I talk to a lot of brands whose real business is on a shelf. Grocery, natural channel, foodservice, specialty retail. Amazon is somewhere between an afterthought and an irritation, run by nobody in particular, and the reasonable-sounding position is that it isn't their channel.
Amazon is your reference page whether or not it's your channel. A shopper standing in the aisle holding your product pulls out a phone and looks you up, and what they find decides whether the thing goes in the cart.
Plenty of grocery shoppers look things up mid-aisle. In FMI's U.S. Grocery Shopper Trends research for 2026, 39% of grocery shoppers said they run a web search on their phone while shopping in person, and 37% said they visit a retailer or manufacturer website. My own rule of thumb after years of watching this is that a customer who discovers you somewhere else double-checks Amazon before committing close to nine times in ten, even when they end up buying elsewhere.
So the question isn't whether you want to be in the Amazon business. It's what your page says about you to the person deciding right now.
Google is where you ask a question, Amazon is where you ask which product
This is the cleanest way I know to explain what the platform is for.
Someone wondering whether paleo or keto suits them goes to Google. That's a question, and it has an answer that isn't a product. Someone who's decided goes to Amazon, because now the question is which one, and Amazon is built to answer exactly that with reviews, images, prices and a comparison to the six things next to it.
More than half of US product searches still start on Amazon, though the number is moving. A Feedvisor survey of 1,006 US consumers, fielded with Zogby Analytics in August 2026, put Amazon as the starting point for 62% of product searches, down from 80% a year earlier as AI assistants take on some of the early research. The same survey found that 74% still name Amazon when they're ready to buy, and that only 5% buy straight away when an AI recommends a product. The rest go and check.
That checking step is where a retail-first brand gets judged. Wherever the research starts, it tends to end on a page with reviews, a price and a photo of the pack, and for most food and beverage categories that page is on Amazon.
For a brand, that means the moment of choosing between you and a competitor increasingly happens on a page you may not be managing.
What your page says when you're not looking
A shopper landing on a neglected listing tends to find the same things.
Images shot years ago, sometimes by a distributor. Copy that describes a formula you've since reformulated. A price set by a reseller who bought a pallet, which may be higher than your own shelf price or oddly lower. A handful of reviews, some of them about shipping damage rather than the product. Maybe a competitor's sponsored ad sitting at the top of your page, which Amazon sells whether or not you're advertising.
It can get worse than stale photos. If a brand's own ads are paused, a shopper who types the brand's name can be served a competitor's ads instead. Resellers create duplicate listings, and it's common for a brand to have several times as many live listings carrying its name as it thinks it has. Nobody at the company decides any of that. It accumulates because nobody owns the page.
None of that is a marketing problem in the abstract. It's the answer a real buyer gets at the moment of decision.
And it isn't only shoppers. Retail buyers look you up before a category review. Investors look you up during diligence. It's the fastest brand lookup available to anybody, which makes it the most consequential page you don't control.
Doesn't our own website do this job?
I get pushback on this one, and I'll be plain about it.
Across our entire client base, we have not had a single brand whose own ecommerce store outsells its Amazon. Not one. These are brands with real sites, real email lists and real investment in direct-to-consumer.
That isn't an argument against having a good site. It's an argument against assuming the site is where online demand lands. Your site is where people go once they've decided they want you specifically. Amazon is where people go when they're deciding, which is a much larger group and includes everyone who hasn't heard of you.
The two also feed each other more than brands expect. Customers discover you on social and look you up on Amazon. We have a client I keep pushing to get their own paid social and search in order, because when those campaigns work, a good share of the sales show up in our Amazon reports anyway. Amazon is extremely sensitive to what happens to your brand off the platform. A good PR hit and we see a lift. A bad one and it's crickets. We can measure that as a surge in branded search traffic, or the absence of one.
It runs the other way too. Amazon shows you which search terms a product gets sales for, and we see brands carry that language onto their own product pages and lift conversion there as well. I think of Amazon search as conversion SEO: Google helps someone answer a question, and Amazon decides what they buy when they search for something specific.
Will selling on Amazon cause trouble with our retail partners?
Usually not, as long as no other marketplace undercuts your Amazon price, because that's the price conflict Amazon punishes, and it punishes it fast.
The worry I hear most is about the shelf: that a buyer will see you on Amazon for less and hold it against you. Once we've priced a brand's Amazon listings so the margins work, it's rare that we still have a problem competing with retail pricing, and rare that the brand's own website causes trouble either. I'll often recommend a brand price its website a little below Amazon. The website is where you own the customer data, and what you learn there about positioning and messaging ends up helping Amazon too. If you drop the website price dramatically, you may hear from Amazon about it, but that's a bridge you can cross when you reach it.
What I can't have is your price sitting lower on another marketplace. Walmart, Target, or a reseller you've never heard of: when the same item is cheaper somewhere else, I see the Buy Box taken away very quickly, and then there's a mess to clean up. It cuts the other way as well. When a reseller lists your product far above what Walmart charges for the same item, Amazon can suppress the Buy Box altogether, and a coupon won't necessarily bring it back.
On Vendor Central, where Amazon buys from you and sets the retail price, the same problem shows up as a spiral. The pattern is that Amazon matches a lower price it finds elsewhere, its own margin on the product goes underwater, and it responds by pausing ads and promotions, suppressing the Buy Box and pushing for cost decreases.
The cleanest way out of all of it is an offer that only exists on Amazon. A different pack size or a bundle gets its own ASIN, and nobody can compare its price like for like. You're probably not going to sell the same pack size on Amazon that you sell in retail or even on your website anyway, and an Amazon-specific pack is often how a brand gets its Amazon margin where it needs to be in the first place.
When your retail doors close
A client of ours is losing access to thousands of stores as a retail partner shrinks its footprint. The demand those stores served doesn't disappear with the shelf. Some of it goes to a competitor, and a meaningful share goes online, and online means Amazon first.
They asked a question I thought was exactly right: if that demand is going to arrive anyway, what should advertising look like when the goal is to hold position rather than to grow?
The honest answer is that it looks different from a growth budget, and smaller. When demand arrives on its own, most of the job is making sure you're the thing it finds: in stock, ranked, converting, with a listing that closes. Spending heavily to buy sales that were coming to you anyway is the most common way an ad budget produces excellent reports and no incremental revenue.
So the read on the ad spend is: enough non-branded presence to catch shoppers who don't know your name, a defensive branded position so a competitor's ad doesn't intercept the ones who do, and nothing beyond that until you can see it earning.
This applies more broadly than shrinking retail. Any time demand moves for reasons that have nothing to do with your marketing, the job is capture rather than creation, and the budget should say so.
The stores you keep matter online too. If you're still stocked in a meaningful number of grocery doors, those products are already for sale on Instacart whether or not anyone is managing them, and Instacart's ad auction is closed to brands without shelf space. The doors you still have can buy you cheaper media there than Amazon will.
Is selling on Amazon worth it for a food brand?
It's worth it when a product clears roughly 30% gross margin after landed COGS and Amazon's fees, before any ad spend, because below that the ads have nothing to be paid out of.
We work this out per SKU, not as a blended average. Under 20% is the red zone, 20% to 30% is workable with care, and 30% and above is where Amazon can pay a food or beverage brand properly. The math is unforgiving at the low end: if you have a 10% margin, even if I'm getting you a five-to-one return on ads, you're losing money.
We politely decline to work with brands that insist on keeping margins at 20% and below. Before it gets to that, we try to lift the number, usually with a price increase or an Amazon-specific pack size, and we only walk away if the brand insists on keeping its retail price and pack.
For the fee side of the subtraction, use Amazon's own figures for your product rather than a number from an article. Fees vary by category, size and price point, and Amazon revises them. Pull your rate from the revenue calculator in Seller Central and plug in your true landed cost.
Getting this wrong is expensive in a specific way. The brands that learn Amazon the hard way tend to spend six to twelve months and somewhere in the five to six figures before they get it right. We started requiring a brand's COGS before we run a single ad years ago, after getting great marketing results for brands whose margins meant they broke even anyway.
Isn't Amazon too crowded now?
I got asked this directly, as a challenge, and I'll give the same answer.
Yes, there's more competition than there was two decades ago, and yes, the easy years are over. What came with the competition is that the barrier to entry fell for you too. Launching a product, testing a price, trying a new pack size and finding out whether the market wants it costs a fraction of what it used to cost, and the feedback arrives in weeks rather than in a retail cycle.
For an emerging brand, that trade is favorable. You don't need to win the category to build a real business on Amazon. You need to own a niche where the demand is specific and the leader is complacent, and the platform will tell you quickly whether you have.
The brands who struggle aren't the ones facing competition. They're the ones spending against a listing that doesn't convert, in a category they picked without checking whether anyone was searching for it.
Can't a distributor or reseller just handle Amazon for us?
They can, and the cost is that someone else sets your price, owns your listing and keeps your customer data on the page shoppers use to judge you.
The usual setup looks like this. The reseller buys wholesale, marks the product up and controls the listings, the ads, the inventory and the Buy Box. The markup is what hurts. A brand that isn't on shelves outside its home region yet ends up on Amazon at a price nobody unfamiliar with it is going to pay, and at that price it won't grow past regional. When a third-party seller dumps old holiday stock below the reseller's price, the reseller's natural move is to ask the brand for a better deal so it can compete. The two companies want different things from the channel.
A reseller is set up to sell a few units at a good markup. A manufacturer trying to grow a brand past one region needs volume, a price that makes sense next to its retail shelf, and the data on who's buying. For a brand that wants to grow, those two sets of goals pull apart quickly. What a reseller keeps goes into the trade-off in more detail, including when a reseller is the right answer.
What a minimum presence looks like
If you're retail-first and you're not going to make Amazon a priority this year, you can still stop it working against you. This is the floor I'd hold.
Enroll in Brand Registry so you control the listing content and can see the reports. Own your own listings rather than leaving them with a reseller or a distributor. Get current images and copy on the page, with the claims that matter to your category visible in the images rather than only in text. Keep price parity clean enough that Amazon doesn't suppress your Buy Box over something a third-party retailer is doing. Watch your review average, because under 4.0 you're being skipped by AI answers as well as by shoppers.
Brand Registry is free, and the requirement is a pending or registered trademark for your brand name or logo from the trademark office of a country where Amazon runs a store. It's what unlocks A+ content, Brand Analytics and the tools for reporting a listing that infringes on yours, and the search reports it opens up are the ones that tell you how much demand is already arriving for your name.
Then hold your own name in search. Branded campaigns are table stakes, and they're cheap, because the shopper typing your brand already wants you. In our audits it's common to find a well-known brand capturing well under half of the searches for its own name plus its category term, with competitors bidding on the rest. Closing that gap is about the cheapest fix on the platform.
That's a few weeks of work and then a light touch. It won't build you a channel. It stops the reference page from costing you sales everywhere else, which is the actual risk for a brand in your position.
And if the reports then show more demand sitting there than you expected, you'll have found that out cheaply. That's how most of the retail-first brands we work with ended up taking the channel seriously. The data made the case, not the pitch.
If your retail plan leans on independent stores rather than national chains, Amazon isn't the only shelf worth setting up this year. Faire reaches those shops through one wholesale marketplace, and the sequence for building that wholesale business is worth reading alongside this.
The bottom line
The brands who tell me Amazon isn't their channel are usually right about where their revenue comes from and wrong about where their customers are looking.
You don't have to build a channel this quarter. You do have to decide that the page carrying your name is going to be accurate, current and yours, because it's being read either way.
If you want to know what yours says right now, send us your brand name. We'll tell you who's holding the Buy Box, what the listings claim, and what a shopper standing in an aisle sees when they look you up. No obligation attached.






