Scaling, International & Exit

Part of the AMZ Operator Series (companion to IDS) | Built July 2026 | Distilling: Joe Valley (Quiet Light, The EXITpreneur's Playbook), Kevin King (Billion Dollar Seller Summit), Tomer Rabinovich (Ride the Amazon Wave), Empire Flippers & Quiet Light market data, Marketplace Pulse (aggregator post-mortem), Amazon Global Selling documentation

---

Every decision after your first profitable month is one question: are you building income or an asset? Income maximizes this month's payout; asset thinking maximizes the multiple paid for your last twelve months. $400k of SDE sells for ~$1.0–1.4M in 2026: every durable SDE dollar is worth ~$3 at exit, and every risk removed is worth more than most growth you can buy.

---

#1. THE SCALING DECISION MATRIX

#1.1 Five Paths, Ranked by Risk-Adjusted ROI

Work top-to-bottom: variations and same-avatar launches compound assets you already own (rank, reviews, suppliers, ad data); international and new channels multiply operational surface area — earn them after the core is systematized (Section 5).

#1.2 Same-Avatar Rule (King) and Batch Expansion (Rabinovich)

King's portfolio rule: build around an avatar, not a niche. Pick the person — trail runner, home barista, van-lifer — and own ten things they buy, across categories if needed. Payoffs: one brand story, cross-sell, an audience that launches every new product, and a brand at exit instead of a pile of ASINs. Buyers pay 4x for "the brand runners trust," 2.5x for "twelve unrelated SKUs sharing a bank account."

Rabinovich's discipline (Ride the Amazon Wave): launch different products to the same people, in batches — not one precious bet at a time; double down on winners, kill losers before they eat a container of cash. His warning: the second product is harder than the first — systematize the launch checklist before multiplying it.

#1.3 When to Kill SKUs — The Bottom-20% Cull

Twice a year, rank every SKU by contribution-margin dollars (not revenue, not units):

CM per SKU = Revenue − COGS − FBA/referral − returns − storage − direct ads
- Bottom 20% by CM$: kill, merge into bundles, or reprice-to-exit.
- Negative CM after ads 2 straight quarters: kill unless it measurably
  feeds a hero SKU (data, not vibes).
- Eating >15% of inventory capital while producing <5% of CM$: kill.

Culling frees capital for winners (AMZ_09 cash math), cuts storage and ad waste, and cleans the P&L cohort a buyer will scrutinize. Kill early.

---

#2. ON-AMAZON GROWTH LEVERS

#2.1 Subscribe & Save — LTV Without Leaving the Platform

Verified July 2026: eligible FBA products auto-enroll at 0% seller-funded discount, Amazon funding 5% on deliveries with 5+ subscription items; you may fund an additional 5% or 10%. No program fee beyond standard fees. (Coupons changed June 2, 2025 — $5 flat + 2.5% of coupon-attributed sales — if you use them to drive S&S trial.)

A subscriber is a recurring order with zero ad cost after acquisition. Consumables: fund 10% on heroes until subscribers exceed 5–10% of units, then test 5%. Failure modes: S&S stacking with coupons into accidental 40%-off orders, and stockouts — subscriptions get cancelled, not backordered. Documented subscription % is a multiple-driver (6.2): FBA's closest thing to recurring revenue.

#2.2 Amazon Business (B2B) — The Ignored 10%

Amazon's B2B channel passed $35B global sales (2024). Amazon's data: business prices ≈ +10% sales lift; quantity discounts ≈ +20% lift / +25% units; B2B orders run 25–30% higher value with ~40% fewer returns. Since Oct 1, 2025, business discounts ≥3% earn reduced fees on bulk orders; ≥5% earns the blue "Business Discount" badge. Setup is an afternoon: business price 3–5% below retail; quantity breaks at 5/10/25+ priced off marginal economics (zero incremental ad cost). Benchmark: 5–15% of revenue within two quarters — and a diversified buyer base DD teams notice.

#2.3 Brand Tailored Promotions & Premium A+

Brand Tailored Promotions (Brand Registry; US, UK, DE, FR, IT, ES, JP): Amazon's only sanctioned "retargeting" — discounts to built audiences: brand followers, repeat customers, cart abandoners, high-spend customers; minimum audience ~1,000 (followers exempt). Standing plays: 10–15% to cart abandoners, thank-you promo timed to reorder cycles, follower promo on every launch.

Premium A+ (verified gate): published Brand Story on brand ASINs + ≥5 approved A+ projects in the last 12 months → unlock notice in A+ Content Manager. Video, larger modules, comparison tables; mid-single-digit conversion lift, free once unlocked — batch five basic A+ approvals now.

#2.4 Multi-Variation Strategy — the 2026 Review-Pooling Reset

Variations pool traffic and (historically) reviews under one parent. That changed: announced Jan 7, 2026, rolling out Feb 12–May 31, 2026, Amazon stopped sharing reviews across variations judged materially different (function, material, use case). Forced splits are hitting abused families; deliberate pooling of unrelated products is enforced as catalog abuse / review manipulation — an account-level threat (AMZ_10 §2). Legitimate variations — true size/color/pack of the same product — still pool reviews and remain your cheapest launch lever. Audit families before Amazon does; unlink anything a reasonable human would call a different product.

---

#3. INTERNATIONAL EXPANSION

#3.1 The Decision Gate — Earn the Right First

GO INTERNATIONAL ONLY WHEN ALL FOUR ARE TRUE:
1. US contribution margin positive after ads, 6+ months (AMZ_09 dashboard).
2. Ops systematized: launches, restocks, CS run without founder daily input.
3. Cash for 2 extra inventory positions without starving US restocks.
4. No compliance landmine on hero SKUs in the target market (electricals,
   toys, food, cosmetics = certification cost before revenue).

Consensus order for US sellers, 2026: Canada (NARF) → UK → EU → Japan. Each step adds one difficulty: CA adds nothing (same account, US stock), UK adds VAT but keeps English, EU adds multi-country VAT + GPSR/EPR, Japan adds language, tax agent, import structure. Broker valuation data: US ≈ 76% of sold-business revenue, UK ~18%, the rest small — international is a margin supplement, not a second business.

#3.2 NARF — The Toe in the Water

Remote Fulfillment with FBA: US FBA inventory serves amazon.ca / com.mx / com.br through the unified North America account. No inventory export, no foreign tax registration — the customer is importer of record, paying duties at checkout. Remote fees run ~30–60% above domestic FBA; delivery 7–12 days CA, 5–9 MX. Thin margin, capped conversion — fine: NARF is a demand test, not an endgame. Read 90 days of sales by SKU, then move real inventory into Canadian FBA for proven SKUs only (local fees, Prime speed, GST/HST registration).

#3.3 Europe — Where Margin Dies Without a System

Biggest prize after the US; biggest compliance stack anywhere:

Pan-EU vs EFN: EFN = stock in one country, Amazon cross-ships EU-wide at higher per-unit fees — one VAT registration + OSS. Pan-EU = stock across DE/FR/IT/ES/PL/CZ/NL — local fees (up to ~53% cheaper per unit) but VAT registration in every storage country. Rule of thumb: Pan-EU pays at ~50+ units/day EU-wide; below that, EFN wins. A 1.5% fuel/logistics surcharge hits EU/UK FBA fees from April 17, 2026.

Sequence: UK first (English, single VAT), prove CM, Germany via EFN, Pan-EU when volume justifies. Native translation for DE/FR/IT/ES — machine-translated listings convert like machine-translated listings.

#3.4 Japan — Highest Barrier, Most Loyal Market

Third-largest marketplace by traffic; thin competition in many niches; quality-obsessed customers. Non-negotiables (verified 2026):

• JCT (10% consumption tax): non-resident FBA sellers must register — no revenue threshold — via a required local tax agent; business buyers expect a qualified-invoice-issuer number.
• Import structure: non-residents cannot ship to Japanese FCs alone — you need an IOR or ACP (Attorney for Customs Procedures).
• Localization: native-Japanese copy and local keyword research (not translated US keywords), Japanese CS (FBA covers baseline), separate JP trademark, product-safety docs for regulated categories.

Enter Japan third or fourth, with an experienced agency on retainer — or don't.

#3.5 Expected Net-Margin Reality by Marketplace (directional)

If a marketplace can't model to double-digit CM at steady state, skip it — the capital compounds faster in same-avatar US launches.

---

#4. OFF-AMAZON — DIVERSIFICATION BUYERS PAY FOR

#4.1 Shopify + MCF + Buy with Prime

Verified 2026: Amazon's combined "MCF and Buy with Prime" Shopify app manages catalog, orders, and returns inside Shopify admin; the legacy MCF US app deactivates Sept 30, 2026 — migrate. Buy with Prime puts the Prime badge and Amazon fulfillment/returns on your own site with integrated checkout; app free, per-order MCF/BwP fees. One FBA pool feeds both channels. DTC at even 10–15% of revenue diversifies traffic and builds the email list Amazon will never give you — launch weapon plus exit asset.

#4.2 TikTok Shop — Real Volume, Regime Risk

2026 state: 6% referral fee most categories (new sellers 3% for 30 days), payment processing ~1–3.8%. Seller-fulfilled shipping eliminated March 31, 2026 — US sellers must use Fulfilled by TikTok (FBT, ~$3.58–4.28/unit, cheaper multi-unit), TikTok Shipping, or CBT. All-in take with affiliate commissions (10–20%) and ads commonly hits 30–45% of GMV: a customer-acquisition engine with fulfillment attached, not a margin channel. Fits demo-able, impulse-priced (<$40) beauty/consumable/gadget products with creator flywheels. Treat FBT inventory as marketing capital; given policy volatility — participate, don't depend.

#4.3 Walmart Marketplace — The Grown-Up Second Channel

No monthly fee; referral 6–15%. Approval wants a real business: EIN + IRS verification letter (W-8BEN-E for internationals), US returns capability or WFS plan, 1–2 day shipping, marketplace track record. WFS: FBA-like fees; storage $0.75/ft³ (Jan–Sep); long-term storage ($7.50/ft³/mo past 450 days) lands June 30, 2026. The 2026 New-Seller Savings program stacks up to ~$75k: 20–40% referral discounts by GMV tier, 10% off WFS, ad credits. MCF may fulfill Walmart orders only with non-Amazon carriers, unbranded boxes; WFS or 3PL is cleaner. Expect 5–15% of Amazon revenue from the same catalog — boring, incremental; buyers like boring.

#4.4 eBay, Etsy, and the Diversification Rule

eBay: liquidation channel (returns, retiring SKUs), durable for parts/hobby niches. Etsy: only if genuinely craft-adjacent. The rule buyers actually price: <80% revenue concentration on Amazon plus one owned-traffic asset (email, social, DTC) earns a visibly better multiple and a deeper buyer pool (6.2). Diversification isn't leaving Amazon; it's not being only Amazon.

---

#5. TEAM & SYSTEMS — THE TRANSFERABILITY LAYER

#5.1 Org Chart: Solo → $1M+

$0–300k/yr   YOU: everything. Write SOPs as you go — hires run on them.
$300k–1M/yr  YOU: strategy, sourcing, cash.
             VA #1 — Catalog/CS (listings, cases, reviews, Seller Support)
             VA #2 or agency — PPC (rules-based, weekly report vs targets)
$1M–5M/yr    + VA #3 — Sourcing/logistics (POs, freight, restocks)
             + Bookkeeper (accrual books — non-negotiable pre-exit, 6.4)
             + Brand manager / senior VA owning launches end-to-end
Your job each stage: fire yourself from the lowest-value seat you hold.

#5.2 Hiring Sequence & Rates (Philippines VA market, verified 2026)

Rules: hire after the SOP exists (a VA without an SOP automates your chaos); paid trial task first; least-privilege sub-accounts only (AMZ_10 §6.3 — shared logins breed related-account suspensions); overpay whoever runs the checklist without you — that person is literally worth a higher multiple.

#5.3 SOP Library Architecture

One folder tree, one format: a Loom recording + a one-page doc per process (purpose → trigger → steps → owner → escalation). Minimum library (~25 SOPs): restock calc, POs, freight, shipments, listing create/edit, keyword refresh, PPC cycle, review/CS responses, account-health triage (AMZ_10 §1.3), promo calendar, month-end books. The test: a competent stranger runs the process from the doc alone — that is what "transferable" means in due diligence.

#5.4 Weekly Operating Rhythm

MON — Metrics Monday: CM$ by SKU, sessions/CVR, TACOS, weeks-of-cover,
      Account Health check #1 (AMZ_10). 60 min, decisions logged.
TUE — PPC cycle: bids, negatives, budgets per ruleset.
WED — Supply chain: PO/freight vs restock deadlines.
THU — Growth block: launches, variations, international/channels.
FRI — CS/review sweep, Account Health #2, one-page week close.

#5.5 Tool Stack (verified prices, July 2026)

Total for a $1M seller: $200–500/mo. Tools are never the constraint; the operating rhythm is.

---

#6. EXIT — THE ASSET EVENT

#6.1 Valuation Mechanics

VALUE = SDE × MULTIPLE       (SDE = Seller's Discretionary Earnings)

SDE = Net profit
    + owner salary AND payroll taxes on it
    + owner personal expenses run through the business
    + one-time costs (trademark/patent filings, lawsuit, rebrand)
    + non-transferring costs (mastermind, unused software)
    ± market cost of owner work a buyer must replace

Joe Valley (The EXITpreneur's Playbook) catalogs ~18 legitimate add-back types and one commandment: document every add-back or lose it in DD. An accepted add-back is worth its amount × your multiple — $2k/mo at 3.2x ≈ $77k of price. His four pillars of value — growth, risk, transferability, documentation — map to every prep action in 6.4.

FBA multiples, 2025–2026 (broker-verified ranges):

Calibration: Empire Flippers quotes multiples monthly — FBA listings ~32x monthly net profit, closing ~26.6x (÷12 ≈ 2.2–2.7x annual closed; ~3.1x average 2024; ecommerce ~3.4x in the 2026 State of the Industry report). Quiet Light: 3.5–4.0x for clean, evergreen brands. Direction is unanimous: 2026 buyers price 30–40% below 2021 aggregator peaks, with real rigor on contribution margin and channel risk. DTC brands with owned traffic command a premium over pure-FBA at equal SDE — one more reason Section 4 exists.

#6.2 What Moves the Multiple

The left column is 12–24 months of deliberate work; the right is why "my buddy got 5x" doesn't apply to unprepared businesses.

#6.3 Buyer Landscape 2026 — After the Aggregator Autopsy

The aggregator era is over as a bid source. 100+ aggregators raised ~$15B+ (2020–22), paid up to 7x for what Thrasio's own leadership later called garbage, and died of it: Thrasio filed Chapter 11 in February 2024, shed ~$495M of debt, emerged June 2024 a smaller brand operator; Perch folded into Razor Group. Marketplace Pulse's verdict — "The Amazon Aggregator is Dead" — holds. Today's buyers pay rational prices:

#6.4 Exit Prep — The 12–24 Month Checklist

□ ACCRUAL books from this month (cash-basis P&Ls torch FBA valuations —
  inventory timing distorts margins). Backfile 24 months if possible.
□ Separate entity: dedicated bank/cards; strip personal spend or document
  it as clean add-backs (Valley: "document or lose it").
□ Supplier contracts: written, assignable, exclusivity on heroes;
  develop supplier #2 (single-supplier is a DD finding — AMZ_04).
□ IP file: trademarks in every marketplace sold, patents/design regs,
  domains/socials, Brand Registry access map.
□ Remove yourself from ops: SOPs + team run the rhythm; your hours
  <10–15/wk, provable.
□ Metric stability: 12 months steady CM%, TACOS, review velocity; no
  deferred restocks or slashed ads to juice SDE — buyers model it out.
□ Concentration: no SKU >40–50% of revenue; B2B/S&S/second channel live.
□ Account hygiene: AHR healthy (AMZ_10), zero open violations, no
  related-account skeletons; case log exportable.
□ Data room: 24-mo P&L, SKU-level CM, supplier list, org chart, SOP index.

#6.5 The Process — LOI → Wire

Sequence: valuation & packaging → buyer calls → LOI (price, structure, ~30-day exclusivity) → due diligence (30–45 days: financials reconciled to bank + Amazon settlement reports; ops, legal, account health) → APA (reps & warranties, 3–5 yr non-compete, 30–90 day transition) → migration (Seller Central transfer, escrow, 2–4 weeks) → wire.

Structure norms (2025–26 verified): smaller deals close 60–80% cash at close; above $1M, earnouts appear in nearly every deal (one broker cohort: 11 of 12 deals >$1M; ~66% average upfront). Balance = earnout (6–24 months on revenue/profit targets), stability payments, or seller notes. Inventory is typically paid separately at landed cost on top of the headline price — negotiate the cap and in-transit treatment explicitly.

What kills deals in DD: P&L that won't reconcile to settlement reports; fee/COGS misclassification inflating SDE; undisclosed black-hat (review schemes surface — AMZ_10 §7 applies to buyers too); inventory surprises (stale stock counted good, unpaid supplier balances); related-account risk; starving ads/inventory during DD to harvest cash. Assume everything will be found — fix it or price it.

Earnout self-defense: cap at-risk ≤30%; tie targets to revenue (buyer-controlled costs corrupt profit targets); define the metric source (Seller Central reports, not buyer's books); acceleration on buyer breach; treat every earnout dollar as bonus, not price.

---

#7. HOW OPERATORS DIE HERE

1. Scaling into cash death — three launches + a Pan-EU stock split funded from float; one slow quarter and payroll bounces. Growth consumes cash before returning it — AMZ_09's cash-conversion math governs every move here.
2. The international VAT hole — a year selling in Germany "while we sort registration": back taxes + penalties + Amazon withholding exceed the revenue earned. Register before the first unit lands.
3. Hiring before SOPs — five VAs, zero documented processes; you now manage chaos at $6/hr instead of doing it free.
4. Waiting for one more good year — multiples reward stability at sale, not your peak. Sellers who rode 2021 "one more year" met 2023 multiples 40% lower. Sell into strength.
5. The earnout trap — 50% of "price" contingent on targets the buyer controls, measured on the buyer's books. Not price; a lottery ticket with your name misspelled.
6. Fad-scaling — capital poured into a trend's peak, then holding the decline cohort AND the inventory. The DD chart is built two years before the sale.
7. Selling a job, not an asset — cash-basis books, founder answering CS at midnight, one supplier, one SKU at 60%. The offer comes in at 2.2x, and it's fair.

---

#MODULE SUMMARY — THE TEN COMMANDMENTS OF SCALING & EXIT

1. Scale in ROI order: variations → same-avatar products → international → channels, B2B on from day one. Never skip a rung for a shinier one.
2. Build around an avatar (King), launch in batches (Rabinovich), and cull the bottom 20% of SKUs by contribution-margin dollars every six months.
3. Earn international: US CM positive 6+ months, ops systematized, spare cash — then CA (NARF test) → UK → EU (EFN until ~50 units/day, then Pan-EU) → Japan last, with a tax agent and native localization or not at all.
4. Register before you sell, everywhere: VAT where stock sits + OSS, GPSR Responsible Person, EPR in DE/FR, JCT from sale one in Japan. Compliance debt compounds faster than revenue.
5. Diversify below 80% Amazon concentration — Walmart for boring volume, Shopify+MCF/Buy with Prime for owned customers, TikTok as paid acquisition that ships. Buyers pay for the second leg.
6. Hire after the SOP, never before. One Loom + one page per process; a business a stranger can run from the docs is what "transferable" means — and multiples are made of transferable.
7. Run the equation daily: price = SDE × multiple. Grow documented SDE (add-backs included); remove multiple-killers — SKU concentration, single supplier, owner-dependence, PPC-only revenue.
8. Know your buyer (2026): aggregators are dead (Thrasio Ch. 11, Feb 2024); real buyers are SBA individuals, search funds, PE platforms, and strategics paying ~2.5–5x SDE, 60–80% cash at close, inventory at cost on top.
9. Prep 12–24 months out: accrual books, assignable supplier contracts, IP filed, metrics stable, yourself out of ops, account health spotless. Every DD surprise is a price cut you funded.
10. Sell into strength, structure with paranoia: earnouts ≤30% on revenue-based, verifiable targets; inventory paid at cost. The multiple rewards stability today — not the year you're hoping for.

---

Sources consulted (July 2026): Empire Flippers (2026 State of the Industry, valuation/earnout guides); Quiet Light (2025 FBA Report); Joe Valley, The EXITpreneur's Playbook; Marketplace Pulse ("The Amazon Aggregator is Dead"); CNBC/TechCrunch (Thrasio Ch. 11); Kevin King; Tomer Rabinovich, Ride the Amazon Wave; Amazon Seller Central / sell.amazon.com (S&S, B2B, Brand Tailored Promotions, Remote Fulfillment, Pan-EU/EFN, GPSR/EPR); Amazon.co.jp JCT docs; Rising Sun Commerce; SimplyVAT-class guides; Walmart Marketplace Learn; TikTok Shop fee analyses; Buy with Prime knowledge center; Helium 10 & Sellerboard pricing; OnlineJobs.ph VA salary reports.