AI Video Services — Go-to-Market Plan
live
Set your name
Strategic Plan · v2 — Triangulated

AI video services for local SMBs: a conditional go.

Unit economics work. Market is real. Cold outreach takes longer than the hype suggests. Triangulated against three independent analyses — Perplexity, Claude, ChatGPT — with the realistic case haircut 20–30% from the original Claude projection.

Path to break-even
Month 3–4 (realistic)
Kill criterion
$7.5K MRR by Day 90
Y1 revenue (mid)
$230K–$290K
Team
1 FT + 1 PT, bootstrap
01 · Bottom Line

The conditional go

This is not an "agency launch" go. It is a 90-day paid validation go with one full-time operator, one part-time operator, and a written kill switch. If both operators need reliable income immediately, this is a no-go.

The unit economics work (85–90% gross margin per retainer, ~$15–30K LTV against fully-loaded CAC of $700–$2,500), the market is real (SMB GenAI adoption jumped to 55–57% in 2025), and a $300–$500/month tool stack can credibly support 5–7 clients before subcontracting. But the cold-outreach math demands 2,000–4,000 emails per signed client, cold email infrastructure needs 6–8 weeks of warmup, and even with conversations the realistic ramp loses deals to skepticism, timing, "send me info," partner approval, and budget hesitation.

The verticals named in the Perplexity research are the wrong three to lead with in 2026. Pivot to home services contractors and med spas first (test both in week 2 before committing the mix), add cosmetic dental/Invisalign later. Lead outbound with phone + LinkedIn + Loom (no warmup) while email warms. Sell a $1,497 one-time pilot as the foot-in-door, not a discounted retainer. Hard 90-day kill criterion: $7.5K MRR + 40%+ pilot-to-retainer conversion.

Gross margin
85–90%
per retainer
Avg LTV
$15–30K
conservative tenure
Fully-loaded CAC
$700–$2.5K
incl. founder time
Tool stack
$272/mo
production
Capacity cap
5–7 clients
before subcontract
Moat window
18–24 mo
before commoditization
02 · Decision Framework

Four hard gates determine continuation

Both operators should not commit full-time on day one. One operator goes full-time on sales + production. The second works 15–20 hrs/week on production and demos, converting to full-time only at the $10–$12.5K MRR trigger with qualified pipeline. This protects against the most likely failure mode — three months of burn with no revenue — without sacrificing upside.

Pre-test · Gate 0
Day 14
500–800 owner touches, ≥1 paid pilot sold
Phone-led, one vertical. If you get conversations but no one pays, the offer or price is wrong — fix before warmup investment.
Gate 1
Day 30
8–10 discovery calls, 2 paid pilots OR 1 retainer + 1 pilot
<5 completed calls after serious outreach → switch vertical or message.
Gate 2
Day 60
≥2 retainers, 4 total pilots sold, <7d time-to-first-video
If pilots don't convert or production is chaotic, pause sales and fix the offer.
Gate 3 · Kill switch
Day 90
$7.5K MRR · 3 retainers · 40%+ pilot conv.
Full go for both operators only at $10–$12.5K MRR with qualified pipeline. 2 usable case-study candidates.

Abandon criteria · any one triggers shutdown:

  • Day 90 under $5K MRR
  • Pilot-to-retainer conversion under 25%
  • Every sale requires heavy discounting
  • Production already breaks below 5 clients

Weekly leading indicators

MetricTargetTrip wire
Positive-reply rate (email, post-warmup)≥0.5%<0.25% × 2 wks
Discovery → pilot conversion≥25%<12% × 2 wks
Pilot → retainer conversion≥40%<25% × 2 wks
Time-to-first-video for new pilots≤7 days>14 days
03 · Market Reality

The opening is real — but narrower than the hype

SMB AI usage jumped from 39% to 55% in a single year (Thryv 2025). Amazon Ads found 43% of SMB marketing leaders feel "excited but don't know where to start" and 30% admit they're "faking it." That confusion gap is the wedge — most SMBs have bought a Canva or HeyGen subscription but lack the time, taste, and strategy to operate it. They want done-for-you, not another tool.

SMB GenAI adoption is accelerating — confusion is the wedge
% of US SMBs reporting active GenAI use · Source: Thryv, US Chamber, Amazon Ads (2023–2025)
75% 50% 25% 0% 2023 2024 2025 2026e 23% 39% 55% ~65%

The price corridor that's left

Done-for-you video subscriptions have already commoditized the $500–$1,500/month tier. Vidchops sells 4 videos/month for $495; Vidpros offers a fractional editor for $1,000/month; Feedbird starts at $99. Below $1,500/month a two-person US shop cannot win on price against offshore-staffed productized services. Above $5,000/month, established agencies dominate via inbound and referral.

The viable price corridor for a 2-person US-based AI-native shop
Monthly retainer pricing landscape, 2026
$0 $1.5K $3K $5K $10K+ Offshore productized Vidchops · Feedbird Commoditized YOUR ZONE $1.8K–$3.5K retainer Strategy + vertical specialism Full agencies LYFE · WebFX · Scorpion Inbound / referral lock

Defensible service layer left: strategy, vertical compliance, real-footage acquisition, and accountability — not generation. Do not sell "AI videos." Sell "patient consults" to med spas, "service calls" to plumbers, "Invisalign starts" to dentists. The done-for-you-with-humans moat for non-tech-savvy SMBs has roughly 18–24 months left. Plenty of time to extract $250K–$500K of LTV across 12–20 relationships.

04 · Verticals

The original three are wrong. Test before committing the mix.

The Perplexity research recommended real estate, independent restaurants, and boutique fitness. Each is a trap in 2026. Real estate is in the wrong moment in its commission-compression cycle. Restaurants have catastrophic churn (~50% close by year five) and the "struggling owners need marketing" argument cuts the wrong way — struggling owners cut retainers first. Boutique fitness is cash-strapped post-COVID.

Triangulation finding: Don't lock med spas at 60% of outreach without a real test. Med spas have budget ($98K/yr/location average) but are already heavily targeted by niche agencies. Home services may be the better first-validation vertical — owners answer phones, ROI is clearer, compliance is lighter. Run the Day 14 pre-test against both before committing the mix.

Test these three · committed mix decided at Day 14

▲ TEST FIRST · home services
Residential contractors
US businesses (>$500K)700K+
Mktg software spend$1.5–3K/mo
Owner answers phoneYes
Gatekeeper riskLow
ROI clarityHigh
▲ TEST FIRST · med spas
Med spas / aesthetics
US locations10,488
Avg revenue$1.4M
Marketing spend$98K/yr
Single-location81%
Niche agency saturationHigh
▲ ADD AT GATE 1
Cosmetic dental / Invisalign
Target practices~30,000
Cosmetic mktg/mo$4,200
Premium concierge$9,600/mo
Case value$4–5K
Achievable ARPU$2.5–3.5K

What to avoid (and why)

✗ AVOID
Individual realtors
Earning <$100K87%
Commission YoY↓ 4.5%
Outreach saturationFloor
✗ AVOID
Independent restaurants
5-yr closure~50%
Net margin3–9%
Annual churn risk30–40%
✗ AVOID
Boutique fitness
Studio count~40K
Owner cash positionTight
ClassPass compressionActive
05 · Cold Outreach Math

The brutal arithmetic

The single most important number: realistic cold email reply rates fell from ~5.1% in 2024 to ~3.43% in 2026 (Instantly, across billions of emails). Top-quartile operators with tight ICP and timeline hooks hit 10–15%, but planning above 5% is wishful. And a 3% reply rate does not mean enough qualified owner conversations at $2,497/mo — early-stage, no-case-study SMB sales lose deals to skepticism, timing, "send me info," partner approval, and budget hesitation. Plan for slippage.

Email-to-client conversion funnel · with realistic early-stage slippage
~1 client per 2,000–4,000 emails sent (no case study), tightening to ~1 per 1,200 once proven
Sent
3,000
Reply
90 (3%)
Positive
36 (40%)
Booked
11 (30%)
Showed
7 (65%)
Closed
1–2 (15–20%)

Fully-loaded CAC is the real number, not hard cash. "$130 hard CAC" ignores list-building, failed samples, follow-up time, no-shows, onboarding, revisions, and the fact that founder time is the scarce asset. Model early CAC as $700–$2,500 fully loaded per retained client, even though hard cash is only a few hundred dollars. This is the number that decides whether you can scale or whether scaling burns you out.

The warmup is the cruel part. Google/Yahoo bulk-sender rules (Feb 2024) and Microsoft enforcement (May 2025) require SPF, DKIM, DMARC, one-click unsubscribe, and complaint rates kept below 0.1%, never reaching 0.3%. New domains require 6–8 weeks of warmup in 2026 before you can ramp to 20–30 sends/inbox/day without burning deliverability.

The fix: multichannel, phone leads

Cold calling to SMB owners produces 15–25% connect rates (much higher than C-suite 4–6%) because owners answer their own phones. Dial-to-meeting: 2.3% average, 4–6% top performers — roughly 33–50 dials per booked meeting. Phone + LinkedIn + Loom can produce 4–8 booked meetings/week from week one, while email warms in parallel and comes online around week 6.

The Loom angle is meta-perfect for a video services business. Practitioner data shows +19% reply lift from Loom inclusion; "Trojan Horse" approach (ask permission first, record only on yes-replies) generated 607 positive replies across one 13-month campaign. Leading with a Loom proves you can produce relevant short-form content — the medium is the demo.

Outreach stack · $370–$420/month

ToolPurpose$/mo
3 sending domains + 9 Google Workspace inboxesCold email infrastructure$27
Instantly GrowthSending + warmup$37
Apollo Basic + OutscraperLead enrichment + Google Maps scraping$79
MillionVerifierEmail validation$30
OpenPhone (2 local-presence numbers)Phone outreach$33
LinkedIn Sales Navigator CoreTargeting$99
HeyReach (both seats)LinkedIn automation$79
Loom BusinessVideo outreach$15
06 · Pricing & Packaging

Three-tier ladder anchored at $2,497

Perplexity's $300–$500 starter is a trap — low-status positioning attracts bad-fit clients. No free production work except possibly a 30-second Loom teardown. Pilot is paid; revisions are capped; no performance guarantee.

SPARK PILOT
$1,497 one-time
14-day delivery · 50% credit to retainer · No performance guarantee
  • 6 short-form videos
  • 1-hour strategy call
  • 1 revision round only
  • Foot-in-door, risk reversal
SCALE
$4,997/mo
6-mo min · $1,497 setup
  • 24 videos · mixed format
  • Weekly production session
  • Dedicated strategist
  • Quarterly creative refresh

Three cash-flow disciplines · non-negotiable

100% upfront monthly
ACH
0.8% capped @ $5 vs 2.9% + $0.30 card. Never Net 30 under $5K MRR.
Setup + first month
$3,494
Collected on signing. 3 signups ≈ $10K cash.
Discount the price
Never
Concede on video count, turnaround, strategist hours instead.
07 · Sequenced Plan

14 · 30 · 60 · 90, then the 12-month fork

Days 1–14 · Pre-test & offer validation Form LLC · 500–800 owner touches phone-led · one vertical · no email yet $33 OpenPhone seat + $30 Outscraper list · <$150 total spend Gate 0: ≥1 paid pilot or offer is broken Days 15–30 · Infrastructure + template library 3 domains + 9 inboxes warming · tool stack · 105 template scripts E&O+GL bound by D21 · phone+LinkedIn+Loom to validated vertical Gate 1: 8–10 discovery calls, 2 pilots OR 1 retainer + 1 pilot Days 31–60 · Convert pilots, email comes online 2,000 emails/wk added · pilots → retainers · VA hired Gate 2: ≥2 retainers, 4 pilots sold, <7d time-to-first-video Days 61–90 · Hit kill-switch gate · expand vertical mix Add second vertical · 6K emails/mo + 200 calls + 300 LinkedIn Build first internal case study Gate 3: $7.5K MRR · 3 retainers · 40%+ conversion Months 4–6 · Compound to $15–22K MRR Case-study-driven outbound · referral program · 1 annual prepay Operator 2 → FT only at $10–12.5K MRR with qualified pipeline Months 7–12 · Choose the fork Path A: niche dominance · raise to $2,997 · subcontract producer Path B: productize $497/mo low-touch tier · commodify the bottom
08 · Financial Projections

Three scenarios · realistic case haircut 20–30%

All scenarios: 1 FT + 1 PT operator, sub-$1K startup, full cold outreach, recommended package structure. Owner cash assumes some subcontract help by midyear and no full salary load. Pre-tax personal income per operator combined.

MRR trajectory · pessimistic vs realistic vs optimistic
Year-1 MRR by month across three scenarios · triangulated
$55K $41K $28K $14K $0 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 Gate 3 · $7.5K
Optimistic — $45–55K M12 · $420–520K Y1 · $130–180K/op
Realistic — $24–34K M12 · $230–290K Y1 · $75–105K/op
Pessimistic — $12–15K M12 · $100–140K Y1 · $25–45K/op

Triangulated scenario summary

ScenarioDay 90 MRRMonth 6 MRRMonth 12 MRRY1 revenueOwner cash (each)
Pessimistic$4K–$5K$8K–$10K$12K–$15K$100K–$140K$25K–$45K
Realistic$7.5K–$10K$15K–$22K$24K–$34K$230K–$290K$75K–$105K
Optimistic$15K$28K–$35K$45K–$55K$420K–$520K$130K–$180K

Realistic year-1: $230K–$290K revenue, $75K–$105K take-home per operator. Break-even on cash basis around month 3–4. This is the central case if execution lands competently. Combined Y1 take-home of $150K–$210K is competitive with senior tech salaries while building an asset.

Pessimistic year-1: $100K–$140K revenue, $25K–$45K per operator. This fails the "cash-flow positive immediately" criterion. Below-market for two operators against consulting day rates. In this scenario, fold by month 6 — either one operator continues solo, or shut down entirely.

Key sensitivities · impact on Y1 take-home

VariableRangeImpact
Reply rate1.5% → 3%Nearly doubles take-home
Monthly churn6% → 3%Compounds dramatically by M12
ARPU$1,800 → $2,800~55% revenue shift
Pilot → retainer conv.per 10pp~$10–15K/yr MRR by M12
Fully-loaded CAC$700 → $2,500Decides scale vs. burnout
09 · Insurance & Legal

The 2026 AI exclusion is real — but it's not automatic

$1M/$1M E&O + GL + Cyber: ~$1,500–$2,500/year for a 2-person services LLC under $200K revenue. Hiscox, Embroker, Vouch lead for AI/creator services in 2026. Bind by day 21 — single-claim defense averages $35–75K, and SMB clients increasingly require COI before signing.

Critical 2026 wrinkle, correctly framed: ISO Forms CG 40 47 and CG 40 48 (effective Jan 1, 2026) are optional endorsements, not automatic universal exclusions. Carriers can attach them to CGL policies, and Berkley's "Absolute AI Exclusion" and Hamilton's named-model exclusions are spreading. Action: line-by-line broker review. Demand in writing whether CG 40 47/48 endorsements are attached and whether E&O contains an AI carve-out. If excluded, get an affirmative AI endorsement, switch carriers, or buy standalone AI E&O wrap (Relm, Armilla, Testudo, Munich Re).

Contract stack · matters more than insurance

  • Limitation of liability capped at three months of fees paid (not aggregate)
  • Performance disclaimer — no guarantee of reach/leads/conversions/revenue
  • 2 revision rounds per deliverable, additional billed
  • AI disclosure — client acknowledges AI use, takes review responsibility
  • IP indemnification flow-down limited to underlying vendor TOS
  • 30-day kill on retainers, fees through notice period
  • IP assignment conditioned on full payment
  • Written releases for any patient/customer likeness; never use without explicit authorization
  • No medical efficacy claims in med spa or dental work; archive approvals

Substantive AI risks · concentrated and avoidable

RiskMitigation
IP claims from generated assetsUse Adobe Firefly Video for deliverables needing IP indemnification — only major model offering enterprise contractual indemnity
FTC Rule on Reviews/TestimonialsNever use AI avatars as fake testimonials. Current FTC civil penalty: $53,088/violation (16 CFR 1.98). "Operation AI Comply" active since Sept 2024.
Music licensingEpidemic Sound · Artlist · Soundstripe for paid placements. Trending TikTok audio is organic-only.
Likeness/voice rightsGet releases for any real person depicted via AI. ELVIS Act (TN) covers voice. NO FAKES Act pending. 47 states have deepfake statutes.
10 · Risks & De-risking

The three killers — and the cheap experiments to test them

Three risks can kill this: cold-outreach channel collapse, AI tool commoditization eating margin, and SMB churn compounding faster than acquisition.

Cheapest de-risking experiment · under $150 · this is Gate 0. Before spending a dollar on email infrastructure, run a 14-day phone-only paid test. $33 OpenPhone seat, $30 Outscraper for 500 owner numbers, 60 dials/day for 14 days. If you can't book ≥10 discovery calls and close ≥1 paid pilot in 14 days, the offer is broken. Fix it before warmup investment.

Defensive rules

Single-client concentration
< 30%
of MRR
Single-channel dependence
< 60%
of new biz by M6
Tool replaceability
Parallel
paths through Veo · Runway · Kling
White-label fallback
Vidpros
$1K/mo fractional editor, negotiated wk 1
11 · The Honest Answer

Should you do this?

The unit economics work. The market is real, growing, and confused enough to want done-for-you. The tool stack is mature enough that two operators can deliver agency-quality output on under $500/month of subscriptions. There is genuinely a $1,800–$3,500/month price corridor where US-based, AI-native, vertical-specialist work outcompetes both productized offshore and full-service agencies. One of you can plausibly take home $75K–$180K in year one if execution lands.

The honest preconditions:

  • One operator FT, one PT until $10–$12.5K MRR with qualified pipeline
  • Day 14 phone-led pre-test validates offer before email warmup investment
  • Test home services and med spas before committing the vertical mix
  • Accept month 1 is sub-$5K cash; email channel takes 6–8 weeks to warm
  • Sell a $1,497 paid pilot, never free work beyond a 30-second teardown
  • First three weeks: template-library buildout before scaling outbound
  • Line-by-line broker review on AI exclusions; use Adobe Firefly for indemnified work
  • Gate 3 ($7.5K MRR by Day 90) written into a formal kill-switch agreement between operators
  • Abandon at Day 90 if MRR <$5K, conversion <25%, or every sale needs discounting

Under those conditions, proceed. Without them, the path is statistically more likely to grind both operators through a six-month plateau at $5–10K MRR and exit with sunk-cost regret. The plan is designed to make the first 90 days legibly testable and the kill criterion non-negotiable, because honest pessimism on the front end is what makes the upside actually reachable on the back end.