Find out before you spend a dollar. Green Light runs your own margin against what a customer costs in your category on Meta in Canada, then gives you one of three answers: not yet, spend to learn, or spend to scale. 15 categories across 7 families, each measured against its own benchmarks. The margin math is free with no email gate. The verdict is $47.
Four questions. You get your real first order contribution, the ROAS you have to beat, and what a customer costs on Meta in Canada in your own category, not a blended CPG average.
And the gap between families is not small. A supplement customer costs roughly twice what a food customer costs, and the margin each family is held to differs just as sharply.
Supplements are observed at $110 to $179 a customer, food at $62 to $73. Same country, same platform, different business.
Against an ecommerce band of $19 to $30 (AdLibrary 2026). Cosmetics buy the most expensive impressions in CPG, which is the trade for carrying the best margins in it.
The best purchase frequency in consumables, against 4 to 6 for household staples on the same plan.
Every CPG benchmark you can find online is American, and most of the American ones are for food. That is how a skincare brand ends up planning a budget against snack economics. The cost of a customer we model runs from $29 at the low end in food and beverage to $179 at the top end in supplements and natural health, so one blended CPG number would be wrong for almost everybody who reads it, and wrong in the direction that hurts: it would tell you the test costs less than it does.
Margin works the same way. A food brand is held to 35% first order contribution and a beauty brand to 45%, because beauty carries 58% to 68% gross margin where food carries 35% to 48%. A skincare brand that clears the food number and stops there is not winning, it is behind its own peers. So the report holds your numbers against your family and says which family it used. Here is the whole set.
| Family | Cost per customer | First order contribution | Repeat rate | Payback | Benchmarks |
|---|---|---|---|---|---|
| Food and beverageSnacks, Beverages, Coffee and tea, Pantry and condiments, Frozen | $29 to $73 | 35% target22% floor | 30% to 40% | 1 to 3 months | Published per vertical |
| Supplements and natural healthSupplements | $85 to $179 | 45% target28% floor | 30% to 40% | 3 to 6 months | Published per vertical |
| Beauty and personal careSkincare, Haircare, Colour cosmetics, Personal care and hygiene | $62 to $160 | 45% target28% floor | 21% to 35% | 2 to 4 months | Published per vertical |
| Household and cleaningCleaning and laundry, Paper and household staples | $73 to $131 | 35% target22% floor | 22% to 32% | 3 to 6 months | Derived, and the report says so |
| Baby and kidsBaby and kids | $69 to $145 | 38% target24% floor | 30% to 42% | 2 to 5 months | Derived, and the report says so |
| PetPet | $53 to $124 | 35% target22% floor | 25% to 40% | 2 to 4 months | Published per vertical |
| Other CPGOther CPG | $76 to $161 | 35% target22% floor | 18% to 28% | 3 to 6 months | Derived, and the report says so |
Margin and retention are vetoes. A brand can score well on creative, tracking and budget and still get told to wait, because no amount of good creative fixes an economics problem.
Your unit economics cannot carry a paid customer yet. You get a fix list in priority order instead of a media plan, with the margin your own category needs to hit before ads make sense.
Most brands land here the first time they run it. That is a sequence, not a rejection, and it is cheaper to find out now.
You can fund a capped test, not a scale. For 90 days the job is finding out what your account pays per purchase instead of what a benchmark says it should.
The report writes the go or no go rule before any money moves, so the decision at the end of the window is already agreed.
Your numbers hold at volume. Margin and retention both clear, which is the profile where more spend buys more customers instead of more losses.
The risk shifts to creative volume and cost per purchase drift, so that is what the plan is built around.
Every one measured against your own family's benchmark, not a blended CPG average.
| Gate | Points | Measured against |
|---|---|---|
| Contribution margin | 25 | Your own family's first-order contribution target, 35% to 45% depending on which one you are in, each with a floor underneath it |
| Retention | 25 | The repeat purchase band for your own family, from 18% to 42% across the 7, against a measured ecommerce average of 18.8% |
| Proof and creative | 20 | 25 reviews, 10 vertical videos, 2 to 3 months of content backlog |
| Tracking and email | 15 | Pixel plus Conversions API, GA4, and the three flows that carry repeat revenue |
| Budget | 15 | Clears the floor that buys 30 purchases a month at your own modeled cost |
Answering not sure on a tracking question scores the same as answering no, and the report says so out loud rather than quietly marking you down. An unverified pixel is an unusable pixel, and that is a finding worth having.
One instruction, with the arithmetic that produced it and the package it routes to.
Every cost line on one order, your contribution in dollars and percent, your breakeven ROAS, and the most you can pay for a customer over 12 months.
Out of 100, each with your family's benchmark, your number, why the gate decides anything, and the one fix that moves it most.
The least you can spend for the test to mean anything, and which event Meta should optimize toward at your volume.
Three windows. Actions, the one number to watch, and a written go or no go rule at the end of each.
Where each dollar goes and why. For a lot of Canadian brands the first dollar does not belong on Meta prospecting.
It adapts to your category. Shipping algebra with the AOV where free shipping stops eating your margin, provincial sales tax with the note for your own catalogue, the Canadian retail media networks that actually carry what you make, 2 federal funding programs with which side of the agri-food line you sit on, and 4 to 5 claim rules from the regulator that governs your ads. 6 regimes across the 7 families, so a supplement brand is never handed a snack brand's checklist.
Three limits, written plainly. A tool that runs on what you typed in cannot verify your numbers, make your creative, or measure incrementality.
Every benchmark in it carries its publisher and what we took from them, so you can check any number we used against the source. Where your family has no published per vertical figure, the report says the number is derived and names what it was derived from. Your $47 is credited toward your first month if we end up working together.
Four questions get you your contribution margin, your breakeven ROAS and the modeled cost per purchase for your own category in Canada, not for food. No email, no signup. Read that first, then decide whether the rest is worth $47.
Most CPG benchmarks on the internet are American, unsourced, or both, and the sourced ones are usually about food. Here is the full list behind the calculator above and the report, with what we took from each one.
First-order contribution ~35%, DTC gross margin 40-50%, food AOV $30-60, repeat 25-30%, 12-month LTV $100-200, subscription churn, trade spend 15-25%
DTC food & beverage CAC $45-53 USD; beauty $90-130; pet care $68-90; fashion $90-120
There is no published Canada-only CAC benchmark, so North America USD figures are the proxy. Food $25-80, beauty $28-120, supplements $80-130 USD. One US dollar of Meta or Google inventory costs about C$1.39, a hidden FX cost specific to Canadian operators
Orders per customer per year, one-time vs subscription cohorts: beauty 2-4 / 4-7, supplements 2-4 / 7-10, food and coffee 2-4 / 6-9, pet 3-5 / 8-12, home goods 1-2 / 2-3
F&B payback 1-3 months, beauty 2-4, pet 2-4, supplements 3-6. Under 6 excellent, under 12 healthy. Below 2:1 is dangerous in food
Breakeven MER = 1 / contribution margin; 2.50x at 40%, 2.22x at 45%, 1.82x at 55%
69% median gross margin across public beauty brands
56.6% median DTC gross margin, 45.6% to 63.8% interquartile. Private brands should subtract 200 to 500 bps from any public range
CAC and gross margin by category: beauty $61 / 60-70%, supplements $61 / 60-70% with 37.7% repurchase, food $53 / 40-55% with 25-35% repurchase, pet $59 / 27-74%, home $58-77 / 45-55% with 15-20% repurchase
Repeat rate by category: consumables 30-40%, beauty 21-35%, apparel 15-26%, home and durables 10-18%. DTC-wide average near 28%
Breakeven ROAS = 1 / contribution margin %; which variable costs belong in the line
Meta CPC F&B $0.42-0.52, Meta CVR F&B 2.02%, Meta CTR F&B 1.85%, subscription CVR floor 30-35%, coffee and snack subscription CVR, replenishment LTV $400-900
Cosmetics CPM runs 30-40% above the ecommerce average, $18-34 vs $14-22 USD. CTR 0.6% for fragrance awareness up to 2.1% for skincare retargeting, 1.40% blended. CPA $28-95. Realistic blended ROAS 2.5-3.5x. UGC beats studio creative by 35-55% on cold-traffic CTR in every sub-vertical except luxury fragrance
Aggregate repeat 18.8% across 156,110 customers; consumables 30-45%; first-to-second 18-30%, second-to-third 35-55%
Sub-$60 AOV median CVR 2.42%; sub-$50 AOV brands convert 3-5%; curation subscription churn 12-18%
Tier-1 markets including Canada report CPM $10-23 USD
Food & Beverage is the cheapest industry CPM; +8.35% year over year, below platform average
Platform-wide 2026 averages: CTR 1.05%, CPM $13.05, CPC about $1.40, ROAS 3.28x across ten industries. Baby CPM up 8% year over year
2-3 months of organic content backlog before paid; repeat purchase % is the strongest CPG product-market-fit signal; Instacart as the accessible first paid channel
Email is 27% of ecommerce revenue; flows drive 41% of email revenue from 5.3% of sends; SMS flows 45.2% from 7.6% of sends
AgriMarketing Market Diversification SME Stream: up to $100K per project, 70:30 cost share, opened February 13 2026, agri-food SMEs under 250 employees and $50M sales
Agriculture and agri-food businesses are excluded from CanExport SMEs starting 2026-27. Non-agri-food SMEs can claim up to $50,000 at 50:50. The 2026 intake runs February 4 to August 31, 2026
B2B accepts $27-$97; start at $47 and test on 100 leads; under 15% conversion means too high, over 30% means raise it
Claim categories and prescribed wording for Canadian food advertising
Front-of-package nutrition symbol requirements and the January 1, 2026 compliance date
Cosmetic Notification Form is mandatory: composition must be notified to Health Canada. As of April 12, 2026 fragrance allergens must be declared on the CNF above 0.01% in rinse-off and 0.001% in leave-on products
The Hotlist names substances prohibited or restricted in cosmetics sold in Canada. A consultation on proposed updates ran November 19, 2025 to February 17, 2026
Natural health products need an 8 digit NPN from the NNHPD. The permitted claims come WITH the licence: the product licence sets the conditions of use, formulation, claims and labelling
Health Canada runs risk-based proactive surveillance of natural health product advertisements, including monitoring for unauthorized claims about serious diseases
Loblaw, Walmart, Sobeys, Metro, Canadian Tire and Amazon are all accelerating Canadian retail media investment in measurement and self-serve control
Sephora is the leading Canadian prestige beauty retailer at roughly 30% of the prestige segment with 100+ stores. Shoppers Drug Mart has the broadest footprint at 1,300+ locations and runs the BeautyBOUTIQUE platform
Benchmarks published in United States dollars are converted once, at a single rate held in one place, so every converted number on this page moves together and none of them is converted twice. 3 of the 7 families have no published per vertical CPG data at all: household and cleaning, baby and kids, other CPG. Their bands are derived from adjacent categories, and every report that uses them says so on the page.
