Advertising

Lineage: upgraded from IDS_04_Advertising.md. Practitioner base: Cody Plofker (Jones Road Beauty), Nick Shackelford, Savannah Sanchez, Andrew Foxwell, Barry Hott, Tim Masek, Sam Tomlinson, John Moran, Kirk Williams, Chase Chappell, Tom Breeze, Ralph Burns, Ben Malol, Nick Peroni, Depesh Mandalia, Maxwell Finn, Harry Hawk, Molly Pittman, Ryan Deiss. Current as of July 2026.

---

#THE ONE-PAGE VERSION

1. Creative is the targeting. Meta and TikTok's algorithms now do the audience-matching work that used to be your competitive edge. Your ad's hook, angle, and format decide who sees it and whether they buy.
2. The Advantage+ ceiling is real for small spenders. Andromeda-era Advantage+ campaigns run ~78% of Meta spend and cut CPA by −38% for accounts spending $10k+/month — but only −14% under $2,000/month. A $1k–$2k operator cannot outsource skill to the algorithm; creative quality is still the lever.
3. Meta mid-2026 e-comm benchmarks: CPC $1.35, CPM ~$10.42, CVR 2.81%, CPA ~$28.40 (median ~$30), blended ROAS ~3.7×. Niche CPA: pets ~$25, apparel ~$22, beauty ~$32, fitness ~$42, electronics ~$46.
4. July 2026: Meta removed the off-platform activity opt-out (folded into "Activity from other businesses"). Retargeting pools grow as a result — expect more signal, but also more regulatory and privacy scrutiny. Don't assume this window stays open indefinitely.
5. TikTok is still the cheapest cold-acquisition channel (CPM $4.80–9.16, 20–40% under Meta) but reach is volatile post-JV as the algorithm retrains on US data. Budget for noisier delivery through 2026, not a broken platform.
6. TikTok Shop changed its rulebook in July 2026: Account Health Rating (AHR) replaced the old violation-points system, Store Rating recalculates continuously, a 60-day After-sales Handling Time metric replaces Customer Complaint Rate, and ad spend now consolidates under GMV Max. Shop-tagged content converts 3.7% vs. 1.8% for non-Shop content — tag everything you can.
7. Google's consensus play is PMax + Standard Shopping hybrid, not PMax alone. Demand Gen is the only conversion-optimized video campaign type left (Video Action Campaigns died April 2025). YouTube Shorts is the cheapest video reach on the platform (CPM ~$4–4.85).
8. Cheap-inventory windows exist outside the big three. Threads ads are in an early-mover CPC window (~$0.68), IG Reels CPM runs ~$8.67, Pinterest retail CPC $0.50–0.70, Reddit CPC ~$1.25. Use them to extend reach once your core channels are profitable, not as a substitute for them.
9. Budget honestly by stage. A $1,000 lean operator runs organic-first and reserves $20–30/day paid tests for after an organic or affiliate signal exists. Paid-first cold testing realistically needs $3,000–10,000 to find a reliable winner — don't let anyone sell you a $500 "guaranteed" launch.
10. Seasonality moves CPMs 2–3×. January e-comm CPMs run $10.80–15.74; November runs $25+. Q4 cold testing is for veterans with a proven creative bank — new operators test in Q1–Q2 and scale into Q4 on known winners.
11. The 20%/48–72h scaling rule still holds. Never raise budget more than 20% in a 48–72 hour window; every change resets learning. Duplicate winners instead of editing them.
12. Retargeting is a defined architecture, not a leftover audience. 30-day site visitors, video/engagement viewers, and past purchasers (90–365 days, excluding recent buyers) are three separate audiences with three separate creative jobs — social proof, objection handling, and winback respectively.
13. Track three numbers daily: MER, CPP, and NC-CPP. Blended ROAS lies at scale. New Customer CPP is the only number that tells you whether the brand is actually growing or cannibalizing its own list.
14. Creative volume is the new moat. AI-assisted hook generation, variant batching, and structured UGC briefs let a solo operator ship 10–15 concepts a week without an agency. The brand that tests more, wins more — this compounds.
15. This module ends where LUCE_14 (Advertising Mastery) begins. Once you're past $5k/month and need contribution-margin bidding, incrementality testing, and full cross-platform attribution, go there. For daily measurement discipline (MER math, dashboards, attribution windows), see LUCE_06.

---

#SECTION 1: THE MENTAL MODEL FOR PAID ADS

#1.1 The Algorithm Reality (Mid-2026)

Meta, TikTok, and Google's ad systems have moved even further from 2018–2021 targeting logic since the original playbook was written. The shift that mattered then — broad beats narrow — has hardened into something closer to a rule.

What changed since the original module: Meta's Andromeda AI overhaul (fully live since October 2025) now handles roughly 78% of Meta ad spend through Advantage+. That sounds like the algorithm has made media buying easy. It hasn't — not for you. Andromeda's CPA advantage is −38% for accounts spending $10k+/month, where the system has enough purchase volume to learn fast. Below $2,000/month, the edge shrinks to −14%. You are not spending enough for the algorithm to learn your customer quickly, which means the old lesson still applies at your budget: the algorithm buys audiences, you compete on creative quality.

The old model (pre-2022):

• Detailed audience targeting was how you won.
• Find the right interest → right demo → show them your ad.
• Advertisers competed on audience precision.

The 2026 model:

• Broad targeting is default, not a testing tactic — Meta and TikTok both recommend it out of the gate.
• The algorithm finds buyers faster than manual targeting once it has signal, but a small account starves that signal.
• You compete on creative quality and signal strength (Pixel + CAPI + first-party data).
• The ad IS the targeting — the platform shows your creative to people who behave like people who engage with that creative.

Barry Hott's line from the original module is more true today than when he said it: "The creative is the targeting." Under $2k/month, it's closer to the whole game.

#1.2 The Creative → Signal → Audience Flywheel

Better creative quality
↓
Higher click-through rate
↓
Better audience signal from clicks (purchase intent)
↓
Algorithm finds more lookalikes of these high-intent users
↓
Lower CPMs (platform rewards relevance with cheaper impressions)
↓
Lower CPA
↓
More profit → more budget → test more creative
↓
Repeat

Operators who skip this and chase interest-targeting rabbit holes are the ones whose CPA runs double the benchmark. The fix is never a smarter audience — it's a sharper hook.

---

#SECTION 2: META ADS (FACEBOOK + INSTAGRAM)

#2.1 Campaign Structure — Testing vs. Scaling

For testing ($0–$5k/month spend):

Campaign: [Product] — Testing
Objective: Sales (Conversions)
Budget: CBO (Campaign Budget Optimization) $50–$100/day

Ad Set 1: Broad (no interests, age/gender only)
  → 3 creative variations

Ad Set 2: Interest stack (5–10 relevant interests)
  → 3 creative variations

Ad Set 3: Lookalike 1% (from customer list or pixel data, if available)
  → 3 creative variations

For scaling ($5k–$50k/month):

Campaign 1: Advantage+ Shopping (let Meta run everything)
  → 5–8 creatives, Meta distributes budget
  → Often your best-performing campaign at this spend level

Campaign 2: Manual CBO — Broad
  → 1 ad set, 10+ creatives
  → Kill creatives with ROAS <2× after 7 days; duplicate winners

Campaign 3: Retargeting
  → Audience: Visited product page, didn't purchase (90 days)
  → Budget: 10–15% of total spend
  → Higher-conversion offer (discount, urgency, new angle)

The Advantage+ note, corrected for 2026: Advantage+ Shopping Campaigns (ASC) remain the top-performing campaign type for most established stores. But "run it and stop thinking" is beginner advice from an earlier era of the algorithm. At your budget (see Section 1.1), Advantage+ still needs your best creative fed into it — the automation handles delivery, not creative strategy. Your job doesn't shrink; it narrows to one thing: creative.

#2.2 Creative Testing System (The 3-2-2 Method)

Cody Plofker (CMO, Jones Road Beauty, scaled to $50M+/year):

The framework:

• 3 different hooks (opening 3 seconds)
• 2 different bodies (middle section)
• 2 different CTAs (closing offer)

Produces 12 unique ad combinations from 7 creative assets. Test all 12, scale the winners.

Hook types that work for e-commerce in 2026:

#2.3 The Creative Brief Template

CREATIVE BRIEF

Product: [Product name]
Target audience: [Who is this person? Be specific]
Their biggest problem: [What keeps them up at night?]
The outcome they want: [What does their dream life look like related to this problem?]
Why our product is different: [The mechanism — how it works differently from alternatives]
Proof/credibility: [Social proof, certifications, founder story, reviews]
The offer: [Price, any discount, guarantee]
CTA: [What action do we want them to take?]

Hook options to test:
1. [Problem-based hook]
2. [Results-based hook]
3. [Curiosity/curiosity-gap hook]

Tone: [Conversational / educational / urgent / aspirational]
Length: [15s / 30s / 60s / static / carousel]
Platform: [Meta feed / Reels / TikTok / Shop-tagged / Threads]
Reference ads: [Link 2–3 competitor ads from the Meta Ad Library or TikTok Creative Center]

DO NOT: [Anything that violates ad policy — health claims, comparison claims, etc.]

#2.4 Scaling Framework (The 20% / 48–72h Rule)

The #1 amateur mistake: touching campaigns too frequently.

The algorithm needs 5–7 days of data to exit the learning phase. Every budget, creative, or audience change restarts it.

The operator rules:

1. Scale budget 20% maximum per 48–72h. Never double overnight.
2. Wait 7 days minimum before judging a new campaign. The first 3 days are noise.
3. Kill underperformers after 7 days. ROAS below breakeven after 7 days is dead — see Section 8 for the breakeven formula.
4. Duplicate winners; don't edit them. If an ad set works, don't touch it. Duplicate → edit the duplicate → test.
5. One change at a time. Changing audience, creative, and budget together tells you nothing about what caused the shift.

Vertical vs. horizontal scaling:

• Vertical: increase budget on existing winning ad sets. Hits a ceiling — diminishing ROAS past a point, faster creative fatigue.
• Horizontal: duplicate winning ad sets into new audiences, countries, or platforms. Extends the runway.
• Best operators do both simultaneously once past the testing phase.

#2.5 The Pixel, CAPI, and the July 2026 Retargeting Shift

Without proper tracking, you're flying blind — the algorithm optimizes on the signal it receives.

1. Meta Pixel: Shopify → Settings → Customer Events → Meta (native integration). Test events in Events Manager → Test Events. Confirm PageView, ViewContent, AddToCart, InitiateCheckout, Purchase all fire.

2. Conversions API (CAPI): Server-side tracking, bypasses browser privacy restrictions. Shopify → Settings → Customer Events → Add connection → Meta Conversions API. Non-negotiable. With Pixel + CAPI, you report 80–90% of actual conversions; without CAPI, 40–60% — an underfueled algorithm.

3. UTM parameters: Every ad URL needs UTM tags for GA4/Triple Whale attribution: ?utm_source=facebook&utm_medium=paid&utm_campaign=cbo_test&utm_content=hook1_ugc. Build a UTM spreadsheet and never publish an ad without them.

4. What changed — July 2026: Meta removed the standalone off-platform activity opt-out, folding it into a broader "Activity from other businesses" setting. In practice, this means Meta's retargeting pools (people it can identify as having interacted with your business off-Meta) get larger, not smaller — more usable signal for your custom and lookalike audiences than you had in 2025. Two things follow: (1) your retargeting audience sizes may grow without you changing anything, so re-check your 90-day retargeting audience size before assuming a stale number; (2) this is a policy environment, not a permanent feature — plan your tracking stack (Section 2.5.1–4) to not be dependent on any single opt-out setting surviving another year.

#2.6 Meta Benchmarks — Mid-2026

Niche CPA (Meta): pets ~$24.56 · apparel ~$22.18 · beauty ~$31.65 · fitness ~$42.30 · electronics ~$46.18.

Budget starting points: $30–50/day per test at the testing phase; Advantage+ wants $50–100/day to learn efficiently; expect ~50 conversions per ad set per week before you trust the data.

#2.7 Ad Copy Formulas

Creative gets the click; copy earns the trust to complete the purchase. Use these structures as starting templates, not rigid scripts — the language should come from your customer-review mining (LUCE_03 Section 2.2), not from a formula alone.

AIDA+ (the DTC standard):

Attention: Hook (first line / first 3 seconds)
Interest: Problem amplification (1–2 lines)
Desire: Solution + differentiation (2–3 lines)
Action: CTA + urgency (final line)
+Social Proof: Woven throughout — not a separate closing line

PAS (Problem-Agitate-Solve) — shortest, best for cold traffic:

Problem: Name the pain in the customer's own words.
Agitate: Make the cost of not solving it concrete (time, money, embarrassment).
Solve: Introduce the product as the specific fix, with one differentiator.

The 4 U's (for headline/hook lines specifically): Useful, Urgent, Unique, Ultra-specific. A hook that hits 3 of the 4 outperforms one that hits 1.

• Weak: "Great skincare tool."
• Strong (hits Useful + Ultra-specific + Unique): "The 3-minute tool dermatologists quietly use before red-carpet events."

Retargeting copy formula (pairs with Section 2.8): lead with the specific product they viewed, name the objection directly (price, trust, "does it work"), answer it in one line, close with urgency or a bonus — never restate the original cold-traffic hook to a warm audience; they've already seen it.

#2.8 Retargeting Architecture — Three Audiences, Three Jobs

Retargeting is not "whoever didn't buy." It's three distinct audiences, each needing different copy and a different offer:

Budget split for retargeting overall: 10–15% of total ad spend once you're past pure testing phase. Each audience above gets its own ad set with its own creative — running one generic "retargeting" ad set against all three collapses the copy job back into one message and underperforms a segmented structure by a wide margin. The July 2026 opt-out change (Section 2.5) makes the 0–30 day pool larger than before; re-segment rather than letting it run as one flat audience (see Failure Modes).

---

#SECTION 3: TIKTOK ADS

#3.1 Why TikTok Is Different — And the Post-JV Reality

TikTok still runs the lowest CPMs of any major platform for e-commerce. Three things haven't changed: it's a less-saturated ad market than Meta, the content-native ad format is inherently cheaper to produce, and the algorithm is aggressive at finding buyers fast.

What's new since the original module: the TikTok USDS joint venture (Oracle/Silver Lake/MGX, ~45–50% ownership; ByteDance retains 19.9%) closed January 22–23, 2026. The ban risk that shadowed every 2025 TikTok strategy is over — but the platform's US algorithm is being retrained on a US-only data environment, and reach and delivery have been measurably more volatile through 2026 than they were pre-divestiture. Budget for noisier week-to-week performance; don't read a bad week as a dying platform, and don't read a great week as the new normal either. Give any TikTok test a full 5–7 days before judging it, longer than you might on Meta.

The TikTok vs. Meta distinction still holds:

• Meta users are used to being sold to → they filter more → higher CPMs.
• TikTok users are used to being entertained → they engage more openly → lower CPMs.
• Meta is stronger for brand-aware audiences and retargeting.
• TikTok is stronger for cold acquisition, especially 18–35.

Savannah Sanchez: "TikTok requires entertainment first, sale second. The moment you break from content-native format and go direct-response hard sell, you're dead."

#3.2 TikTok Creative Rules

What works:

1. UGC-style: someone on camera, talking like they're talking to a friend. No studio, no professional lighting.
2. Problem-solution-product story: hook with the problem (5s), show the struggle (10s), introduce the product (10s), show transformation (10s), CTA (5s). ~40s total.
3. Product demonstration: show exactly how it works — satisfying product moments are TikTok gold, often with no voiceover, just text overlays.
4. Trending audio (licensed): boosts both organic reach and ad engagement.
5. Text overlays: ~70% of TikTok is watched on mute. Critical information must be on screen as text.

What doesn't work: polished, produced ads; talking at the camera for 3+ seconds without a cut; mismatched music tone; heavy product claims in the first 5 seconds.

#3.3 TikTok Campaign Structure

Campaign: [Product] — [Month/Quarter]
Objective: Conversions → Purchase (or Website Conversions)
Budget: Campaign level, not ad group

Ad Group 1: Broad — Female 18–35
  → Creative: 3 UGC hooks
  → Placement: TikTok only (not partner networks)

Ad Group 2: Interest targeting — [Relevant interests]
  → Creative: 3 different UGC hooks, different angle from Ad Group 1

Ad Group 3: Custom audience — Website visitors 30 days (retargeting)
  → Offer-based creative (discount, urgency)

Budget starting point: $50–100/day minimum for the algorithm to learn. TikTok Creative Center (ads.tiktok.com/business/creativecenter) is free and shows top-performing ads by category — study format, hooks, and duration of category winners before you shoot.

Spark Ads: boost organic TikTok posts instead of running dark posts directly. Benefits: native format (comments/likes/shares accumulate), lower CPM than dark posts, visible social proof — particularly powerful for real customer UGC.

#3.4 TikTok Shop, GMV Max, and the July 2026 Rulebook

TikTok Shop is the single biggest structural opportunity in the channel, and its rules changed materially in mid-2026.

What changed since the original module:

• GMV Max now consolidates TikTok Shop ad spend — ad campaigns, LIVE boosts, and affiliate-driven content buying all run through one automated budget-allocation system rather than separate manual campaigns.
• Account Health Rating (AHR) replaced the old violation-points penalty system as of July 2026. AHR is a continuous, composite score — a slow supplier or a spike in returns now shows up here, not just in your return-rate report.
• Store Rating recalculates continuously rather than periodically.
• 60-day After-sales Handling Time replaces Customer Complaint Rate as the primary service metric TikTok watches.
• Shipping still requires USPS labels purchased through TikTok Shipping (in effect since January 2026); 2-day shipping and 48-hour tracking-scan rules remain in force. No retail-arbitrage dropshipping (foreign-branded direct shipping) is allowed.

The stat that should change your content strategy: Shop-tagged content converts at 3.7% vs. 1.8% for non-Shop-tagged content. Tag every product mention, every video, every LIVE — the in-app checkout removes friction that a bio link never will.

Setting it up:

1. TikTok Seller Center → create a shop, sync your Shopify catalog.
2. Enable the Creator Marketplace, set affiliate commission (start at 15–25% for an unproven product — you need pickup more than you need to protect margin early).
3. Creators post content on commission only; your job is to approve creators, ship samples, and monitor quality — which now flows directly into your AHR.

The affiliate math (referral fee, mid-2026): TikTok Shop's referral fee is 6% most categories, 3% for a new seller's first 30 days. At a 20% affiliate commission on a $50 item: $10.00 commission + $3.00 referral fee (first 30 days) or $10.00 + $2.00... wait — always recompute per your own price point; the formula is:

Per-order acquisition cost (TikTok Shop affiliate) =
    (commission % × sale price) + (referral fee % × sale price)

This is the closest thing to "free" paid traffic in e-commerce — creators take the creative and distribution risk, you take the inventory risk, and the cost only triggers on a sale.

---

#SECTION 4: GOOGLE ADS FOR E-COMMERCE

#4.1 When Google Makes Sense

Google is not where you start. It's where you add once you've proven demand elsewhere.

Start Google when:

• You've proven product-market fit on Meta or TikTok (roughly >$20k/month cumulative validated spend).
• Your product has clear search intent (people actively Google "buy [product]").
• Your blended MER is healthy and you need scale beyond social.

#4.2 The PMax + Standard Shopping Hybrid

What changed since the original module: the original taught "PMax is the only way to run Shopping now, Standard is being phased out." That's stale. The mid-2026 consensus among Google specialists is a hybrid: run PMax broad for automated reach across Search/Shopping/YouTube/Display/Gmail/Maps, and run Standard Shopping alongside it for brand-query isolation and manual control over your top SKUs. Standard Shopping is not dead — it's your control lever when PMax's opacity becomes a liability (you can't see which specific queries triggered a sale in PMax the way you can in Standard).

PMax setup:

1. Upload the product feed via Google Merchant Center, synced from Shopify.
2. Create the PMax campaign: 5 headlines, 5 descriptions, 5 images, 1–2 videos minimum.
3. Set the conversion goal to Purchases, with conversion value.
4. Target ROAS 300–500% (3–5×) for most categories — don't set it lower.
5. Let it run 30 days before judging; PMax needs more data than Search alone.

Standard Shopping (run in parallel):

• Use for your top 20% of SKUs where you want bid control.
• Isolate brand-name queries so PMax doesn't cannibalize your cheap brand-search traffic with broad-match spend.

#4.3 Brand Search Campaigns (The Hidden Profit Center)

Bidding on your own brand name. Competitors can and do bid on it — without your own bid, their ad can outrank your organic listing.

• Brand CPCs: $0.10–$0.80 — extremely cheap, highest-intent traffic.
• Conversion rates from brand search: 8–15%, roughly 10× cold traffic.
• Never turn this off; it's the single highest-ROI line in your Google account.

Campaign: Brand Search
Keywords: [brand name], [brand name] + product, [brand name] + reviews, [brand name] + discount
Match type: Broad match only (prevents competing with yourself)
Budget: $10–50/day is enough for most brands under $500k/year

#4.4 Demand Gen and YouTube Shorts

What changed since the original module: Video Action Campaigns were killed in April 2025. Demand Gen is now the only conversion-optimized video campaign type on Google. If you're repurposing Meta/TikTok video creative for Google, it runs through Demand Gen, not a legacy VAC setup.

YouTube Shorts is the cheapest video reach on the platform — CPM roughly $4–4.85, CTR ~1.24% (the best CTR on the platform). For a lean operator adding a second video channel, Shorts inside Demand Gen is a lower-cost on-ramp than in-stream YouTube before you have a dedicated creative budget for long-form (see LUCE_14 Section 5 for the full YouTube playbook).

#4.5 The Google–Meta Interplay

John Moran: "Meta creates demand. Google captures it."

The typical journey: see your ad on TikTok/Instagram → interested, don't buy → Google "[your product]" or "[your brand]" later → click your Shopping or brand-search ad → purchase. This is why Google's reported ROAS often looks artificially strong — it's capturing demand Meta already paid to create.

Implication: don't optimize Meta and Google independently. They're the same funnel, measured in two dashboards. See LUCE_06 for how to build a single blended-MER view across both.

---

#SECTION 5: THE CHEAP-INVENTORY PLAYBOOK

New for 2026. Meta and TikTok aren't the only channels with usable CPMs, and a lean operator should know when each of the following earns a place in the budget — not run all four at once on day one.

Sequencing rule: don't add a fifth channel until your first two (typically Meta + TikTok, or Meta + organic) are consistently hitting your MER target. Each new channel needs its own creative variant, its own measurement, and its own 60-day ramp budget (see LUCE_14 Section 7.3 for the full decision framework on adding channels).

---

#SECTION 6: CREATIVE PRODUCTION AT SCALE

#6.1 The UGC Creator System

UGC remains the backbone of e-commerce advertising: it looks organic, converts on trust rather than polish, is cheap to produce, and iterates fast (24–48h vs. weeks for studio work).

Where to source it:

• Billo.app — UGC creator marketplace, $50–150/video, fast turnaround.
• Minea.com ($49–99/mo) — shows which UGC-style creatives are performing right now, useful as a reference for briefs.
• TikTok Creator Marketplace — partner with real TikTok creators directly.
• Direct outreach — DM micro-influencers (5k–100k followers) in your niche: free product + $50–200 for content rights.

Briefing rule: give creators talking points, not a script. Scripted UGC sounds scripted and converts worse. Provide the problem the product solves, the key features, and any claim restrictions — then ask for 3 hook variants minimum, in 15s/30s/60s if possible, and let them keep their own voice.

#6.2 Static vs. Video — When to Use Each

Static works best for: retargeting, product-detail shots, promo announcements, Google Shopping (static required), high-ticket consideration purchases.

Video works best for: cold acquisition, TikTok (almost always video), any product that needs demonstration, first touchpoint in a new market.

The 2-format rule: run static and video simultaneously — they reach different users at different moments. Running only one format leaves money on the table.

#6.3 Ad Fatigue and Creative Refresh

Fatigue signs: CTR declining week-over-week, CPC rising, frequency >4, ROAS declining despite stable budget and audience.

Prevention: never stop producing. Even at $5k/month, test 10+ new creatives monthly. Cody Plofker's team at Jones Road Beauty tests 30–50 new creatives/month — that discipline is why their CPAs stay low.

#6.4 The AI-Assisted Creative Production System

New for 2026. Creative volume used to require a studio or an agency. It now requires a repeatable system a solo operator can run with an AI drafting layer and a review pass. This is not "let AI make your ads" — it's a workflow that turns one good creative concept into a dozen testable variants without a dozen new shoots.

1. The Hook Matrix. Build a spreadsheet: rows = hook types (Section 2.2's 7 types), columns = your top 3 customer angles (mined from reviews per LUCE_03 Section 2.2). Use an LLM to draft 3–5 hook-line variants per cell from your product brief and customer language — you get 21–35 candidate hooks in under an hour. Read every line for compliance (health/comparison claims) before it reaches a script.

2. Variant Batching. From one UGC shoot (one creator, one session), extract multiple ad variants by re-cutting: swap the first 3 seconds (hook), swap the CTA, swap the caption/text overlay. One 90-second raw UGC clip can become 6–9 distinct ad executions without asking the creator to reshoot. Use a video editor with a template (CapCut, Canva Pro at $15/mo) to standardize the cut points.

3. The UGC Brief Template, AI-drafted, human-approved:

INPUT: product brief, top 3 customer angles, 2 competitor reference ads
OUTPUT (draft by AI, edited by you before sending to creator):
  - 3 hook lines per angle (9 total)
  - Suggested shot list (3–5 shots)
  - Suggested on-screen text overlays (mute-proofing)
  - Explicit DO NOT list (claims, comparisons, restricted words)

Draft this with an AI assistant from your product's creative brief (Section 2.3), then edit for accuracy and compliance before it goes to a creator — never send an unreviewed AI draft as a real brief.

4. Weekly cadence: batch-generate the week's hook matrix and briefs in one sitting (60–90 minutes), then spend the rest of the week shooting, cutting, and testing. This converts creative production from a daily scramble into a weekly system — which is the actual difference between an operator who tests 3 concepts a week and one who tests 15.

---

#SECTION 7: BUDGET LADDERS

#7.1 The $1,000 Lean Tier — Organic-First

If you're starting with roughly $1,000 in proof-of-concept capital (see LUCE_03's validation ladder), paid ads are not where you start spending. Organic content and TikTok Shop affiliates are cheaper and — more importantly — tell you whether the product works before you pay to find out.

Rung 1 (organic, $0):    Post daily/near-daily on TikTok + Reels, 14–30 days.
Rung 2 (affiliate, $0–150): Seed 10–20 TikTok Shop affiliates on commission-only.
Rung 3 (paid test, $20–30/day): ONLY after Rung 1 or 2 shows a real signal
                                 (engagement, purchase-intent comments, or an
                                 actual affiliate-driven order). Run 4–5 days
                                 per platform, one angle per ad set.

Skipping straight to Rung 3 without a signal from Rung 1/2 is the single most common way a $1,000 test budget disappears with nothing learned. See LUCE_03 Section 3.2 for the full validation ladder this budget nests inside.

#7.2 The Honest Paid-First Tier — $3,000–10,000

If you're starting with paid acquisition as the primary channel (larger starting capital, an existing audience, or a product too complex for organic virality), be honest about the cost of finding a winner: realistically $3,000–10,000 before you have a creative-audience combination that clears breakeven with confidence. Budgets under this compress into 1–2 creative tests, which isn't enough data to separate a bad angle from a bad platform from a bad week.

Phase 1 — Foundation ($0–5k/month spend): establish pixel/CAPI signal,
  find 2–3 creative winners, target breakeven ROAS. This is learning,
  not scaling.
Phase 2 — Acceleration ($5k–30k/month): transition winners to CBO,
  add a second channel, target MER 3–4×.

(Full 4-phase model in LUCE_14 Section 8.)

#7.3 Seasonality Table

---

#SECTION 8: THE AD OPS DASHBOARD

#8.1 The Three Metrics That Actually Matter

There are 200+ metrics in Meta Ads Manager. Track three, daily:

1. MER (Marketing Efficiency Ratio) = Total revenue ÷ Total ad spend. Your true blended return across all spend. Target >3× for most e-commerce with 30–40% margins. Track in a spreadsheet, not off the platform's self-reported ROAS.
2. CPP (Cost Per Purchase). What it costs to acquire one order. Must stay below your breakeven CAC (Section 8.2) at all times. Track by channel — Meta CPP vs. TikTok CPP vs. Google CPP are not interchangeable numbers.
3. New Customer CPP (NC-CPP). Strip out repeat purchasers. Taylor Holiday's framework: "New customer revenue is the only metric that tells you if your brand is growing." If overall CPP looks healthy but NC-CPP is rising, you're burning your existing list, not growing.

#8.2 Breakeven ROAS — The Number Before Any Campaign Decision

Breakeven ROAS = 1 ÷ Contribution margin % (before ad spend)

Example: 55% contribution margin → breakeven ROAS = 1 ÷ 0.55 ≈ 1.82×. Target ROAS should run at breakeven × 1.3–1.5 to leave room for overhead and reinvestment. Never benchmark a new campaign against the market's blended ~3.7× average — that average includes mature, optimized, branded advertisers. Benchmark against your own margin math. Full contribution-margin bidding mechanics live in LUCE_14 Section 10.1; full MER/attribution dashboard build-out lives in LUCE_06.

Worked example — a $1k lean operator's first week of paid testing: Product retails at $39.99, landed/fulfilled cost $12.10 (per the LUCE_03 landed-cost formula), payment processing ~3%, giving a contribution margin of roughly 65% before ad spend. Breakeven ROAS = 1 ÷ 0.65 ≈ 1.54×. Target ROAS ≈ 2.0–2.3×. Spending $25/day for 5 days ($125 total) on TikTok, the operator sees 4 orders at $39.99 = $159.96 revenue → ROAS = 1.28×. That's below breakeven — per Decision Tree 2, this is a hold-and-refresh signal, not an automatic kill: refresh the hook once, re-test, and judge again at day 7 before spending further.

#8.3 The Weekly Ad Review Process

Monday — Weekly reset: pull last week's spend, revenue, ROAS, CPP by campaign; compare to the prior week and month; flag any campaign where ROAS dropped >20% week-over-week.

Wednesday — Creative review: check learning-phase status; kill creatives with spend >$100 and ROAS below breakeven; identify the week's top 3 creatives by volume and brief 2 variations of each.

Friday — Scaling decisions: scale budgets 20% on anything above target ROAS for 7+ days; pause anything below target for 10+ days; review results from Tuesday's creative launches.

This is the entire job. Elite operators spend roughly 80% of their time on creative and 20% on this review cycle — the ones with the worst results spend most of their time tweaking audiences and budgets inside Ads Manager instead.

---

#DECISION TREES

#Tree 1 — Which platform gets the next dollar?

START: You have $X to spend on paid acquisition this week.

IF you have <$1,000 total remaining AND no organic/affiliate signal yet
  → Do not spend on paid ads. Run Rung 1/2 of the lean tier (Section 7.1) first.

IF you have an organic or affiliate signal (engagement ≥2%, or ≥1 affiliate order)
  → Run a $20–30/day TikTok Spark Ad test on the proven organic hook (5 days).
     IF CTR >1.5% and CPC <$1.50 → advance to a full testing-phase campaign
        (Section 3.3) and consider adding Meta (Section 2.1).
     ELSE → rotate to the next angle before spending further.

IF you're already spending $5k+/month profitably on Meta and/or TikTok
  → Add Google (Section 4.1) once cumulative validated social spend
     exceeds ~$20k and MER is healthy.
  → Add a cheap-inventory channel (Section 5) only after your primary
     two channels hit MER target for 2+ consecutive weeks.

IF CPMs are in the Nov/Q4 spike window (Section 7.3) AND you have
   fewer than 10 proven creative winners
  → Do not cold-test in Q4. Scale known winners only; bank creative
     for Q1 testing instead.

#Tree 2 — Scale, hold, or kill a live campaign

AT day 3 of a new campaign/ad set:
  → Do not judge yet. Noise dominates before day 5–7.

AT day 7:
  IF ROAS ≥ target (Section 8.2 breakeven × 1.3–1.5)
    → Scale budget 20% max, wait 48–72h, re-check.
  IF ROAS is between breakeven and target
    → Hold. Refresh one variable (hook, CTA, or audience) — not all three.
  IF ROAS < breakeven
    → Kill. Duplicate a different winner's structure with a new angle
      instead of resuscitating this ad set.

AT frequency >4 (Meta) or 5–7 days of declining CTR (TikTok):
  → Fatigue. Refresh the hook first (fastest fix); if that doesn't
    recover CTR within 3–4 days, replace the format or the angle.

---

#KPI TABLE — TARGETS, WARNINGS, KILL SWITCHES

---

#THE 2026 REALITY LAYER

Andromeda changed who Advantage+ actually helps. The CPA advantage of Advantage+/Smart+ automation is real but tiered: −38% at $10k+/month, only −14% under $2k/month. If you're a lean operator, don't expect the algorithm to compensate for weak creative — it can't, yet, at your spend level. This is the single most important recalibration versus the original module's "run Advantage+ and mostly stop thinking about it" advice.

Meta's July 2026 retargeting policy change grows your pools, not just your options. The removal of the standalone off-platform activity opt-out (now folded into "Activity from other businesses") means your custom and lookalike audiences likely have more usable signal than they did a year ago. Re-check retargeting audience sizes; don't assume last year's number still holds. Build your tracking stack assuming this kind of policy shift happens again — CAPI-first, not opt-out-dependent.

TikTok's post-JV algorithm is retraining, not broken. Expect real week-to-week reach volatility through the rest of 2026 as the US-only data environment settles. Extend your judgment window on TikTok tests to a full 5–7 days minimum, longer than you'd wait on Meta.

TikTok Shop's July 2026 rulebook (AHR, recalculated Store Rating, 60-day After-sales Handling Time, GMV Max consolidation) means supplier reliability is now an ads problem, not just an ops problem. A slow 3PL doesn't just cost you returns anymore — it drags your Account Health Rating, which affects your reach and affiliate eligibility.

Google killed Video Action Campaigns; Demand Gen is the only conversion-optimized video type left. If your Google video strategy still references VAC, it's stale — rebuild it in Demand Gen, and use YouTube Shorts placement for the cheapest reach while you build a video creative bank.

PMax alone is not the consensus anymore — PMax + Standard Shopping hybrid is. The original module's "PMax is the only way to run Shopping" advice cost operators visibility into which queries actually converted. Run both.

Threads' cheap-CPC window (~$0.68) is an early-mover opportunity, not a permanent baseline. Test it now, while it's under Meta's other placements — expect it to normalize toward Reels/Feed pricing as more advertisers arrive.

---

#FAILURE MODES

---

#SOPs & CADENCES

Daily (15–20 min):

• Check MER, CPP, NC-CPP vs. target (Section 8.1).
• Scan for account issues: disapprovals, policy flags, AHR changes (TikTok Shop sellers).
• Confirm yesterday's revenue in Shopify/GA4 matches platform-reported spend roughly.

Weekly — the Monday/Wednesday/Friday cycle (Section 8.3):

• Monday: pull last week's numbers, flag >20% ROAS drops.
• Wednesday: creative review, kill underperformers, brief next batch (Section 6.4's weekly hook-matrix session fits here).
• Friday: scale winners 20%, pause laggards, review new creative test results.

Monthly:

• Recompute breakeven ROAS (Section 8.2) against current margins — landed costs and platform fees drift.
• Review channel mix against the budget ladder stage you're actually in (Section 7).
• Re-check retargeting audience sizes and segmentation (2026 Reality Layer note on the opt-out change).
• If running TikTok Shop, review AHR and Store Rating trend, not just sales.

Quarterly:

• Reassess whether you're ready to add a channel (decision framework in LUCE_14 Section 7.3).
• Rebuild the creative bank ahead of the next seasonality shift (Section 7.3).

---

#WEEK-1 ACTION PLAN

1. Day 1: Determine your budget tier (Section 7.1 lean vs. 7.2 paid-first) honestly, based on actual capital, not aspiration. Set up Pixel + CAPI (Section 2.5) regardless of tier — you need signal from day one.
2. Day 2: Build your hook matrix (Section 6.4) from your top 3 product angles. Draft 9–12 hook lines and one full creative brief (Section 2.3).
3. Day 3: If lean tier: post your first organic content batch (3–5 videos) using the hooks from Day 2. If paid-first tier: launch a testing-phase Meta campaign (Section 2.1) with $50–100/day CBO.
4. Day 4: Set up TikTok Shop seller account if not already live; seed your first 5–10 affiliate outreach messages (commission 15–25%).
5. Day 5: Review early signal — organic engagement rate, or CTR/CPC on the paid test. Do not scale yet; keep collecting data.
6. Day 6: Draft the next batch of creative variants using variant batching (Section 6.4) from whatever content performed best so far.
7. Day 7: Run the Monday weekly-review process a week early — set your baseline numbers for MER, CPP, and whichever platform metrics apply, so next week's review has something to compare against.

---

#SELF-TEST

1. Advantage+/Andromeda's CPA advantage is −38% for accounts spending $10k+/month. What is it for accounts spending under $2,000/month, and what does that number imply about where a lean operator's effort should go?
2. Your product runs a 55% contribution margin before ad spend. What is your breakeven ROAS, and what target ROAS should you actually aim for?
3. Name the three metrics tracked daily in Section 8.1, and explain why NC-CPP can diverge from blended CPP even when blended ROAS looks healthy.
4. What changed about Google's video campaign options in April 2025, and what's the one conversion-optimized video campaign type left?
5. A TikTok Shop seller's reach and affiliate approvals drop suddenly in August 2026 with no change in return rate. What's the first thing to check, and why might the old "check my violation points" instinct fail here?

<details> <summary>Answers</summary>

1. −14% under $2,000/month. It implies the algorithm can't yet compensate for weak creative at that spend level — effort should go into creative quality and volume (Section 2.2, 6.4), not into trusting automation to find efficiency it doesn't have the data to find.
2. Breakeven ROAS = 1 ÷ 0.55 ≈ 1.82×. Target ROAS should run at breakeven × 1.3–1.5 ≈ 2.4–2.7×, not the market's blended ~3.7× average, which reflects mature branded advertisers.
3. MER, CPP, and NC-CPP. Blended CPP can look healthy while NC-CPP rises if a growing share of "purchases" are repeat customers rather than new ones — the brand isn't actually growing new-customer volume even though the blended number looks fine.
4. Video Action Campaigns (VAC) were killed in April 2025. Demand Gen is now the only conversion-optimized video campaign type on Google.
5. Check Account Health Rating (AHR) in TikTok Seller Center. The old "violation points" system was replaced in July 2026 — AHR is a continuous composite score that reflects fulfillment SLA, After-sales Handling Time, and other factors beyond return rate, so a problem can show up there before it shows up in a traditional return-rate report.

</details>

---

#CROSS-REFERENCES

• → LUCE_03 (Product Selection): feeds this module the winning angle(s) and Rung 4 paid-test results that seed your first campaign structure.
• → LUCE_06 (MER & Measurement): the full attribution and dashboard build-out for MER, CPP, NC-CPP, and blended-channel measurement referenced throughout Section 8 — go there before you build a permanent reporting system.
• → LUCE_05 (Marketing): organic, email, SMS, and influencer channels that feed the same funnel this module's paid acquisition captures at the bottom.
• → LUCE_07 (Brand Building): the destination for creative angles and positioning once a product graduates past pure performance testing.
• → LUCE_08 (Store CRO): the fix when ad metrics (CTR, CPC) look healthy but on-site conversion doesn't follow.
• → LUCE_09 (Finance & Scaling): contribution-margin and cash-flow mechanics behind the breakeven ROAS formula in Section 8.2.
• → LUCE_14 (Advertising Mastery): the advanced twin of this module — go there for contribution-margin bidding, incrementality testing, full cross-platform attribution architecture, YouTube and Pinterest deep dives, media-buying operations, and the Q4/BFCM playbook.

---

LUCE — Launch. Unit Economics. Compound. Exit.

Next: → LUCE_14_Advertising_Mastery.md — the advanced operator's playbook for scaling past $5k/month: contribution-margin bidding, incrementality testing, full attribution architecture, and the complete Q4 scaling system.