Amazon vs D2C Channel Math for Independent Brands

A framework founders use when they cannot agree

We sat in on a board meeting last spring where the founder and the CEO disagreed on Amazon. The founder had avoided Amazon for five years because it was "not the brand." The CEO argued they were leaving eight figures on the table. Neither had run the math. They were arguing based on instinct.

The math resolves the argument. Not always in one direction. This post is the framework we use to resolve it, with the places it breaks down honestly flagged.

▸ Amazon is a channel, not a brand. Do not expect brand equity to live there. ▸ Unauthorized sellers change the math. If they exist, you need presence. ▸ Contribution margin, not revenue, is the number that matters. ▸ Incrementality is the hidden variable and nobody measures it well.

The revenue illusion

Amazon revenue looks like growth. A brand doing two million a year on D2C adds Amazon and hits three million. Growth, on the surface. Three problems with the surface.

First, the Amazon margin is much thinner than D2C. Referral fees, FBA, advertising, returns. A product with a sixty percent D2C contribution margin often drops to twenty-five percent on Amazon. Same revenue line, different profit line.

Second, some Amazon customers are not incremental. They would have bought on D2C if Amazon did not exist. Cannibalization eats the revenue headline.

Third, Amazon does not build the asset of an owned audience. Every sale on D2C builds an email, a SMS subscriber, a retention cohort. Every sale on Amazon builds an Amazon customer. You do not own the relationship.

The question is not "does Amazon add revenue." It does. The question is "does Amazon add profit net of cannibalization and opportunity cost." That math is harder.

The channel math equation

Here is the simple form. For each channel, calculate.

▸ Revenue. ▸ Direct product cost including fulfillment. ▸ Marketing cost attributable to the channel. ▸ Platform fees. ▸ Returns cost.

The result is channel contribution. Then adjust for.

▸ Cannibalization. What share of this channel's customers would have bought on the other channel? ▸ Retention value. What is the one-year LTV of a customer acquired on this channel? ▸ Brand protection value. Is this channel defending against unauthorized sellers?

The adjusted contribution is what you compare.

Where Amazon clearly wins

Amazon math works for specific brands. The consistent patterns.

High search intent, low brand loyalty categories

If customers search for the product category on Amazon rather than the brand name, Amazon presence is mandatory. Categories like supplements, household goods, basics apparel, and tools fit this pattern. The customer is on Amazon with intent. You are not going to pull them to D2C for a first purchase. Get the first sale on Amazon, try to retain them with packaging inserts and brand education.

Unauthorized seller defense

If unauthorized sellers are already selling your product on Amazon, you need presence. Otherwise they control the listing, the price, and the customer experience. Brand registry plus an authorized seller account lets you control the listing even if you prefer D2C for direct sales.

Incremental geography or demographic

If Amazon's customer base reaches zip codes or demographics your D2C does not, the customers are incremental by definition. Measurable through zip-code overlap.

Subscription-resistant categories

Some categories do not convert to subscription well. Customers want to buy on demand, compare options, switch brands. Amazon fits that shopping mode. D2C with a subscription engine does not.

Where Amazon clearly loses

The reverse patterns.

High-margin, high-loyalty brand categories

Premium skincare, designer apparel, specialty food. The brand equity is most of the purchase decision. Amazon strips it. Customers who would have built a relationship with the brand on D2C instead buy once on Amazon and churn. Long-term value is negative.

Low-order-value categories

If AOV on Amazon is under about twenty-five dollars, the math rarely works. Referral fees, FBA, and advertising eat most of the margin. Unless the category has a genuine replenishment cycle that builds LTV through repeat Amazon orders, it loses money.

Brands with strong D2C flywheel

If your D2C is acquiring customers profitably through paid social, email is driving repeat, and retention is healthy, Amazon adds distraction without adding growth. Every hour the team spends optimizing Amazon listings is an hour not spent on the channel that is working.

Subscription-first brands

Shopify Subscriptions customers are worth multiples of a one-time buyer. Amazon Subscribe and Save has much weaker retention. A customer acquired on Amazon rarely converts to a D2C subscription later. The LTV differential is the whole game.

The hybrid models that work

Most brands are not pure D2C or pure Amazon. The working hybrids.

D2C premium, Amazon entry

Flagship SKUs at full price on D2C. Entry SKUs or older variants on Amazon at a lower price point. Customer enters the brand on Amazon, learns the brand, upgrades to D2C for the premium. Works when the product ladder is real.

D2C subscription, Amazon single-unit

Auto-ship core product through D2C subscription. Same product available for one-time purchase on Amazon at a higher unit price. The economics push replenishment customers to subscribe.

D2C owned experience, Amazon defense

Minimal Amazon presence to control the listing and prevent unauthorized sellers. No active Amazon marketing. Amazon revenue is small but the brand controls its presence.

The measurement problem

Every framework above depends on measuring incrementality. Incrementality is hard. Real incrementality testing requires holdout audiences, geographic experiments, or pre-post analyses with controls. Most brands do not have the volume to run a clean experiment.

The practical alternatives.

▸ Subscriber survey. "Where did you first hear about us?" for the last hundred D2C customers. Compare to Amazon cohorts. ▸ Zip-code overlap. Export D2C customer zips and Amazon customer zips. Overlap suggests cannibalization, gap suggests incrementality. ▸ Pause test. Drop Amazon advertising for a month. Watch D2C traffic and sales. A spike suggests cannibalization. Flat or declining suggests independent demand.

None are perfect. Triangulated, they give a directional answer.

The table version

Count the signals. If five or more favor Amazon, run Amazon seriously. If five or more favor D2C-only, skip or minimize Amazon.

What the decision actually looks like

A supplement brand we advised scored six of seven on favors-Amazon. We launched their top three SKUs on Amazon with a full listing strategy, FBA, and a defensive pricing model. Eighteen months later Amazon does thirty-four percent of revenue at twenty-three percent contribution margin. D2C grew twelve percent over the same period. Net business profit up materially. The founder had been wrong.

An apparel brand we advised scored six of seven on favors-D2C-only. We recommended minimal Amazon presence for brand defense only. The team stayed focused on D2C. Revenue grew forty percent year over year. If they had chased Amazon they would have diluted focus and lost the growth year.

For operational context see retention marketing, which is the function that wins on the D2C side of this equation. For platform context see Shopify development and platform migration.

What to do this week

▸ Run the channel math equation for both channels if you are on both. ▸ If you are D2C-only, run a one-week audit of unauthorized seller presence on Amazon. ▸ Run a zip-code overlap analysis between D2C and any marketplace presence. ▸ Pull the subscriber survey data if you have it. If not, add the question to the post-purchase flow. ▸ Write down the five factors above that matter most for your category and score them honestly.

Related reading

For brand-side launch context see the D2C launch calendar and the product launch playbook for boutique brands. For the retention lever that beats Amazon math for the right brand see retention marketing. For migration context if a platform change is part of the strategy see platform migration and Shopify vs Squarespace.

FAQs