Deep Research · 5 Waves · Internal / Critical-Advisor

The $99 AI course is the wrong product.
The market underneath it is real.

67agents dispatched
5waves + audit
~300sources cited
whitespace verified
June 6, 2026run date
The Verdict — three statements
01

Drey's plan, as stated, loses money.

A $99 course sold cold on Facebook doesn't self-liquidate: education cost-per-purchase runs $52–$90 — and the audit found education purchases cost 45% MORE than the all-industry baseline, not less. audit-verified The market's converged cold entry price is $27. Real estate is the most defended vertical at every layer, and the one open wedge (Claude workshops for realtors) drew a copycat 11 days after the first mover. $99 is an upsell price wearing a front-end costume.

02

The slot Carter described is genuinely empty — and the audiences respond.

Active ads, inactive ads (the graveyard test), and audit-wave alternate methods all agree: nobody sells low-ticket vertical AI training to professionals via paid social. verified 3 ways Meanwhile SaaS operators run AI-framed Meta ads to these exact audiences for 145–325 days straight — the demand and the ad economics exist. The gap is real: 62% of workers considering AI upskilling, 4% in training.

03

The winning shape is different from the original idea.

Not a recorded course → a live challenge/workshop with delivered artifacts. Not real estate → bookkeepers (channel proven, whitespace uncontested). Not funnel-first → workshop-first in Des Moines, where the lane is verifiably open and the same hours double as agency bizdev. The national funnel becomes a capped $3–5K experiment that buys the one number nobody on earth has published. backend conv. unmeasured

Decision Surface

Three ways to play it

Switch between the paths. Each shows the economics, the odds the evidence supports, and the pre-committed kill criteria.

Not recommended as stated · viable only re-shaped + immediate

The $99 "AI for Real Estate" course on Facebook

Fails on price point, vertical saturation, format, and window — four independent strikes.

Cost per purchase
$52–90
vs. $99 price = no margin
Window remaining
4–6 mo
copycat appeared in 11 days
Incumbents
REAIS · Buffini · NAR
$179 / $599yr / own bundle
TAM trend
↓ 1.2M
NAR members by end-2026
  • The ONLY viable shape: $27 live workshop + CRM-specific templates (Follow Up Boss / KVCore), time-back hook — the Yashu model executed with real ad infrastructure.
  • Fair-housing compliance hook: demoted by audit — zero agent-level enforcement exists; use as body-copy trust layer & broker-B2B angle, never the cold hook.
  • Makes sense only as Drey-runs-it / Carter-supplies-infrastructure — the window doesn't justify Carter's own focus.
  • Kill: not launched within ~8 weeks → skip entirely. Front-end CPA >$50 sustained → kill.
★ Recommended — highest expected value

Workshop-first: "AI for [vertical]" live in Des Moines

The convergence play — product revenue AND agency pipeline from the same hours. Lane verifiably open.

Event revenue
$1.5–3K
private/association · $97–497 tickets public
P(retainer in 90d)
70–80%
vs 20–25% for info product >$5K
Iowa AI CE courses
0
approved statewide — empty lane
Delivery load
3–6 h/event
after one-time 30–48h build
  • Local competition = two Drake professors doing generalist literacy (Innovation Profs). Nobody does vertical implementation.
  • Mapped targets w/ contacts: Iowa SBDC co-host → DMAAR lunch-and-learn (2,700 members) → GDM Partnership → PHCC-IA → HBA → Rotary. Outreach week 1; first events <60 days.
  • Warm/referral close rates: 50–70% vs 20–35% cold. US Chamber/Google free "AI basics" tour is creating demand to follow behind.
  • Iowa tax quirk: live-interactive training plausibly sales-tax-exempt where recorded courses are taxable (CPA confirmation needed).
  • Generates the proof assets (testimonials, outcomes, creative) any national funnel needs anyway.
Run as a capped $3–5K experiment — buys the missing industry data

The bookkeeper national funnel ("The 5-Day Close")

Best vertical by evidence. Channel proven, whitespace uncontested — but profitability hinges on one unmeasured number.

Channel proof
Live ads
Ambitious Bookkeeper ~$37 · Mavashev $27 → 1,700 buyers
AI-skills competitors
0
at cold-traffic prices (nearest: $499+$79/mo)
Market
1.6M+
employed + solo layer · FB-native
The linchpin
10%?
buyer→backend conv. — no public data exists
  • Offer: $27–47 front-end, client-communication anchor ("the email you've been dreading? drafted in 30 seconds"), artifact-bundled, live 5-day delivery.
  • Hook register: competitive displacement + client-narrative pressure ("your clients read the same AI headlines") — NOT replacement fear (only 9% feel it).
  • Partnerships in parallel: Ambitious Bookkeeper · Workflow Queen · Successful Bookkeeper podcasts (100–250 of first 500 buyers over 3–6 mo). Avoid the Bookkeepers.com/Robinson empire — competing $499 AI course.
  • Kill: front-end CPA >$50 sustained · OTO take <7% · buyer→backend <5% at day 60.
  • Scale trigger: buyer→backend ≥10% measured → push annual billing (92% vs 68% retention) and scale spend.
Stress-Test It Yourself

Funnel economics simulator

Per 1,000 cold ad clicks, using the audited benchmark ranges. The amber slider is the linchpin — no published data exists for it anywhere; it can only be measured with live traffic. Watch the verdict flip as you move it.

Per 1,000 clicks

Ad spend
Buyers
Cost per buyer (CPA)
AOV (FE + bump + OTO)
Front-end revenue (post-refund)
Front-end P&L
Backend subscribers
Backend LTV (lifetime)
Total P&L incl. backend
Blended LTV per buyer
Where to Aim

Vertical scorecard

Ranked by the full run's evidence. Expand for the load-bearing facts.

BookkeepersBest betChannel proven by live incumbent ads · AI-skills slot empty

Channel proof: Ambitious Bookkeeper runs Meta ads to working bookkeepers at ~$37 right now; Yuri Mavashev's $27 advisory blueprint ran Oct–Dec 2025 with 1,700+ claimed buyers — exact price, exact audience. Whitespace: zero AI-skills-for-existing-bookkeepers offers at cold prices (nearest is $499 + $79/mo). Anchor workflow: client-communication drafting — most-cited pain, perfect 30-second ad demo. Hook: competitive displacement + client-narrative pressure ("our clients see all the same headlines" — Jason Staats). Only 9% fear replacement personally; don't sell fear. Partnerships mapped: Ambitious Bookkeeper, Workflow Queen, Successful Bookkeeper. Conflict: the Bookkeepers.com empire (owns I Love Bookkeeping, BKX, DBA + a competing AI course).

Real estate4–6 mo windowWedge open, copycat in 11 days, incumbents everywhere else

First mover (Realtor Marketing Lab: $27 Claude workshop + CRM templates → $2,997 backend) launched May 30; a near-identical $297 copycat launched June 10. ≥10 entities active; NAR sells its own AI bundle; Realtor.com + Google announced a joint tool June 2. Credential layer owned (REAIS $179, Buffini $599/yr — same parent company as McKissock). 82% adoption / 17% impact = the shallow-adopter gap is the message. 2023 "amazement" copy is stale — competitive-pressure framing is current. Fair-housing hook: audit-demoted (zero agent-level enforcement; 5th of 5 concerns) — trust layer and broker-B2B angle only.

Trades / home servicesFormat mismatchHuge gap, but they buy coaching & workshops, not courses

DEWALT (n=2,481): 90% say AI will be indispensable in 5 years, 8% use it, 3% cite cost — a pure education gap. Anchor workflow: quoting/estimating (5hr → 1hr documented cases). But the purchase category is coaching/community ($397–$2,500/mo: Contractor Fight, EGIA, Nexstar) and free YouTube — no evidence any $27–99 self-paced product has sold to trades in volume. One organic Skool community exists (ContractorOS, $69/mo, 326 members). Enter via local workshops + associations (PHCC-IA, HBA) — which is exactly Path B. Seasonal windows: Oct–Nov, Jan–Feb.

Churches / ministrySmall ceilingFlank verifiably open, but $160–640K ARR realistic max

The incumbent (Kenny Jahng, $27/mo, 8,100-member community) runs zero paid ads — verified — and doesn't cover admin/operations. But the real market is 38–76K early adopters (Lifeway: only 10% of pastors are regular AI users), 46% budget <$100/yr, and church staff are natively untargetable on Meta (religious targeting removed; email-match only). Sermon-AI is theologically contested; admin/comms is clean. Beachhead only.

Office managers (cross-vertical)UnexploredAudit-wave surfaced — one curriculum across SMB types

The what-did-we-miss sweep flagged admin/office staff across all SMB verticals as a single audience with universal curriculum (the buyer persona behind much of the VOC). ASAP (admin professionals' association) as a warm partner. Thin evidence — noted as an option, not a finding.

The Numbers That Decide It

Evidence highlights

$72.40
Average education cost-per-purchase on Meta — 45% above the all-industry baseline. The audit's biggest correction: education leads are cheap, education buyers are not.
325 days
Uplinq's continuous AI-framed ad run to bookkeeper-adjacent audiences (ChurchSocial 220d, RealEstateContent 145d) — proof these audiences sustain ad economics.
18.5%
Spuckler's published challenge-buyer → $67/mo membership conversion (warm list, audit-verified). Cold-discounted to a 10% planning base — the linchpin nobody has measured.
64% vs 48%
Cohort vs paid self-paced completion (Ruzuku, 32K courses). Real but de-dramatized by audit — the "5% completion" doom stat is free MOOCs only.
5/5
Regulatory claims audit-confirmed: BizOpp rule dead, FTC targets only $10K+ income schemes, Meta permits skill training, zero HUD agent enforcement, typical-results disclosure required.
$3–5K
Honest test budget for one offer (the "$500 test" is a myth at education CPLs). Treat it as buying the missing industry data.
2 chargebacks
…per month at 200 sales approaches Stripe's review threshold under Visa VAMP (0.5%). Payment processing — not the FTC — is the #1 operational risk.
75%
…of course creators earn <$1K/yr. Power-law outcomes are the base rate. The infrastructure Carter already owns (media buying + creative pipeline + GHL) is the main counter-argument.
70–80%
Probability that 8–12h/wk of agency bizdev lands a retainer in 90 days (~$13K LTV) vs 20–25% for >$5K from a cold info product. The workshop is where both paths converge.
Creative Ammunition

Hook bank — from ads that are demonstrably working

Verbatim copy pulled from long-running advertisers (run length = working economics). Filter by archetype. Full 25-hook catalog: wave-3/findings/agent-10.md.

Imagine transforming your followers into guests with just one hour of work each week.
ChurchSocial.ai · 220-day run · #1 archetype for professional verticals: time-back
Everyone's talking about using AI to make money online... But 99% are using it completely wrong. While everyone else is asking ChatGPT to write their Instagram captions... smart people are using newly released AI technology for something way more profitable.
Richard Yu · 10 identical ads, 30+ days · most transferable frame — swap the mechanism slot per vertical. (Audit note: emulate the frame, never the operator — fraud record.)
Stop paying humans to make human errors.
Uplinq (bookkeeping AI) · 325-day run · status-quo-absurdity variant of time-back
Visual media is being rebuilt around AI. Brands are no longer willing to overpay agencies and wait weeks for a single project.
Histairy Films · 150+ days · competitive displacement with a closing window
For Six-Figure Business Owners Only: Build 3 AI Employees in 3 Days.
Chelsey & Stephen Diaz · exclusivity qualifier + named mechanism + compressed timeframe
78% of potential clients now research agents on social media before they ever reach out — it's not years of experience that win them over, it's the trust they feel through your content.
RealEstateContent.ai · 145-day run · new-rules-of-the-game stat
Florida man sold his house in just 5 days after letting ChatGPT handle the entire process instead of a real estate agent.
Smart Agents (organic→paid) · skeptic→amazed third-person story; polarizing = reach
Somewhere between the glue gun and the shipping labels, the fun of running a craft business can start to feel a little heavy... I found a way to make that part faster and so much lighter. — $27, made for crafters.
Julie's Wreath Boutique · 52-day single creative · emotional-relief + price-in-headline — the template for "made for bookkeepers"
❌ No need to show your face. ❌ No need to create products from scratch. ❌ No "build a personal brand" nonsense.
Richard Yu · negative requirement stripping — removes objections before they form
A.I. is here to stay, and your company's growth in 2026 depends on your next move.
aiforcompanies.com · 46-day run · soft-consequence variant safe for mainstream business audiences
Some of our older guys have learned to ask ChatGPT the right questions, and they're kind of amazed with some of the answers it comes up with.
Dan Callies, Oak Creek Plumbing (VOC, CNN) · the trades buyer's own voice — owner-pride register for trades creative
The biggest thing accountants are facing right now is the narrative — our clients see all the same headlines that we see every day.
Jason Staats (VOC, Karbon) · the bookkeeper hook nobody is running: client-narrative pressure
If You Do It

The 90-day plan

Workshop-first, funnel-second. Checkboxes are yours to use.

Phase 1 · Weeks 1–4 — local circuit + build

Goal: 3+ speaking slots booked, workshop curriculum built once, used everywhere

Phase 2 · Weeks 4–8 — first events + funnel build

Goal: 2 paid local events delivered; bookkeeper funnel staged

Phase 3 · Weeks 8–13 — the capped experiment

Goal: spend $3–5K, measure the linchpin, decide on data
What the Research Could Not Resolve

Known unknowns

Unknown

Buyer→backend conversion from cold traffic. Zero public data. The model's profitability pivots on it. Only resolvable with $3–5K of live spend and 60–90 days.

Unknown

Cold-traffic challenge economics (paid-entry CPL, free-vs-paid head-to-head). Proprietary industry-wide.

Thin

Special-ad-category classification for education-to-professionals ads: policy-implied exempt, enforcement-untested. $100 live test before scaling.

Thin

Pre-2025 graveyard blind spot: commercial ads age out of the library in ~1 year; a quiet 2023 failure cohort can't be fully excluded (no brand artifacts found, though).

Thin

Own-client-vertical risk: every documented agency-owner course success taught peers, not their client vertical. Vendor-vs-trainer positioning blur is unstudied in either direction.

Gap

Reddit comment-level VOC stayed locked behind auth walls across all 5 waves — the richest raw practitioner voice is unmined. A logged-in manual session would fill it.

How This Was Built

Wave log

Wave 1 · 24 agents
Wide net: economics, compliance, 14+ verticals, formats, funnels, channels, history. Found: $99 doesn't self-liquidate; whitespace hypothesis; RE saturated.
Wave 2 · 17 agents
Focus: live Ad Library scrape confirmed the empty slot; bookkeepers elevated; churches downgraded; economics corrected; compliance playbook.
Wave 3 · 12 agents
Drill: graveyard test → whitespace; operator teardowns (warm-audience dependence, challenge-as-setter); fair-housing demoted; Iowa recon; hook bank; VOC.
Wave 4 · 8 agents
Final pass: workshop-first backend; targeting recipes; RE window 4–6mo; partnerships; Stripe/tax fine print; opportunity-cost comparison; naming.
Wave 5 · 6 agents
Audit: 5 claim clusters verified, 4 partial, 3 contradicted (education CPP, completion doom-stats, Yu purged). Miss-sweep surfaced licensing model + framing bias.