Finance & Unit Economics
P&L, Customer Economics, CAC, LTV, Contribution Margin, Pricing
- Revenue is vanity, profit is sanity - manage the business on contribution margin, not topline or gross margin.
- Run every acquisition decision through three gates: first-order floor, payback under 6 months, and nCAC under ~33% of contribution LTV.
- Model on landed cost, never the factory price - post-2025 it runs 40-80% above FOB for China-sourced goods.
- Hold discount discipline: a 20% discount at 50% margin needs 67% more units just to stand still.
On this page
Revenue is vanity, profit is sanity. If you don't know your contribution margin, you don't know if your brand actually works. Build in fat because you'll need it. Budget for CAC. It's often a multiple of COGS and the thing most founders forget. Your brand will sink or swim based on unit economics.
Finance is where operating reality shows up. If the numbers do not work at the unit level, growth just scales the problem. Most founders obsess over revenue and ignore the one metric that keeps them alive: payback. Get payback under six months and growth funds itself.
Jump straight to what you need: the contribution margin view, the three gates for acquisition maths, or pricing and discounting. Cash flow, working capital and funding now live in Section 32: Cash Flow & Funding.
At Quad Lock, Chris and I handled the bookkeeping, reconciled all payments, and paid every invoice ourselves until about $10M. Not because I enjoy bookwork. Because it was the best way to be completely across the finances in real time. When you're bootstrapped and growing fast, spending heavily on ad media and stock simultaneously, you need to feel the cash flow, not just read a report about it. Doing the books myself meant I always knew exactly how fast we could grow and when we needed to pull back.
Our first financial controller came on around $10M. We didn't have a full-time CFO until after our PE sell-down north of $50M. And we didn't start building a C-suite until past $100M. See Section 25: Team & Culture for the full team evolution.
This isn't a blueprint. If you've raised capital or have a co-founder with a finance background, your path will look different. The principle is: stay as close to the numbers as your situation demands. In the early days of high growth with limited cash, that meant doing it myself.
Being bootstrapped forces financial discipline in a way that raising capital never does. Every dollar of inventory is a dollar you can't spend on marketing. Every pricing decision is a bet on whether you'll have enough margin to fund growth. When nobody is writing cheques, you learn to read the numbers very quickly. That's not a disadvantage. It's the reason bootstrapped brands often have the cleanest unit economics in the market.
The ground under DTC finance has shifted, and it all points one way: economics matter more than they used to. Blended CAC has roughly tripled since 2015, so the first order carries more weight than when traffic was cheap. Venture money has largely dried up, so capital efficiency is now the price of entry. Tariffs and freight turned landed cost into a live risk. Returns have climbed from around 11% in 2020 to ~14% for DTC, with overall ecommerce running 19-20.5%. And after iOS 14, MER replaced platform-reported ROAS as the honest operating number. None of this is a reason to panic. It's the reason the gates and the landed-cost honesty in this section - and the cash discipline in Section 32: Cash Flow & Funding - are survival tools, not nice-to-haves.
The DTC P&L Structure
Breaking out returns and discounts from gross revenue lets you spot margin erosion early. Contribution Margin sits between Gross Profit and EBITDA (earnings before interest, tax, depreciation and amortisation) because it is the number you manage daily - what you keep from each order before fixed costs and marketing. Platform costs are split: per-order transaction fees (Shopify, payment processing) belong in variable costs, while SaaS subscriptions belong in overhead.
What "Healthy" Typically Looks Like by Stage:
These ranges represent what well-run DTC brands target - not the median. They vary significantly by category (beauty brands will hit the high end, apparel and electronics the low end). Use them as aspirational benchmarks to spot where your P&L has room to improve.
| Line Item | $1M Revenue | $10M Revenue | $50M Revenue | $100M+ Revenue |
|---|---|---|---|---|
| COGS | 25-35% | 20-30% | 18-25% | 15-22% |
| Gross Margin | 65-75% | 70-80% | 75-82% | 78-85% |
| Fulfilment & Variable | 12-18% | 10-15% | 8-12% | 7-10% |
| Contribution Margin (before marketing) | 47-63% | 55-70% | 63-74% | 68-78% |
| Marketing | 25-35% | 20-28% | 15-22% | 12-18% |
| Tech/SaaS benchmark* | 3-5% | 2-4% | 1.5-3% | 1-2% |
| Team | 10-15% | 12-18% | 15-20% | 15-22% |
| Overhead | 5-8% | 4-7% | 3-5% | 3-5% |
| EBITDA Margin (target range)* | -5% to 8% | 8-18% | 15-25% | 20-30% |
*Tech/SaaS is shown as a benchmarking lens only. In practice, some of it sits inside variable costs and some inside overhead, so don't add every row mechanically.
*EBITDA is shown as a healthy target range at each stage, not the strict arithmetic result of summing the worst-case or best-case combination of every line above. A brand at the worst end of every input would produce a wider negative EBITDA. The bands are what well-run brands actually deliver.
At $1M, you might not be profitable, that's okay if unit economics are sound. At $10M, profitability becomes non-negotiable, target 10%+ EBITDA. At $50M+, target 15-20%+ EBITDA margins. At $100M+, you should be seeing volume benefits on COGS and marketing efficiency improving. These EBITDA ranges also directly determine your valuation multiple, see Section 28: Valuation & Exit.
When we first started out, well before we had a Financial Controller, we ran the business on five main ratios.
Performance marketing as a percentage of revenue - holding this ratio kept us disciplined and helped protect profitability, as long as gross margin, fulfilment, team costs and overhead stayed within target.
Gross margin - strict, always strong.
Revenue per employee - kept us disciplined on hiring (Section 25: Team & Culture).
Stock days - at current sell-through, how many days of stock do we have, and does it cover the lead time?
Runway - if no one paid us tomorrow, how many weeks could we survive and still fulfil every liability?
That was it, but we watched them ruthlessly.
This is not financial advice. This is founder math, but it worked for us.
The Contribution Margin View
For daily and weekly management, contribution margin is the most useful lens:
Two numbers, two uses. Contribution margin before marketing is what's available to fund acquisition and overhead. This is the number you use in LTV calculations. Contribution margin after marketing is what's left per first order once you've paid to acquire the customer. This is the number you use to assess first-order profitability and payback.
Landed Cost Is Your Real COGS
The factory price on your quote is not your COGS. The number that actually hits your margin is landed cost: what one unit costs you sitting in your warehouse, ready to ship.
Founders model off the FOB price because it's the number on the invoice, then wonder why the real margin comes in light. For China-sourced goods post-2025, landed cost typically runs 40-80% above the factory price once you stack freight, duties, the 2025 tariff layers, and handling on top. FOB is only about 55-70% of what the unit really costs you. Model off FOB and you've overstated your contribution margin before you've sold a thing.
This isn't a static line any more. Tariff and freight volatility through 2025-26 can move your landed cost by hundreds of basis points with little warning, and it lands straight on contribution margin - the number that funds all your acquisition. Brands that locked in pre-tariff freight and duty rates are running materially fatter gross margins than peers buying at today's rates. If your entire supply sits in one country, that's not just an operational risk, it's a financial-modelling risk: a single tariff change re-prices your whole P&L. Build the sensitivity into your scenarios, not the footnotes. For mitigation - dual sourcing, terms, reactive ordering - see Section 7: Supply Chain & Operations.
The practical rule: every margin number in this section runs on landed cost, never FOB. When you stress-test the P&L, model a tariff or freight spike as one of your scenarios. The brands that get caught are the ones who treated landed cost as a fixed input and woke up to find it had moved underneath them.
Aim for 50%+ before marketing as a starting point, 30%+ after marketing on first orders. The right number depends on your category and margin structure. Critical insight: first orders might be 30-50% (after CAC), but repeat orders should be 50-70% (no acquisition cost). That's why retention matters so much to the financial model. See Section 21: Customer Retention & Loyalty.
Returns Are a Margin Line, Not an Afterthought
Returns sit at the top of the P&L waterfall for a reason, but most founders treat them as noise. They're a structural drag on contribution margin, and the rate has climbed hard.
The rate is only half of it. Each return carries a real cost: reverse shipping, inspection, restocking, and on soft goods the write-down when the item can't be resold at full price. Reverse shipping and processing labour alone run $8-18 per return; fold in inspection, restocking and unsellable units and the all-in figure runs higher, soft goods worst of all. At a 14% return rate, even just that $8-18 processing layer quietly clips a couple of points off contribution margin before you've counted a single write-off.
A point off your return rate flows straight to contribution margin, same as a point off CAC - and it's often cheaper to win. The levers are unglamorous and they work: accurate sizing guidance, brutally honest PDP photography and copy so expectations match reality, and a return policy whose economics you've actually modelled rather than copied from a competitor. Track return rate per SKU. The ones that spike are telling you something about the product or the listing, not the customer. This is the financial lens; the root-cause reduction playbook (reason-coding, quality systems) is Section 8: Quality & Returns, and the retention lens (exchange-first) is Section 21: Customer Retention & Loyalty.
Unit Economics Deep Dive
Most founders ask one ratio to answer three different questions: am I profitable today, when does my cash come back, and is this customer ultimately worth acquiring? One number can't carry that load. This section answers the three questions separately, in order, with three gates. Get the definitions right first, then run every acquisition decision through the gates.
Rule Zero: Definitions or Nothing
An undeclared metric is a vanity metric. Before any LTV or CAC number enters a meeting or is used for a decision, it must declare itself:
Revenue LTV against paid CAC looks roughly twice as good as the honest version. Blended CAC hides an expensive paid channel behind a cheap email list. A 36-month window makes any CAC look fine. None of these are lies exactly. They're undeclared assumptions, and undeclared assumptions are how brands scale themselves broke.
CAC: Total marketing spend / new customers acquired. Include ALL acquisition spend (ads, agency fees, creator costs aimed at acquisition) for blended CAC, not just your best channel. Track YOUR blended CAC trend over time - that matters more than outdated industry averages.
nCAC vs Blended CAC: nCAC (new customer acquisition cost) isolates the cost of acquiring genuinely new customers. Blended CAC includes all customers (new and returning). Both matter. nCAC tells you the true cost of growth. Blended CAC tells you overall efficiency. Track both. For how this connects to your ad account structure, see Section 14: Meta Ads.
Contribution LTV:
Use contribution profit before acquisition cost. Gross margin ignores fulfilment, processing, and platform costs, so it overstates what a customer is worth. Contribution after marketing double-counts CAC when you compare to CAC. The right number is revenue minus COGS, fulfilment, payment processing, returns, duties, and platform fees, but excluding acquisition CAC.
As a sanity check expressed against revenue (revenue LTV, not contribution LTV), single-purchase brands run ~1.2-1.5x AOV, moderate repeat 2-4x, high repeat or subscription 4-8x. Use revenue LTV for a quick gut check only. Every decision runs on contribution LTV.
The First Order Is Not an Average Order
Before you run a single gate, look at the shape of one customer's orders. Store AOV blends new and returning customers into one flattering number, and the gates don't run on blends - Gate 1 runs on the first order, because that's the order your nCAC bought. Three shapes cover almost every DTC brand:
| Shape | What it looks like | Who runs it | What flat AOV does to your gates |
|---|---|---|---|
| Front-loaded | High first order, smaller repeats | System and kit brands - the customer buys in, then accessorises. This was the Quad Lock pattern: a bigger initial purchase, then smaller ongoing orders | Understates the first order, so Gate 1 reads a false RED. You under-spend on acquisition that was actually self-funding |
| Flat | First and repeat orders about the same | Broad catalogues and most replenishables, where order values average out | Fine - flat AOV is your honest number |
| Back-loaded | Discounted first order, bigger repeats | Intro-offer and subscription-led brands | Overstates the first order, so Gate 1 reads a false GREEN. You think acquisition self-funds while every new customer digs a hole |
Run the gates on your first order value and your repeat order value as two numbers. Shopify splits new-customer and returning-customer AOV natively - this is one report, not a cohort project. If the two are within a few dollars of each other, you're flat, and one number is fine.
The Three Gates
Run every acquisition decision through these in order. Each gate answers one question. Failing an early gate isn't automatically fatal, but it must be a conscious, funded decision, never a discovery.
The three gates at a glance - the whole framework on one screen before the detail below:
| Gate | The question it answers | Passes when | Healthy band |
|---|---|---|---|
| 1. The Floor | Am I profitable on this customer today? | First-order contribution covers nCAC, or the day-one hole is capped and funded | First-order result at or above zero |
| 2. The Payback | When does my cash come back? | Payback lands inside your category window | 3-6 months; over 12 is fragile |
| 3. The Value | Is the customer ultimately worth it? | nCAC is a small share of contribution LTV at a stated window | nCAC under ~33% of LTV (3:1+) |
Gates 1 and 2 are the cash gates and bind at all times; Gate 3 is the value screen and, run across cohorts, the ranking that points your budget at the most valuable customers. Everything below is how to run each one honestly.
To make the gates concrete, one brand runs through all three below. The numbers are deliberately middle of the road, not elite; swap in your own as you go. A replenishment product: $80 an order at 50% contribution margin before CAC, so $40 contribution profit per order. nCAC of $60. The cohort table shows 3.5 orders per customer in the first 12 months, a reorder roughly every 3-4 months. One assumption to declare per the shapes above: this brand's first and repeat orders are the same size - a flat shape. If yours aren't, run the split; a front-loaded contrast follows the gates.
Gate 1: The Floor (First-Order Contribution After CAC)
Positive means acquisition self-funds: every new customer adds cash on day one. Negative means you're investing: the repeat curve has to fill the hole. Neither is wrong. What's wrong is not knowing which one you're doing, or letting the hole deepen without a decision. Strict first-order profitability caps growth for brands with strong, proven repeat; ignoring the first order entirely is how brands with weak repeat scale themselves into the ground.
(Inputs: $40 contribution/order, 3.5 orders in 12 months, $60 nCAC.) $40 first-order contribution minus $60 nCAC = -$20. The first order doesn't cover acquisition: each new customer starts $20 underwater. The cohort table shows the repeat curve is real, so this brand sits in the amber row: a capped, deliberate day-one hole, not a gamble. A bootstrapped operator would push that $20 toward zero before scaling.
First-order contribution looks best the day the order ships and worse three weeks later when the return lands. Run Gate 1 on contribution net of expected returns - a category-specific reserve, see Returns Are a Margin Line - not the point-of-sale number, or you bank a day-one profit that doesn't survive the returns window. The rule: never judge a cohort on Gate 1 until it's older than your returns window. In returns-heavy categories, apparel especially, the point-of-sale read is an upper bound, not the result.
Gate 2: The Payback
The most important operating metric in this section. It connects acquisition spend to the only resource that actually kills companies: cash.
DTC payback is lumpy, not smooth: the first order's contribution lands on day one, then reorders arrive months apart. The repeat curve, not the first order, sets the payback clock.
Adjust by category and by cash position. Consumables and replenishables (food, beauty, pet, supplements) should land 1-6 months; fashion and home goods 3-6 when repeat is genuinely strong, 6-9 when it isn't (fashion is the category where this band is most contested, so set it from your own observed repeat, not the optimistic end); durables and electronics can stretch 6-12+, which is exactly why those categories punish bootstrapped operators who scale on faith. And read every band through your funding: these assume you're growing from cash flow. A venture-funded brand with observed repeat can rationally carry a longer payback - it's buying market share - but it finances every extra month, and the historical failure rate of that bet is high, so the bar for a genuine land-grab is high too.
(Inputs: $40 contribution/order, 3.5 orders in 12 months, $60 nCAC.) Gate 1 left a $20 hole. Customers reorder roughly every 3-4 months, so the second order lands around month 4 and adds another $40 of contribution: cumulative $80 against $60 nCAC. Payback is roughly month 4. That's the healthy band for a replenishable, with no working-capital alarm.
Two brands can have identical LTV:CAC and completely different survival odds. The one with 3-month payback recycles its marketing dollars four times a year and compounds. The one with 14-month payback needs someone else's money to grow and dies in the first sustained CAC spike. The ratio ignores time. Your bank account doesn't.
Gate 2 measures how fast your marketing dollar comes back - it is not the whole cash story. DTC also sinks cash into inventory you buy months before you sell it, and the two compound: a 4-month marketing payback on a brand growing 80% a year with a 90-day inventory cash cycle still finances months of working capital at all times. A green payback band tells you the marketing maths works, not that you can afford to scale. Multiply your scale rate against your cash conversion cycle before you believe it - the Cash Conversion Cycle and the Growth Death Spiral in Section 32: Cash Flow & Funding are the other half of this gate.
Gate 3: The Value (Is the Customer Worth It?)
The long-run screen, and the gate to hold most lightly: it's a hurdle to clear, not a number to maximise. A pristine ratio usually means you're underspending, not winning (that's Maximise Dollars, Not Ratios below). State it whichever direction is useful, it's the same number:
nCAC % of LTV = nCAC / Contribution LTV (stated window)
(Inputs: $40 contribution/order, 3.5 orders in 12 months, $60 nCAC.) 3.5 orders x $40 = $140 contribution LTV at 12 months. nCAC of $60 is 43% of that, about 2.3:1. That's the workable band, and it only passes because payback landed inside 6 months at Gate 2. The job now is pushing toward 3:1 by lifting first-order value or repeat rate, not by spending harder.
(Inputs: $150 first order, $60 repeat orders, 50% margin, $70 nCAC, 3 orders in 12 months.) First-order contribution is $75 against $70 nCAC: +$5, Gate 1 green - acquisition self-funds on day one. Payback is immediate. The 12-month LTV is $75 + 2 x $30 = $135, a 1.9:1 ratio - thin, but with a profitable first order that's a brand question (why is the tail thin?), not an acquisition veto. That's the front-loaded shape working as designed: the buy-in funds growth, repeat compounds on top. Now run the same brand through flat AOV instead ($90 an order) and Gate 1 reads -$25 - the maths would tell a self-funding brand to stop scaling. That's why the shapes come before the gates.
Want the ceiling as a number rather than a verdict? The allowable CAC calculator turns these same inputs into both ceilings - the nCAC that keeps you profitable on day one, and the one that clears your target ratio.
A few notes on reading this honestly:
- The window changes the verdict. A brand failing Gate 3 at 12 months can pass at 36, but only with observed multi-year cohort data. A retention curve you're hoping for is not an asset.
- The old "3:1 rule" came from SaaS, and the number was never the problem - the dropped assumptions were. SaaS runs 80%+ margins, recurring revenue and multi-year predictability, so revenue and profit are nearly the same number and a long horizon is safe to bank. DTC has none of those. Importing the 3:1 threshold while silently dropping the SaaS assumptions is the actual error. Run the gate on contribution LTV at a window you can observe, or don't bother running it.
- A positive Gate 1 changes the red verdict - if it's a real one. If you bank profit on day one, a thin 12-month ratio can't sink you - there's no hole for repeat to fill. For low-repeat categories (durables, gifts) the gate becomes a brand question (why is the tail thin?) rather than an acquisition veto. That was the Quad Lock shape: profitable first orders, repeat as compounding on top. But the carve-out only holds if that day-one profit survives scrutiny: positive on your paid/incremental nCAC, not a blended number flattered by organic, and on returns-settled contribution, not the point-of-sale figure. A profit that evaporates once you isolate paid and subtract returns was never a cushion, and it can't neutralise a thin Gate 3.
- No 12-month cohort yet? Use the window you have. A young or fast-growing brand whose customers are mostly weeks old can't run a true 12-month gate - and extrapolating a hopeful tail is the cardinal sin above. The fix isn't to guess; it's to run the gate at the longest window most of your cohort has actually reached (often 60 or 90 days), label it as that per Rule Zero, and scale conservatively against it - widening the day-one hole you'll tolerate only as longer cohorts mature and confirm the curve. A short observed window beats a long imagined one.
One quality check alongside the gate: if a customer's cumulative value grows roughly 30% beyond their first order within 60 days, and roughly doubles within 12 months, you've built a genuine brand connection and your economics deserve aggression. Flat after the first order means you're a product, not a brand, and Gate 3 will only pass at windows you can't afford to wait for.
Run Gate 3 across your cohorts, not just one, and it stops being a filter and becomes a ranking. Two cohorts can both clear it - and the one with the higher contribution LTV can sustain a higher nCAC and still pass, which is the maths telling you to pay more to win your best customers, not less. The reflex most brands get wrong is the opposite: chasing the lowest CAC. A cheap-to-acquire customer with no repeat nets out behind an expensive one with a long tail, so the cheapest cohort is usually not the one to scale. Chase the highest net value, and the gate turns from a yes/no test into an allocation engine - a higher-LTV cohort earns a higher allowable nCAC, more budget, and more creative. The full budget-allocation mechanics (cat-geo and LTV-weighted allocation) are in Section 14: Meta Ads; Section 21: Retention & Loyalty applies the same lens by channel, and Section 4: Know Your Customer by cohort.
Scaling Happens at the Margin
The gates above describe your average customer. Scaling decisions live somewhere else: the next dollar.
Average View (What Happened)- Blended CAC looks healthy
- Account ROAS still above target
- Early efficient spend averages down late expensive spend
- You discover the inflection two months after you passed it
Marginal View (What the Next Dollar Does)- Cost of the LAST increment of spend, not the average
- Marginal CAC rises before blended ROAS falls
- Spend up to the point where the marginal customer still passes Gates 1 and 2
- Pull back the moment it doesn't
The operating rule: spend down the efficiency curve until the marginal customer fails Gate 1 or Gate 2. Those are cash gates and they hold at all times. Gate 3 can be consciously relaxed at the margin, but only when all three conditions are met: topline is exceeding plan, the LTV tail is observed in your cohort tables, and the relaxation is a pre-agreed ceiling that steps back down, not a ratchet that only moves up.
There's a subtlety in which nCAC the gates run on, and it flips between diagnosis and decision. For the steady-state question - is this brand healthy? - blended nCAC is fine. For the spend-more question, the gates have to run on the marginal customer's incremental nCAC, because that's the cost the next dollar actually pays. Blended nCAC averages in your cheap, organically-acquired and returning customers, so it systematically understates what growth costs - and a brand whose blended Gate 1 looks self-funding can be badly underwater on paid alone, propped up by an organic base that doesn't scale with the ad budget.
And it isn't only the cost that's marginal - so is the value. Paid traffic usually converts a worse mix than your organic and referral base: more one-and-done, fewer subscribers. So the customer your next dollar buys can be both more expensive and lower-LTV than your blended cohort average implies. When paid and organic acquire genuinely different populations, run the gates on the paid-acquired sub-cohort - the people your next dollar actually buys - not the flattering blend of everyone.
How do you get a marginal number when the dashboard only shows the blend? You estimate it, three honest ways: step spend up in deliberate increments and watch where the next tranche's cost per new customer jumps; run periodic geo-holdouts or a media-mix model to calibrate true incremental CAC against platform-reported figures; and hold a stop-rule - keep spending while the marginal customer's incremental return clears your profitability floor, stop the moment it doesn't. Blend for the daily dashboard, margin for the decision. Before you commit the next increment, run it through the Can I Scale Ads? calculator - it puts the marginal customer through Gates 1 and 2 for you:
Revenue above budget during a high-intent period proves nothing about what your extra spend caused. The only honest test is a holdout: run the spend in one geography, hold another flat, compare. Once you're spending roughly $1M a year on the channel you're testing (often $3M+ in revenue) and your sales are steady enough that the expected lift clears weekly noise, you have the volume to do this - and the tooling floor is dropping fast as platforms move to Bayesian models. Until you've run one, treat every "incremental" claim from your own dashboards as an upper bound.
Maximise Dollars, Not Ratios
A falling ratio with rising total contribution dollars and intact payback is success: you're spending further down the curve and banking more money. A pristine 5:1 with a flat bank balance and flat growth is the failure mode wearing a medal. Acquiring 5,000 customers at $70 contribution each beats acquiring 1,000 at $100 each, provided the marginal customers clear Gates 1 and 2 and your cash position carries the timing.
At Quad Lock, first orders were profitable after CAC at the blended nCAC level. Worth being precise about that, because it's Rule Zero in action: on average, a new customer's first order covered the average cost of acquiring them, which made growth self-funding. It doesn't mean every individual customer was profitable on day one. Some cells and some weeks ran more expensive, and the cohort tables told us where. The marketing dollar came back inside the first transaction on average, and the repeat curve, roughly half of any day's orders, was pure compounding on top rather than a rescue plan we were depending on. The discipline behind it was simple. We tracked nCAC per category-geography cell, measured it against what the cohort tables said those customers were worth, and the spend ceilings came out of that maths. They didn't drift upward because we'd had a good month. The gates don't tell you to hold the line where we held it. They tell you to know exactly where your line is, what holding it costs you, and to move it only by decision, never by drift.
First-order value levers: Bundles (+15-25%), free shipping thresholds (set 20-30% above AOV for flat shapes; anchored to the first order for front-loaded brands - see Section 7: Supply Chain & Ops for the shape carve-out), post-purchase upsells (ReConvert, Zipify: +5-15%), cart cross-sells (+5-10%). See Section 11: Website & Conversion for implementation. Be precise about which order you're lifting: every dollar of first-order value flows straight through Gates 1 and 2 - higher first-order contribution shrinks the day-one hole and shortens payback without touching CAC. A returning-customer cross-sell lifts store AOV and moves neither gate. For front-loaded brands, engineering the first-order bundle - attach at the moment of buy-in - is the core Gate 1 lever; that was the Quad Lock move.
Contribution Dollars Are Not Profit Dollars
The three gates have a hard limit worth stating plainly: they validate the customer, not the business. All three run on contribution - the money left after the variable cost of one more order. None of them sees your fixed base: the team, the agency retainers, the software stack, the rent, the brand marketing that never ties to a cohort. A brand can pass all three gates on every customer and still lose money, because contribution is what pays for that fixed base, and "maximise contribution dollars" says nothing about whether there are enough of them to cover it.
This is where the maximise-dollars rule needs its governor. Scaling rarely adds only variable cost - more customers usually mean more support headcount, more agency fee, more 3PL and returns complexity - so the contribution-maximising level of spend can sit past the profit-maximising one. The honest marginal rule is therefore stricter than the one above: spend down the curve until the marginal customer stops covering its fully-loaded marginal cost, not just its marginal CAC.
So before the gates become a licence to scale, run one check against the business rather than the customer: does the total contribution your spend throws off clear your total overhead and land you in your stage's EBITDA band? (See The DTC P&L Structure - roughly break-even is fine at $1M, but 10%+ EBITDA is non-negotiable by $10M.) Revenue is vanity; contribution is the truth about a customer; EBITDA is the truth about a company. The gates get you the first two. Don't mistake them for the third.
Put real numbers on it with the break-even calculator - it folds your fixed costs and monthly ad spend into a single revenue target, then grades it against your stage:
Break-Even ROAS, MER, and the Margin That Sets Them
Gate 1 asks whether the first order clears nCAC. ROAS is the same question from the ad account's side, and your contribution margin sets the line you have to clear. The breakeven point isn't a number someone hands you - it falls straight out of your own margin.
The higher your margin, the lower the ROAS you can survive on. That's the whole game in one line. Run your own numbers with the break-even ROAS calculator - it takes your margin, hands you the bar, and grades your current ROAS against it.
A 40% contribution-margin brand breaks even at 2.5x ROAS. Drop to 30% margin and you need 3.33x. At 25% margin the bar jumps to 4.0x. Same ad performance, three different verdicts - because the margin, not the ad, decides what "good" means. This is why two founders can run identical campaigns and one prints money while the other bleeds.
Manage the blended number with MER (Marketing Efficiency Ratio): total revenue divided by total ad spend, across every channel. With platform attribution gutted post-iOS 14, MER is the honest read - the platforms each claim the same sale, but your bank statement shows one revenue figure and one spend figure. Because MER counts the email, SMS and organic revenue that paid media drives on top of itself (see Customer Retention & Loyalty), it reads higher than any single platform's attributed ROAS. Healthy MER climbs with scale as brand demand compounds: roughly 1.5-2.5x at $1-5M, 2.5-3.5x at $5-10M, 3.0-4.5x at $10-25M, and 3.5-6.0x+ past $25M (the same stage bands as Section 27: Measurement & Data, which owns the full MER treatment). The split within a stage is telling: brands growing 30%+ a year run a median ~3.8x MER against ~2.1x for the plateaued ones. Watch your own MER trend monthly. It's the cleanest single read on whether the marketing engine is paying for itself.
Pricing Strategy & Discipline
A 1% realised price improvement can lift operating profit by 8-11%, assuming volume holds.
Source: McKinsey's "The Power of Pricing" (2003, S&P 1500 average: 8%) and Marn and Rosiello's "Managing Price, Gaining Profit" (Harvard Business Review, 1992, 2,463 companies: 11%). The multiplier is your operating margin inverted, so thinner-margin businesses see a bigger lift per price point - and the "volume holds" assumption is the hard part, since price rises usually cost some volume.
Pricing is the single highest-leverage financial decision in your brand. Most founders set prices in five minutes and never revisit. See Section 5: Product for how to set your price (value-based vs cost-plus, price architecture, psychological tactics). This section covers the financial discipline side.
The Discounting Trap
This kills more DTC brands than bad products. Here's the cycle:
Discounting is a slippery slope. The further you discount, the more you need to. The better question is: how can we manufacture a result without discounting?
At Quad Lock, we held discipline for about a decade: a 10% popup always-on, and 30% off once a year for BFCM. Only once a year, so customers were never trained to wait for a sale. BFCM gave us cover to discount without discounting the brand - cultural, expected, time-limited. It didn't stop people buying in March. Come November, that same customer grabs the accessory they've been meaning to get. Incremental revenue from the existing base.
Direction: Be cautious discounting >20% on core products. Use value-adds instead of price cuts. Keep genuine sitewide sale events rare (cadence rule in Section 31: Peak Season & Promotions). Protect your hero SKU - discount entry products or bundles. Loyalty discounts > acquisition discounts.
The discount math founders forget:
| Starting gross margin | Discount | Extra units needed to keep same gross profit |
|---|---|---|
| 60% | 10% | 20% more units |
| 60% | 20% | 50% more units |
| 60% | 30% | 100% more units |
| 50% | 20% | 67% more units |
| 40% | 20% | 100% more units |
A 20% discount on a 50% margin product means you need to sell 67% more units just to stand still. That's before you account for the long-term brand damage of training customers to wait for sales.
Pricing Benchmarks by Category (Premium DTC Target Ranges, 2026)
These are target ranges for well-run premium DTC brands, not category averages. Mass-market or commodity products in any of these categories typically run 10-20 percentage points lower on gross margin. Sources: Eightx 2026 ecommerce margin benchmarks, Northstar Financial Advisory 2026 ecommerce profit margins.
| Category | Typical Retail Range | Premium DTC Gross Margin | COGS % of Retail | Notes |
|---|---|---|---|---|
| Skincare / Beauty | $25-85 | 65-82% | 18-35% | 80%+ possible for premium skincare/fragrance, but CAC is often high |
| Supplements | $30-70 | 60-78% | 22-40% | Stronger when subscription/replenishment economics work |
| Apparel | $35-150 | 50-65% | 35-50% | Returns and markdown erode realised margin |
| Home goods | $40-200 | 40-60% | 40-60% | Bulky shipping and breakage compress contribution |
| Electronics / Hardware | $50-300 | 30-55% | 45-70% | Branded accessories sit higher; commodity electronics lower |
| Food & Beverage | $15-60 | 35-55% | 45-65% | Cold chain, co-packing and freight matter |
| Pet products | $20-80 | 45-65% | 35-55% | Consumables lower, accessories higher |
Common Pricing Mistakes
- Racing to the bottom. If your only advantage is price, you don't have a brand.
- Ignoring contribution margin. Gross margin doesn't matter if CAC eats it. Track contribution margin per order.
- Not testing prices. Tools like Intelligems can reveal customers would pay 15-20% more than you're charging.
- Forgetting unit economics at scale. Model at 10x your current volume - shipping, CS, and return costs become material.
Section 26 Checklist
Benchmarks for this section
See what good looks like on the numbers that matter here:
- Unit economics benchmarks for DTC - The margin, payback, and cost ratios that decide whether growth makes money. These are...
Tools for this section
Free Excel tools that pair with this section:
- DTC P&L Template - A profit and loss built for direct-to-consumer brands - the right line items, in the right order, from revenue down to EBITDA.
- DTC Economics Calculator - What a customer costs to win, when the cash comes back, and what they're worth at 12 months - run the three gates in your browser, or download the full Excel model with the EBITDA bridge.
- Customer Acquisition Planner - Plan acquisition like a portfolio - customer value and allowable CAC by persona and geography, in one place.
Go from reading to doing.
You've read the section. Sign in to score your ecommerce brand with the Health Check, track your progress across 472 checklist items, and get your tools and history in one dashboard. Free, and always will be.