compounding channel with no cost per click and no auction to lose.

Milestone: know their SOURC-E score, know exactly where competitors are cited instead of them, and hold a 90-day plan that says what to fix first.

Tiny: find out today whether ChatGPT names them or names a competitor. This is the one the ad sells. It is free, it is immediate, and it is the only promise the ad is allowed to make.

Amplifiers we are cleared to use

  • Fear: a competitor is being recommended right now, in answers you cannot see.
  • Frustration: paying every month for a report you cannot act on.
  • Envy: naming the specific competitor the engines cite. Strongest one we have, and it is factual, we can pull it live.
  • Identity: the owner who saw it early instead of the one who got told about it.

Urgency is off limits in its usual form. No countdowns, no invented seat counts, no "holding your spot until Friday." The one real urgency we have is that citation position compounds, so being cited second costs more to fix later than being cited first cost to win. Say that, and nothing stronger.

The Bridge: ad promise → funnel process → sales personalization

SOP-35 is blunt that these three must be congruent or the funnel leaks at the joins. Ours:

  • Ad promises: find out whether AI search recommends you or your competitor.
  • Funnel processes: the VSL explains what the audit is and what it costs, the application qualifies, the confirmation page proves we are real.
  • Sales personalizes: on the call we open their scorecard and their citation gap. Not a deck, their data.

Test for congruence: a lead should never encounter a number, a promise, or a qualification bar in one stage that contradicts another stage. See the gap list at the bottom, we are currently failing this test on the revenue threshold.

SOP-03 · Mastering VSLs

The nine beats, and what each one owes us

Hook, Credibility, Opportunity, Offer, Objections, Qualify, CTA, Testimonials, Reiterate. The script at /call-vsl already runs this order. What follows is the job each beat has to do in this funnel, so the beat can be judged rather than admired.

1. Hook · 0:00 to 0:15

Job: say exactly what this is, in the first sentence, with no windup. Affluent buyers leave during throat-clearing.

Test: if you muted everything after 15 seconds, could a viewer state what is being offered? Ours passes.

2. Credibility · 0:15 to 0:40

Job: earn the next two minutes with numbers and names, not adjectives. Founded at 17, teams in Melbourne, London and Atlanta, QuickBooks, Weber, HelloFresh, seven-time best large SEO agency.

The Samsung-found-us-through-ChatGPT line is doing double duty as credibility and as proof the mechanism is real. That is the single strongest asset in the script. Do not let it get cut for time.

3. Opportunity · 0:40 to 1:10

Job: the industry "why now," told as a shift that is already happening to them rather than a prediction. Ten blue links became two or three recommendations.

This beat sells the category, not us. If it is doing brand work it is off-job.

4. Offer · 1:10 to 1:40

Job: state the price upfront. Ours is free, which SOP-03 treats as a price like any other, and free raises a suspicion that a dollar figure would not. Say what happens before the call, on the call, and after.

No value stacking. Naming a fake dollar value for a free audit is exactly the move SOP-03 says destroys trust with this audience.

5. Objections · 1:40 to 2:05

Job: answer "what is the catch" before they think it. Because free is the price, the catch objection is the only one that matters here, and the honest answer is that some audited businesses hire us.

Say the mechanic out loud. Prospects who cannot see how you get paid assume the worst version.

6. Qualify · 2:05 to 2:30

Job: name who it is not for, out loud, and make the application feel like a consequence of that rather than a hurdle. Turning people away is the strongest available proof that the call is worth having.

This is the beat that has to carry the revenue threshold, and it is currently the one that contradicts the landing page. See gaps.

7. CTA · 2:30 to 2:45

Job: remove uncertainty about what happens next. Sixty seconds, eight questions, we review every one, fits book straight onto the calendar. Effort, count, and outcome.

8 and 9. Proof and reiterate · 2:45 to 3:05

Job: a specific named result, then the risk reversal restated. Worst case they leave with a graded site and a plan they own.

Open: the script still carries a TODO here. SOP-03 wants real screenshots, real names, specific results. Until the first cohort completes we are reiterating value without proof in the proof slot, which is the weakest beat in the film.

Runtime rule. SOP-03: affluent audiences want 2 to 10 minutes, less affluent need 15 to 45. A $1M+ owner is firmly in the first group, so the 2:45 to 3:15 target is correct and should be defended against every request to add a section.
SOP-32 · Setter pre-call

Segment by interest level, not by database state

This is the biggest change to how the funnel is currently wired. Today the nurture segments are defined by mechanical triggers: form submitted, meeting exists, meeting missed. SOP-32's Interest Spectrum says the useful variable is how hot they are, because the same database state can hold two people who need opposite treatment.

Interest levelWho that is in our funnelCorrect treatment
No interestCold ad viewer. Landed on /vsl, watched under 25%, gone.No offer. Venus Fly Trap content only (SOP-14). Pushing the CTA at this level burns the impression.
CuriosityWatched 25 to 60% of the VSL, never opened the application.Retarget with one proof asset, not the CTA again. They have heard the offer, they do not believe the proof yet.
General interestOpened the application, abandoned partway. We know the exact field.Answer the objection the drop point implies. Dropping at turnover is a fit worry, dropping at website is an embarrassment worry. Different message.
High interestCompleted the application, no booking.They said yes to the idea and no to the calendar. Remove friction and name the objection. Highest-value recoverable segment in the funnel.
ConvictionBooked.Protect the show. Manual contact, engagement bait, selfie video. Per SOP-32 this is worth roughly 30% on show rate, which is the highest-leverage number in the whole build.

We can already measure this, which is unusual

The per-event tracking on the funnel writes a KV record for every video milestone and every application field touched. That means interest level is not a guess here, it is a query. video:* gives the VSL watch depth that separates No Interest from Curiosity, and fill:* gives the exact abandon field that separates General from High. Most funnels running this SOP are estimating the spectrum. We can read it.

Engagement bait, written for this offer

SOP-32's technique: text two specific content pieces and let them choose. The choice itself is the qualification signal, and a reply of any kind restarts the conversation.

"Quick one before we speak. Want me to send over (A) the SOURC-E scorecard we grade you against, so you know what is coming, or (B) the ChatGPT citation check we ran for another business in your category? Either is fine, just reply A or B."

Both assets are real and already exist, which is the requirement. Whoever answers B is further along than whoever answers A, and the call should open differently for each.

The selfie video

SOP-32 and SOP-18 both put personal video at the center. Sixty seconds, phone, the person actually running the call, sent by text. It is obviously not automated, which is the entire point.

Script shape: name, who I am, I am the one on your call [day], I had a look at [their site] and one thing jumped out, see you then. Do not reveal the finding. The unclosed loop is what gets them on the Zoom.

SOP-28 · Confirmation page

The due-diligence machine

A $1M+ owner who books a call with an agency will Google that agency before the call. SOP-28's argument is that the window between booking and calling is the most under-used real estate in the funnel, and that you either control what they find or let a search engine decide.

What we have today

Confirmation video, client logo wall, awards, and analytics on its own video ID. That is the main video and the trust assets, two of the five parts.

What the framework asks for next: breakout modules

Short standalone answers, one objection each, 60 seconds to 2 minutes, sitting under the main video. Written for this buyer:

  • "Is this actually a sales call?" The honest answer, given plainly, kills the biggest pre-call anxiety.
  • "Why is the audit free?" Same answer as VSL beat 5, said again by a different face.
  • "We already have an SEO agency." The most common real situation for this buyer, and it needs an answer that is not a pitch to fire them.
  • "What is SOURC-E?" The framework, in two minutes, so the call does not spend its first ten minutes on definitions.
  • "How is AI search actually different from SEO?" The category question.
  • "What happens in the thirty minutes." Removes procedural uncertainty.

Full-spectrum testimonials

SOP-28 is specific that you show home runs and average wins and honest outcomes. A page of only home runs reads as selected, and this buyer is sophisticated enough to discount it automatically. An average result presented plainly buys more trust than a spectacular one.

SOP-34 tie-in, and it is on-brand for us specifically

The SOP says to make sure AI chatbots return positive results when a prospect researches you. For us that is not just a tactic, it is a proof obligation. We sell an audit of whether the answer engines recommend a business. If a prospect asks ChatGPT about StudioHawk before the call and gets something thin, the offer refutes itself. We should run our own SOURC-E audit on studiohawk.com and be able to show the score. Passing our own test is the most defensible credibility asset available to this funnel.
SOP-13 and SOP-18

Follow-up rules this funnel has to obey

Send timing (SOP-13)

3PM in the recipient's timezone. Tuesday and Friday carry the highest open rates.

We are running US and AU leads through one sequence, so "3PM" has to be timezone-aware in HubSpot rather than a fixed clock time, or half the list gets it overnight.

Credibility-first email (SOP-13)

The first email should carry a screenshot of the ad they responded to, their own application answers played back, review links and socials.

Our current first email carries none of these. It is a good email that is doing a different job than the SOP asks of it.

Framework conflict worth a decision, not a fix

SOP-18 says contact everyone, do not pre-qualify, leads lie on forms, and warns that some of the best buyers understate themselves. Our funnel does the opposite: it auto-scores the application and sends a rejection email to anyone who scores badly. Both positions are defensible, ours protects senior calendar time and Jeremy's protects revenue. Suggested middle: keep the auto-score, but the rejection email should invite a reply from anyone who thinks we read it wrong, and that reply should reach a human. Costs nothing and recovers the understaters.
Gaps

Where the funnel is currently off-framework

Closed
Revenue threshold now agrees across the funnel. Settled at $1M+. The VSL script hook and qualify beat, the /call landing page (hero, meta description and the who-it-is-for list), the application, the nurture and all seventeen ad angles now state the same bar, which closes the SOP-35 Bridge failure. The answer bands on the application form still offer sub-million options on purpose, so we can measure who applies under the bar rather than pretending they do not.
Blocker
A2P 10DLC registration before any US SMS sends. SOP-09 requires it, and the whole nurture design leans on SMS. Sending unregistered US traffic through a 10DLC-gated carrier gets messages silently filtered rather than bounced, so this fails quietly and looks like poor engagement.
Gap
Proof slot in the VSL is still a TODO. Beat 8 is the credibility payoff and currently holds a placeholder. First completed audit call that produces a nameable result should be filmed into it.
Gap
No breakout modules on the confirmation page. Six scripts listed above. They are 60 to 120 seconds each and can be filmed in one session with the VSL.
Gap
No no-show retargeting. SOP-06's remedy for high no-show is video texts plus a $5 to $10 per day retargeting pool. We have the video text half and none of the ad half.
Gap
Engagement bait and selfie video are not in the sequence yet. Both are manual by design and both are load-bearing for the 30% show-rate claim. They need an owner and a task-creation workflow, not an automated send.
Watch
Interest-level segmentation is designed but not wired. The tracking data supports it today. The HubSpot workflows still branch on mechanical state only.