MER, Measurement & Attribution

Lineage: upgraded from IDS_06_MER_Measurement.md. Practitioner base: Taylor Holiday (Common Thread Collective) · Cody Plofker · Andrew Faris (4× CFO) · Zachary Murray (Triple Whale) · Northbeam team · Rockerbox team · John Cavendish · Ryan McKenzie (Tru Earth) · Jordan West · Richard Gaffin · Stephanie Liu · Michael Pelpola. Current as of July 2026.

---

#THE ONE-PAGE VERSION

1. Platform ROAS lies. MER doesn't care about attribution models. MER = Total Revenue ÷ Total Ad Spend, across every channel, calculated from your own bank account and Shopify — not from Meta's, TikTok's, or Google's dashboard.
2. July 2026 made platform ROAS lie harder, not softer. Meta folded its standalone off-platform activity opt-out into "Activity from other businesses," growing the identified audiences it can claim credit for. Expect platform-reported ROAS to drift further from reality this year, not converge toward it.
3. TikTok's post-JV algorithm is retraining on US-only data — reach and reported performance are noisier through 2026. Judge TikTok on a 10–14 day window, not a bad Tuesday.
4. Track three MER numbers, not one: daily MER (noisy, informational only), aMER — a 7-day rolling average (your actual decision metric), and NC-MER — new-customer revenue only (your actual growth metric).
5. Breakeven MER = 1 ÷ contribution margin (before ad spend). Target MER = breakeven × 1.3–1.5. Never benchmark against the market's ~3.7× blended average — that number belongs to mature branded advertisers, not you.
6. Section 7's kill/scale table is canonical — LUCE_04, LUCE_09, and LUCE_16 all point back to it. Know your own contribution margin before you touch a budget.
7. You do not need paid attribution under $30k/month. A spreadsheet, GA4 (free), and a one-question post-purchase survey outperform a $129/month tool you don't have the order volume to make useful.
8. Triple Whale (from $129/month) earns its keep at $30k–$50k/month spend (the Section 4.2 hard gate). Northbeam or Rockerbox + Elevar earn theirs at $50k+/month. Buying up the stack early buys confusion, not clarity.
9. Incrementality testing answers the question attribution can't: would this revenue have happened anyway? Geo holdouts and cost-cap "pseudo-experiments" are the two methods a solo or small-team operator can actually run.
10. Cohort analysis is the only way to know if this month's customers are as good as last month's. LTV:CAC below 2× at 6 months means you're slowly destroying business value even if MER looks fine today.
11. The weekly measurement ritual is non-negotiable and short: a 5-minute daily check, a 30-minute weekly review, a 90-minute monthly cohort look. That's the entire job at solo-operator scale.
12. Generic dropship margins (3–7% net) leave almost no room for a marketing budget at all — their breakeven MER often sits at 4–5×. Branded/US-fulfilled margins (15–35% net) breakeven around 1.8–2.2×. Your margin profile decides which row of Section 7 applies to you.
13. The worked $1,000 example in Section 8 is honest, not aspirational — a real first paid week comes in below breakeven, gets refreshed once, and only clears target in week 4. That's a normal 30 days, not a failure.
14. This module cross-references LUCE_04/LUCE_14 (the spend decisions MER math feeds) and LUCE_09 (the contribution-margin and cash mechanics behind the breakeven formula).
15. LUCE_06 ends pointing to LUCE_16 (MEO Framework) — the advanced twin for operators past $10k/month who need channel-level MER allocation, formal incrementality programs, and a full MEO tech stack.

---

#SECTION 1: THE ATTRIBUTION PROBLEM, 2026 EDITION

#1.1 What most operators track — and why it tells you nothing

Here is what most e-commerce operators look at every morning:

• Platform-reported ROAS (Meta says 3.8×, TikTok says 2.9×, Google says 5.2×)
• Revenue in Shopify

Here is what that combination actually tells you: nothing you can act on. Add the three platform ROAS numbers together and you get a picture of your business that's mathematically impossible — you did not generate 300% of your actual revenue.

Four structural reasons this happens:

1. Multi-touch journeys claim triple credit. A customer sees a TikTok ad, sees an Instagram Story two days later, Googles your brand name, clicks a Google ad, and buys. TikTok claims the conversion. Meta claims the conversion. Google claims the conversion. One sale, three claimed wins.
2. Tracking gaps understate the denominator, not the numerator. Server-side tracking (Conversions API) has closed much of the post-iOS14 gap, but even a well-instrumented pixel + CAPI setup sees roughly 80–90% of actual conversions — never 100% (see LUCE_04 Section 2.5). Platforms fill the gap with modeled conversions, which skew toward the platform's own narrative.
3. View-through attribution inflates retargeting. Meta's default attribution window credits a sale if a user merely saw your ad — no click required — within a day. This manufactures false credit for retargeting campaigns aimed at people who were already going to buy.
4. Cross-device journeys break the chain entirely. Customer browses on a phone at lunch, buys on a laptop that night. Most platform pixels cannot stitch that together, so the sale either goes uncredited or gets misattributed to whichever platform happened to fire last.

The result: every platform's dashboard reports a ROAS meaningfully higher than reality, in the platform's own favor. An operator who allocates budget off platform ROAS is making decisions from fiction dressed as data.

#1.2 What changed in July 2026: Meta's off-platform data shift

The original version of this problem just got worse. In July 2026, Meta removed the standalone "off-platform activity" opt-out and folded it into a broader "Activity from other businesses" setting. The mechanical effect: Meta's identifiable retargeting and custom-audience pools are now larger than they were a year ago, because Meta can connect more of a user's off-Meta browsing and purchase behavior back to their Meta identity by default.

Larger identified pools mean two things for your reporting, and they point in the same dangerous direction:

• More conversions get attributed to Meta, including view-through conversions from people who were always going to buy — the exact mechanism described in Section 1.1, now operating on a bigger dataset.
• Retargeting audience sizes may have grown without you changing anything. If you haven't re-checked your 90-day retargeting pool size recently, do it before you trust last year's segmentation (see LUCE_04 Section 2.5 for the audience-architecture fix).

The practical instruction: treat 2026 platform ROAS as more inflated than 2025 platform ROAS, not less. MER — which by definition ignores which platform claims what — becomes more valuable precisely because the platforms' own numbers are drifting further from ground truth.

#1.3 What changed in 2026: TikTok's post-JV volatility

The TikTok USDS joint venture (Oracle/Silver Lake/MGX, ~45–50% ownership) closed January 22–23, 2026, ending the ban-risk overhang that shadowed every 2025 TikTok strategy. But the platform's US algorithm is now retraining on a US-only data environment, and reach and reported performance have been measurably noisier through 2026 than pre-divestiture.

For measurement, this means: don't read a bad TikTok week as a broken channel, and don't read a great TikTok week as the new baseline. Extend your judgment window on TikTok to 10–14 days, roughly double what you'd use on Meta, before drawing a conclusion from either platform-reported ROAS or your own channel-level MER contribution.

#1.4 The fix: MER as ground truth

Taylor Holiday (CEO, Common Thread Collective, managing $100M+ in ad spend):

"MER is the most important number in your business. Everything else is a supporting metric."

MER works because it structurally ignores every failure mode in Section 1.1:

• It doesn't care about attribution models.
• It doesn't care about view-through windows.
• It doesn't care which platform claims what.
• It captures everything — direct, organic assisted by ads, email revenue from ad-acquired customers.

It is not a perfect number. It cannot tell you which channel is working (Section 4 covers that). But it cannot lie to you about whether your marketing, in aggregate, is efficient.

---

#SECTION 2: THE MER FRAMEWORK — DEFINITIONS AND FORMULAS

#2.1 MER — the master metric

MER = Total Revenue ÷ Total Ad Spend (all channels combined)

Also called "blended ROAS" or "true ROAS." The difference between platform ROAS and MER, illustrated with representative mid-2026 numbers:

MER of 3.4× does not mean you're profitable. MER tells you the efficiency of spend relative to revenue; profitability depends on your margin structure. A store with 25% contribution margin and MER 3.4× is thriving. A store with 12% contribution margin and MER 3.4× is losing money on every ad dollar. Section 2.6 and Section 7 give you the exact math.

#2.2 aMER — the rolling average that removes the noise

Daily MER is real but useless for decisions on its own — it swings on single-day events that have nothing to do with underlying efficiency.

aMER (Average MER) = Total Revenue ÷ Total Ad Spend, over a rolling window
                      (7-day rolling is standard; use 30-day rolling for
                      strategic, month-level calls)

Two distortions to flag, not react to:

• Email/SMS send days distort daily MER upward. A blast drives a revenue spike with zero ad spend attached to it — great for the day's number, meaningless for judging your paid channels.
• Sale days distort daily MER upward for a different reason — discounted AOV plus a demand spike inflates the ratio temporarily.

DATE  | TOTAL REVENUE | META SPEND | TIKTOK SPEND | GOOGLE SPEND | TOTAL SPEND | DAILY MER | NOTE
6/01  | $8,240         | $1,500     | $500          | $300          | $2,300      | 3.58×     |
6/02  | $6,180         | $1,500     | $500          | $300          | $2,300      | 2.69×     | Email send
6/03  | $10,450        | $1,500     | $500          | $300          | $2,300      | 4.54×     |
7-day aMER (6/01–6/07): 3.51× ← this is the number you act on

Rule: never adjust spend off a single day's MER. React to aMER trend across the canonical table in Section 7.

#2.3 NC-MER — the growth indicator

Regular MER includes repeat purchasers — people who were already going to buy from you regardless of this week's ad spend. That inflates MER without telling you anything about whether the business is growing.

NC-MER (New Customer MER) = New Customer Revenue ÷ Total Ad Spend

If MER is 3.5× but NC-MER is 1.8×, you are: spending heavily to reach existing customers who'd have bought anyway, running retargeting too aggressively, and not acquiring new customers efficiently. Your growth is decelerating even though the headline number looks fine.

Taylor Holiday's NC-MER benchmark for growing brands: >2× in year 1, >2.5× in year 2+.

How to track it:

• Klaviyo segment: "Customer placed first order > 365 days ago" = existing customer; everyone else = new.
• Shopify: Reports → Customers → New vs. Returning.
• Triple Whale (if you're past the $30k/month gate in Section 4.2): tracks natively in the Customer dashboard.

#2.4 Contribution margin and breakeven MER

MER is meaningless without knowing what fraction of revenue survives after product, fulfillment, and platform costs — before ad spend even enters the picture.

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

Worked P&L (landed-cost aware, per the 2026 fact sheet):

Revenue:                                    $100,000  (100%)
- COGS (factory price + duty on invoice):    $16,000   (16%)
- Domestic 3PL pick/pack + last-mile:         $9,000   (9%)
- Payment processing (~3%):                   $3,000   (3%)
= Contribution margin before ad spend:       $72,000   (72%)

- Ad spend:                                  $28,000   (28%)  → MER = 3.57×
= Contribution margin after ads:             $44,000   (44%)

- Returns/chargebacks:                        $2,000   (2%)
- Overhead (tools, contractors, owner draw): $10,000   (10%)
= Operating profit:                          $32,000   (32%)

Breakeven MER = Revenue ÷ (Revenue − COGS − Fulfillment − Payment processing − Returns)
              = $100,000 ÷ ($100,000 − $16,000 − $9,000 − $3,000 − $2,000)
              = $100,000 ÷ $70,000 = 1.43×

Target MER = Breakeven MER × 1.5 = 2.14× (for a comfortable cushion)
At MER 3.57×: comfortably profitable, well past target.

This is the formula the whole course points back to. LUCE_04 and LUCE_14 use it at the campaign level; LUCE_09 uses it at the P&L level; LUCE_16 uses it at the channel-allocation level. Section 7 turns it into a table you can use without redoing the algebra every time.

#2.5 Putting the three numbers side by side

---

#SECTION 3: THE WEEKLY MEASUREMENT RITUAL

This is the entire measurement job at solo-operator scale — three checkpoints, none of them long.

#3.1 The spreadsheet — exact columns

Build this once in Google Sheets. Every subsequent number in this module reads from it.

DATE | TOTAL REVENUE | NEW CUSTOMER REVENUE | RETURNING CUSTOMER REVENUE |
META SPEND | TIKTOK SPEND | GOOGLE SPEND | OTHER SPEND | TOTAL AD SPEND |
EMAIL/SMS SEND (Y/N) | SALE DAY (Y/N) | DAILY MER | 7-DAY aMER | NC-MER |
ORDERS | RETURNS $ | NOTES

Populate DATE through TOTAL AD SPEND daily from Shopify and each ad platform. Formula columns (DAILY MER, 7-DAY aMER, NC-MER) calculate automatically. Flag EMAIL/SMS SEND and SALE DAY manually — this is the discipline that keeps you from reacting to noise (Section 2.2).

#3.2 Daily — 5 minutes

• Pull yesterday's total revenue and total ad spend; log in the sheet.
• Calculate daily MER; glance at it, don't act on it.
• Scan for anomalies: revenue spike or drop >20% versus the 7-day average.
• Confirm Shopify revenue roughly matches what each platform reports as spend (catch billing errors early).

#3.3 Weekly — 30 minutes

• aMER (7-day rolling) vs. your target (Section 7 table).
• NC-MER vs. your target (Section 2.3).
• Cost per purchase by channel — Meta CPP, TikTok CPP, Google CPP are not interchangeable numbers (see LUCE_04 Section 8.1).
• Top 3 creatives by spend and platform-reported ROAS (directional only).
• Email open rate + revenue per send.
• Return rate and chargeback rate for the week.

#3.4 Monthly — 90 minutes

• Update the cohort revenue table (Section 6.2).
• LTV:CAC by acquisition cohort (Section 6.3).
• P&L vs. budget — recompute contribution margin; landed costs and platform fees drift monthly, and a stale contribution margin number quietly corrupts every breakeven-MER calculation downstream.
• CAC payback period.
• Channel mix shift — are you more or less reliant on any single channel this month?
• New customer cost trend — rising, flat, or falling over the trailing 90 days?
• If running TikTok Shop: review Account Health Rating and Store Rating trend, not just sales (see LUCE_04 Section 3.4).

#3.5 Red flags — when to act immediately, not on the next scheduled check

Most of Section 3's rhythm is deliberately unhurried — daily numbers are informational, weekly and monthly reviews carry the real decisions. A short list of exceptions overrides the schedule:

Stop and investigate the same day if:

• aMER drops >25% week-over-week (algorithm change, creative fatigue, tracking break, or a platform policy shift — Section 1.2's kind of event).
• Return rate spikes above 15% in any 7-day window.
• Chargeback rate exceeds 0.8% in any month.
• Email deliverability drops below 85% — stop sending immediately; a damaged domain reputation compounds the longer you keep mailing through it.
• An ad account gets restricted with no backup account ready to activate.

Flag for the next scheduled review, not an emergency, if:

• NC-MER is declining but hasn't hit the 3-consecutive-week threshold in the KPI table yet.
• CAC is rising but under the 20%-over-30-days threshold.
• Email click rate has dipped under 1% for one send (check the next one before concluding a trend).

The distinction matters: treating every yellow flag as a same-day fire drill burns the exact attention this module is trying to protect — the weekly creative and budget work in LUCE_04/LUCE_14 that actually moves MER.

---

#SECTION 4: ATTRIBUTION TOOLS AND THE MEASUREMENT STACK LADDER

#4.1 The landscape

First-party (your own data, free):

• Google Analytics 4 — data-model-based attribution, free, limited DTC-specific utility but solid for traffic-source and journey analysis.
• Shopify Analytics — basic, last-click only; fine for revenue and new-vs-returning, not for channel attribution.

Third-party attribution tools:

• Triple Whale (from $129/month) — most widely used DTC attribution tool under $50k/month spend. Native Shopify integration, Pixel + server-side tracking, strong UI, native post-purchase survey ("Sonar").
• Northbeam — stronger multi-channel ML attribution modeling, higher learning curve; built for $50k+/month spend.
• Rockerbox — mid-tier, multi-touch attribution (MTA) focus; good for brands running Google/Meta/TikTok/TV together.
• Elevar — not an attribution tool itself; a server-side tracking layer that sits underneath the others and improves the data quality feeding into all of them, particularly for Meta's pixel post-iOS14.

#4.2 The ladder — including the gate most operators skip

You do not need paid attribution under $30,000/month in spend. This is a hard gate, not a soft suggestion. Below that threshold, three things are true simultaneously: your order volume is too thin for a tool's ML/segmentation to find a reliable signal, a wrong budget call at your spend level costs less than the tool's monthly fee times twelve, and the spreadsheet in Section 3.1 already answers every question you actually need answered. Buying Triple Whale at $2k/month spend doesn't buy clarity — it buys a second dashboard to be confused by.

#4.3 The post-purchase survey — the cheapest attribution tool you'll ever run

Triple Whale, Northbeam, and standalone tools (Fairing, Enquire Labs, or a plain Typeform) can all run a "How did you hear about us?" question on the order confirmation or thank-you page. It costs $0–$50/month and produces the most accurate attribution data available: self-reported, unfiltered by any platform's algorithm.

Survey options to include: TikTok (ad or organic) · Instagram (ad or organic) · Facebook (ad or organic) · Google search · A friend/referral · YouTube · Podcast/press · I'm a returning customer · Other.

What to do with the data: build a channel contribution index comparing PPS % against each platform's self-claimed %.

• PPS % > platform-claimed % → the platform is under-attributing its real impact (common for TikTok organic, podcasts, and word-of-mouth — none of which show up cleanly in any dashboard).
• PPS % < platform-claimed % → the platform is over-claiming, usually by capturing credit for demand another channel created (classic Google brand-search behavior — see LUCE_16 Section 5.2).

Industry benchmarks: Meta over-claims by 30–60%; Google over-claims by 20–40%. Treat these as directional, not gospel — run your own survey and compute your own gap.

#4.4 Attribution frameworks, briefly

You will encounter these terms in every tool's settings menu. Know what they mean and, more importantly, what they're for.

The operator's approach: use MER for budget-level decisions (how much to spend overall — Section 7). Use your attribution tool's Pixel model, once you're past the $30k gate, for channel-level decisions. Use the post-purchase survey to reality-check platform over-attribution at any spend level. Use cohort analysis (Section 6) for actual customer LTV. No single model replaces this combination.

#4.5 Setting up Triple Whale, once you cross the gate

When spend clears $30,000/month (Section 4.2), the setup takes under an hour:

1. triplewhale.com → Connect Shopify (one-click OAuth).
2. Connect Meta Ads, TikTok Ads, and Google Ads accounts.
3. Install the Triple Whale Pixel on Shopify: Settings → Customer Events → Triple Whale.
4. Configure the post-purchase survey ("Sonar") — this is the same "how did you hear about us" logic from Section 4.3, now native to the tool instead of a standalone Typeform.
5. Confirm test events fire before trusting any reported number: Pixel test mode should show PageView, ViewContent, AddToCart, InitiateCheckout, and Purchase events landing correctly.

What operators actually do with it once it's running: compare Triple Whale's Pixel-based channel attribution against each platform's native reporting. The gap between the two is a live, ongoing version of the post-purchase-survey exercise in Section 4.3 — except now it updates daily instead of requiring a manual pull. Use it to rank channels by NC-MER (LUCE_16 Section 4.3 covers the full budget-allocation framework this feeds), not to replace the aMER-based scale/hold/kill decisions in Section 7, which stay spreadsheet-driven regardless of what attribution tool sits on top.

---

#SECTION 5: INCREMENTALITY TESTING

#5.1 The question attribution can't answer

The question that matters in paid advertising is: would these customers have bought even if I hadn't run this ad? Attribution models — even the good ones — can only tell you which ads were seen or clicked. They cannot tell you whether those ads caused the purchase.

Incrementality testing creates a holdout — a control audience that never sees your ads — to measure the causal effect of spend. For many brands, reported ROAS is 3.5× and incremental ROAS is 2.1×; the gap is purchases that would have happened anyway (organic, email, brand search, existing customers).

#5.2 Geo holdout tests

Manual holdout (works on any platform, no vendor tool required):

1. Define your test: run ads in cities/states A, B, C. Withhold spend from city/state D (control) — ideally matched to A/B/C on population, historical revenue, and seasonality.
2. Compare revenue-per-capita in the test region vs. the control region over the same period.
3. If test regions grew meaningfully more than the control → your ads are driving incremental growth.
4. If growth is roughly equal → your spend is not incrementally driving purchases; you're mostly buying demand that existed anyway.

Meta's built-in tool:

1. Ads Manager → Experiments → Conversion Lift.
2. Select the campaign, choose a holdout size (10–20% of audience sees no ads).
3. Run 2–4 weeks minimum.
4. Meta reports incremental purchases and incremental ROAS directly.

#5.3 Cost caps as pseudo-experiments

Formal geo holdouts need scale and patience most solo operators don't have in month one. There's a cheaper, faster substitute that produces a real (if less rigorous) read on incrementality: using a cost cap as a stepped, self-administered experiment.

The logic: a cost cap tells the platform's algorithm the maximum you'll pay per result. As you loosen the cap in controlled steps and watch how MER responds, you're watching the platform reveal the marginal value of additional spend — which is exactly what an incrementality test is trying to measure, just without a formal control group.

The protocol:

1. Run a campaign under a strict cost cap for 2 weeks. Record aMER for that period.
2. Loosen the cost cap ~20% (spend can now rise where the algorithm finds room). Hold for 2 more weeks. Record aMER again.
3. Compare:
   • aMER held steady or improved as spend rose → the additional spend found genuinely incremental demand. Continue loosening in the same 20% steps.
   • aMER declined meaningfully (>10–15%) as spend rose → you've hit the edge of incremental demand; the marginal dollars are buying non-incremental conversions (existing customers, brand search cannibalization, retargeting overlap). Hold the cap where it was.
4. Repeat in 2-week steps until you find the cap level where aMER starts to break — that level is your channel's practical MER-preserving spend ceiling, re-checked monthly since it moves with competition and seasonality.

This is the same underlying logic as the "controlled scale test" described in LUCE_16 Section 4.2 (increase budget 20–25%, hold 2 weeks, read the MER response) — cost caps give you a more granular, lower-risk version of the same experiment, one step at a time instead of one big jump.

#5.4 Interpreting incrementality results

• Incremental ROAS < 1.5×: the channel is not efficiently driving new purchases. Reduce spend, reallocate to a channel with a stronger incrementality read.
• Incremental ROAS > 2.5×: the channel is genuinely driving incremental growth. This is where the next marginal dollar of your budget should go.
• Use the gap to correct your mental model of platform ROAS — if a channel's incremental ROAS runs 40% below its platform-reported ROAS, discount that platform's number by roughly that much every time you glance at Ads Manager going forward.

---

#SECTION 6: COHORT ANALYSIS — UNDERSTANDING YOUR BUSINESS OVER TIME

#6.1 What a cohort is, and why most operators skip this

A cohort = the group of customers who made their first purchase in the same time period (usually monthly). Cohort analysis asks: "Of the customers who first bought in January, how much have they spent, total, by June?"

This is the only way to answer four questions MER cannot:

• What is true LTV, tracked over time rather than assumed from AOV?
• Is LTV improving or declining cohort over cohort?
• Are new cohorts as good as old cohorts — critical before you scale spend?
• What's the actual payback period on customer acquisition cost?

#6.2 The cohort revenue table

Build this monthly in Google Sheets, pulling from Shopify Reports → Customers.

COHORT TABLE (cumulative revenue per cohort, by months since first purchase)

           Month 1   Month 2   Month 3   Month 6   Month 12
Jan cohort $45,000   $12,000   $8,000    $5,000    $3,000
Feb cohort $51,000   $14,500   $9,200    (pending)
Mar cohort $48,000   $13,000   (pending)
Apr cohort $52,000   (pending)

Jan cohort 12-month cumulative LTV = $45,000+$12,000+$8,000+$5,000+$3,000 = $73,000
(divide by number of customers in the cohort = LTV per customer)

What to look for:

• Month 2 revenue as % of Month 1 = your 30-day repeat rate. Target: 20–40% depending on category.
• Month 6 total as % of Month 1 = your 180-day LTV multiple. Target: 1.3–2.0×.
• Month 2/Month 1 ratio falling over successive cohorts → your product or post-purchase experience is degrading.
• February cohort LTV trending below January cohort LTV → the customers you're currently acquiring are lower quality, usually a targeting or offer problem, not a product problem.

#6.3 LTV:CAC — the exit multiplier

LTV = lifetime revenue per customer, across all purchases
CAC = Total ad spend ÷ New customers acquired

CAC payback period:

CAC Payback Period = CAC ÷ (Monthly revenue per customer × Contribution margin %)

Example: CAC = $45, monthly avg spend after acquisition = $25, CM = 40%
Payback = $45 ÷ ($25 × 0.40) = $45 ÷ $10 = 4.5 months

Target: under 6 months for sustainable growth; under 3 months is exceptional.

Healthy LTV:CAC progression by cohort age (from an aggregate view across scaled 7–9 figure operators):

If your cohort isn't tracking toward these bands by the relevant age, you have a retention problem, not (yet) an acquisition problem — see LUCE_09 for the cash-flow implications of a slow-maturing cohort.

---

#SECTION 7: THE CANONICAL KILL/SCALE DECISION TABLE

This table is the reference point for every other LUCE module that touches spend decisions (LUCE_04, LUCE_09, LUCE_14, LUCE_16). Find your contribution margin, read your breakeven and target MER, and use the signals as-is — don't improvise thresholds mid-campaign.

Scale signal: aMER (7-day rolling) at or above your target for 2 consecutive weeks → increase spend 15–20%, then re-measure for another 2 weeks before scaling again.

Hold signal: aMER sitting between breakeven and target → maintain spend, diagnose (creative fatigue, seasonality, tracking break) before touching budget in either direction.

Kill signal: aMER below breakeven for 7+ consecutive days, after ruling out a tracking break or an attribution-inflation event (Section 1.2) — reduce spend or pause the underperforming channel/campaign.

Read the fact sheet into this table honestly. Generic dropshipping in 2026 runs 3–7% net margins, which typically maps to a 15–25% contribution margin before ads — a breakeven MER of 4–6×. That's why so many dropship stores fail even with a working product: they never had enough margin to fund a marketing budget in the first place. Branded, US-fulfilled operators running 15–35% net margins typically sit in the 45–55% contribution-margin band — a breakeven MER of 1.8–2.2×, an entirely different game with far more room for error. Know which game you're playing before you read a "3.7× MER" headline and assume it applies to you.

---

#SECTION 8: WORKED EXAMPLE — A $1,000-BUDGET OPERATOR'S FIRST 30 DAYS

This continues the worked example from LUCE_04 Section 8.2: a product retailing at $39.99, landed/fulfilled cost $12.10 (duty + freight + 3PL + last-mile, per the 2026 fact sheet), payment processing ~3%, giving a contribution margin of roughly 65% before ad spend. From Section 7: breakeven MER ≈ 1.54×, target MER ≈ 2.0–2.3×.

The operator follows LUCE_04's lean tier (Section 7.1): organic-first, no paid spend until an organic or affiliate signal appears.

Reading the 30-day blended MER (2.40×) against the two paid-only periods (1.28×, then 1.92× and 1.87×) is the actual lesson here. The blended number looks comfortably above both breakeven and target — but that's partly because Weeks 1–2's organic and affiliate revenue count in MER's numerator with zero spend in the denominator, exactly as MER is designed to work (Section 1.4). Judge campaign decisions on the paid-period numbers using LUCE_04's Tree 2 rules. Judge the business's 30-day health on the full blended MER. They will diverge in month one — that divergence is diagnostic, not an error.

Honest verdict for month 2: the proof-of-concept passes (blended MER clears both thresholds), but the paid-acquisition engine on its own is only just clearing breakeven. The operator's job in month 2 is to prove the paid channel can independently hold ≥2.0× before scaling budget — not to celebrate the blended number and increase spend prematurely. Since this is an early-stage store with no repeat purchasers yet, NC-MER ≈ MER for this period; that will diverge once returning-customer revenue appears (Section 2.3).

---

#DECISION TREES

#Tree 1 — Is my MER healthy right now?

START: You have a current 7-day aMER reading.

IF aMER ≥ target MER (Section 7 table) for 2 CONSECUTIVE WEEKS
  → Scale spend 15–20%. Re-measure for 2 more weeks before scaling again.

IF aMER is between breakeven MER and target MER
  → Hold spend. Diagnose before touching budget:
     - Creative fatigue? (check frequency/CTR trend, LUCE_04 Section 6.3)
     - Seasonality? (check against LUCE_04 Section 7.3 CPM table)
     - Tracking break? (confirm Pixel + CAPI still firing, LUCE_04 2.5)

IF aMER is below breakeven MER for FEWER than 7 consecutive days
  → Do not act yet. This is noise (Section 2.2). Keep monitoring daily.

IF aMER is below breakeven MER for 7+ CONSECUTIVE DAYS
  → Rule out an attribution-inflation event first (Section 1.2 — did a
     retargeting pool change or platform policy shift distort the reading?).
  → If ruled out: reduce spend on the underperforming channel/campaign,
     or kill it, per LUCE_04 Tree 2.

IF NC-MER is declining 3+ consecutive weeks WHILE overall MER holds steady
  → Do not treat this as healthy. Your acquisition engine is weakening
     even though the blended number looks fine. Audit new-customer
     targeting and creative before the overall number follows it down.

#Tree 2 — Which measurement tool do I actually need right now?

START: What is your current monthly ad spend?

IF spend < $10,000/month
  → Spreadsheet (Section 3.1) + GA4 (free) + post-purchase survey
     (Section 4.3). Do not buy a paid attribution tool. Full stop.

IF spend is $10,000–$30,000/month
  → Same stack. This is the range operators buy tools too early.
     If you genuinely can't answer "which channel is working" from the
     spreadsheet + PPS combination after 60 days of disciplined tracking,
     the problem is usually the discipline, not the tooling.

IF spend is $30,000–$50,000/month
  → Add Triple Whale (from $129/month). Order volume now supports its
     channel-level modeling.

IF spend is $50,000+/month
  → Northbeam or Rockerbox + Elevar. See LUCE_16 Section 7 for the full
     build-out at this scale.

IF you're running formal incrementality tests (Section 5) at ANY spend level
  → A paid attribution tool is optional; a geo holdout or cost-cap
     pseudo-experiment (Section 5.2–5.3) answers the incrementality
     question directly, tool or no tool.

---

#KPI TABLE — TARGETS, WARNINGS, KILL SWITCHES

---

#THE 2026 REALITY LAYER

Meta's July 2026 off-platform data change makes platform ROAS less trustworthy, not more. Folding the standalone opt-out into "Activity from other businesses" grows Meta's identifiable retargeting pools — expect reported ROAS to drift further from your actual incrementality this year. Re-check retargeting audience sizes; don't assume last year's segmentation still holds (see LUCE_04 Section 2.5 for the audience-architecture fix).

TikTok's post-JV algorithm is retraining, not broken. Extend judgment windows on TikTok to 10–14 days. Reading a volatile week as a dead channel is one of the most common measurement mistakes in 2026.

Pixel + CAPI still tops out around 80–90% conversion visibility, never 100%. This is the structural reason MER — which needs no pixel data at all — remains the ground-truth metric even as tracking infrastructure improves.

Triple Whale's entry price (from $129/month) makes it tempting to buy early. Resist it below $30k/month spend (Section 4.2). The tool isn't the bottleneck at that stage; the weekly ritual (Section 3) is.

Generic dropship margin compression (3–7% net) is a measurement problem as much as a strategy problem. At those margins, the breakeven MER in Section 7 sits at 4–6× — a bar most cold-traffic testing simply cannot clear. If your contribution margin doesn't support a breakeven MER under 3×, the fix is margin (landed cost, pricing, fulfillment — LUCE_09), not a better attribution tool.

Cost caps as pseudo-experiments (Section 5.3) are a genuinely new-for-2026 addition to the incrementality toolkit — not because the tactic is new, but because Advantage+/Andromeda-era automation makes cost-cap responses a cleaner signal than they were under manual bidding; the algorithm's reaction to a looser cap is now closer to a real demand-elasticity read than it was in 2022.

---

#FAILURE MODES

---

#SOPs & CADENCES

Daily (5 minutes) — Section 3.2:

• Log revenue, spend, daily MER.
• Scan for >20% anomalies.
• Confirm Shopify and platform numbers roughly reconcile.

Weekly (30 minutes) — Section 3.3:

• aMER and NC-MER vs. target.
• Cost per purchase by channel.
• Top creatives, email metrics, returns/chargeback rate.

Monthly (90 minutes) — Section 3.4:

• Cohort table update, LTV:CAC by cohort.
• P&L vs. budget; recompute contribution margin.
• CAC payback period, channel mix shift, new-customer cost trend.
• Re-check retargeting audience sizes (2026 Reality Layer).
• TikTok Shop sellers: AHR and Store Rating trend.

Quarterly:

• Recompute your Section 7 row (contribution margin band) — margin structure shifts with pricing, supplier terms, and shipping cost changes; don't run on a stale row.
• Re-evaluate your position on the Section 4.2 tool ladder as spend grows.
• Run (or re-run) a formal incrementality test on any channel now carrying >20% of spend (Section 5).

---

#WEEK-1 ACTION PLAN

1. Day 1: Build the MER spreadsheet with the exact columns from Section 3.1. Backfill the last 30 days if the data exists.
2. Day 2: Compute your contribution margin % before ads, pulling landed-cost figures from your product-selection or finance work (LUCE_03/LUCE_09). Find your row in the Section 7 table; write down your breakeven MER and target MER.
3. Day 3: Set up GA4 if it isn't already installed. Confirm Shopify's new-vs-returning customer report is working — you need it for NC-MER.
4. Day 4: Install a free/low-cost post-purchase survey (Typeform, Fairing free tier) with the single "How did you hear about us?" question (Section 4.3).
5. Day 5: Run your first daily 5-minute check. Log the numbers.
6. Day 6: Run your first weekly 30-minute review, even on a partial week — this establishes your baseline aMER.
7. Day 7: Confirm honestly which rung of the Section 4.2 ladder you're actually on. If you're under $30k/month (nearly every reader of this module is), commit to spreadsheet-only. Do not buy a paid attribution tool this week.

---

#SELF-TEST

1. Your contribution margin before ad spend is 50%. What is your breakeven MER, and what target MER should you aim for (using a 1.3–1.5× cushion)?
2. Tuesday's daily MER came in at 1.1× because of an email send that day. What should you actually act on, and why?
3. Your post-purchase survey shows 40% of customers say Meta influenced their purchase, but Meta's dashboard claims 55% of orders. What do you do with that 15-point gap?
4. What is the mechanism by which Meta's July 2026 removal of the off-platform activity opt-out risks inflating reported ROAS?
5. You're spending $8,000/month. Should you buy Triple Whale? Why or why not?

<details> <summary>Answers</summary>

1. Breakeven MER = 1 ÷ 0.50 = 2.00×. Target MER = 2.00 × 1.3–1.5 = 2.6–3.0×.
2. Nothing, off that single number. Flag the email-send day in your spreadsheet (Section 3.1's rules) and read the 7-day aMER instead — daily MER is noisy and informational only; email/sale days distort it in predictable, non-actionable ways (Section 2.2).
3. Treat the 15-point gap as an attribution-inflation signal — weight Meta's reported ROAS down by roughly that amount going forward. Don't cut Meta budget off that number alone; check whether aMER has actually declined, and consider a geo holdout or cost-cap pseudo-experiment (Section 5) before reallocating spend based on the survey gap alone.
4. Folding the standalone off-platform opt-out into "Activity from other businesses" grows the audience Meta can identify and claim credit for via retargeting and view-through attribution — more purchases get claimed by Meta even when the ad didn't cause them. Trust aMER over platform ROAS more than you did before this change (Section 1.2).
5. No. At $8,000/month you're under the $30,000/month gate (Section 4.2) — your order volume doesn't yet support Triple Whale's channel-level modeling, and the spreadsheet + GA4 + post-purchase survey stack already answers the questions you need answered at this spend level.

</details>

---

#CROSS-REFERENCES

• → LUCE_04 (Advertising) / LUCE_14 (Advertising Mastery): the spend decisions — testing budgets, scaling rules, breakeven ROAS at the campaign level — that this module's MER math feeds and disciplines.
• → LUCE_03 (Product Selection) / LUCE_13 (Product Selection Science): the landed-cost inputs (duty, freight, 3PL, last-mile) behind the contribution-margin calculations in Section 2.4 and Section 7.
• → LUCE_09 (Finance & Scaling): the full contribution-margin, cash-flow, and P&L mechanics behind the breakeven-MER formula; where a declining CAC payback period (Section 6.3) becomes a cash-runway problem.
• → LUCE_16 (MEO Framework): the advanced twin — channel-level MER allocation, formal incrementality programs, and the full MEO tech stack for operators past $10k/month.

---

LUCE — Launch. Unit Economics. Compound. Exit.

Next: → LUCE_16_MEO_Framework.md — the advanced operator's system for scaling past $10k/month with channel-level MER allocation, formal incrementality testing, and a full measurement-infrastructure build-out.