Marketplace vs Own Store: Where Price Competition Happens

Same Product, Two Different Games

Sell a pair of headphones on a marketplace, then sell the same headphones on your own store, and you are not running one business through two channels. You are playing two games that disagree about what a price is for.

On a marketplace you compete inside a single listing, against other sellers of the identical item, on a page you do not control. The buyer is looking at one product with a set of offers behind it, and those offers differ in almost nothing except the number. Price is the only free variable you have left, so it absorbs all the competitive pressure in the system.

On your own store you compete across the open web against products that are not the same product. They substitute for yours. The shopper on your page is weighing a whole proposition: the brand, the shipping promise, the return policy, whether they have heard of you before. Price is one of six variables, so the pressure disperses across all six.

Almost every article comparing the two treats them as channels and scores them on reach, cost and control. That framing hides the part that actually governs your pricing behaviour. These are different competitive structures, and an identical price cut means something different in each one.

Nine Dimensions Where Marketplace and Own-Store Competition Diverge

Here is the comparison, before the arguments that come out of it.

Dimension	Marketplace listing	Your own store
Who you compete against	Other sellers of the identical SKU, inside one page you do not control	Different brands and products that merely substitute, across the open web
Unit of competition	The offer: one item, several sellers, one winner surfaced	The whole page: product, brand, shipping, returns, trust
Price visibility	Total and instant. Every rival's price sits on your own listing, machine-readable to them too	Partial and effortful. You have to go find each competitor's page, and they have to find yours
Speed of movement	Continuous and algorithmic. Rivals reprice against you within minutes	Slower and campaign-shaped. Competitors move on promotions, seasons and cost changes
Customer relationship	Owned by the marketplace. Buyer data, email and remarketing are largely withheld	Owned by you. Email list, retention, lifetime value, offer design
What a price cut buys	Share of a fixed demand pool for that SKU. You take units from another seller, often triggering a counter-cut	Marginal new demand plus positioning. No rival is forced to notice
Data you can get	Rich but contractual. The listing shows everything; the API restricts what you may store or use	Thin but unrestricted. No API, but public product pages and structured data are yours to read
Cost of the same headline price	Net of referral fee plus closing and fixed fees	Net of payment processing plus acquisition cost
Failure mode	Race to the bottom inside the listing	Invisibility: nobody is comparing you at all

Two rows carry most of the weight, and the rest of this post is mainly about them: what a price cut actually buys, and the two failure modes, which are opposites rather than variations.

One clarification on the data row before moving on. Which prices you can read, and by what means, is a per-platform question with a long answer, covered in the platform-by-platform guide to competitor price tracking. The point that belongs here is narrower: on a marketplace the numbers are all in front of you and the limit is what your agreement lets you do with them, while on the open web the limit is finding them at all.

A Price Cut Buys Different Things in Each Structure

Inside a marketplace listing, demand is fixed and you are dividing it. The shopper has already chosen the product; the only remaining question is who ships it. Every unit you win by cutting is a unit some other seller does not sell, and that seller is looking at your price on the same screen you are looking at theirs. The counter-cut is not a risk, it is the normal response, and both of you end the day with the same unit split at a lower price.

The Buy Box is the clearest evidence that this structure collapses competition to a single winner rather than spreading it: one offer gets surfaced, and the rest of the sellers are a click away that most buyers never make. How that selection works, and what kind of repricer belongs against it, is the subject of the Amazon repricing and Buy Box guide.

The competitor in that listing is usually another independent seller rather than the platform itself. Third-party sellers accounted for 60% of Amazon's worldwide paid units in the first quarter of 2026, against 61% in the prior-year quarter (Amazon, Q1 2026). Note the metric: units, not sales. Third-party seller services net sales came to $41.6 billion in the same quarter, up 14% year over year, against total net sales of $181.5 billion.

On your own store the arithmetic changes shape. A lower price can pull in demand that did not exist at the old price, because your visitor was not already committed to buying this exact SKU today from somebody. That is ordinary demand response, and the size of it is a property of your category rather than of your competitors, which is why price elasticity in ecommerce is a more useful thing to measure on your own site than on a listing. The flip side is that no rival is forced to notice. Your cut may buy you volume, or it may quietly buy you nothing but a thinner margin, and you will not get a counter-move to tell you which.

There is a long-running direction of travel underneath both. Cavallo's 2018 study of US retailers found that online competition has increased how often prices change and made pricing more uniform across the country (NBER, October 2018). That work is now several years old, so treat it as a direction rather than a current measurement, but the direction has not reversed.

The Two Failure Modes Are Opposites

Get a marketplace listing wrong and you end up in a race to the bottom. Everyone can see everyone, everyone is selling the identical item, everyone has software watching. The floor is not set by what the product is worth, it is set by whichever seller has the lowest cost base or the least discipline, and once that seller moves, matching them is the path of least resistance for everybody else. The discipline that prevents this is boring and unglamorous: know your floor per unit and refuse to go under it, even while you watch units go elsewhere.

Get your own store wrong and you fail in the opposite direction. Nobody is comparing you at all. No shopper has your page open next to three rivals, so your price never gets tested and never gets punished. That sounds comfortable and it is dangerous, because you can hold a price 20% above the market for a year and read the resulting silence as acceptance rather than as absence. Ecommerce reached 16.9% of total US retail sales in the first quarter of 2026, at $326.7 billion (US Census Bureau, May 2026), and none of that traffic is obliged to find you.

This is also the honest place to flag a gap. It is intuitive that prices are packed tightly together inside a marketplace listing and spread widely across the open web, and the structural reasons are clear enough: total visibility plus identical goods compresses, partial visibility plus differentiated goods disperses. I could not find a tier-one measurement of that dispersion contrast for 2025 or 2026, and there is no reliable current comparison of how often prices change on marketplaces versus independent stores either. Take the contrast as a structural argument, not as a number, and be suspicious of any article that quotes you one.

The Fee Arithmetic, Without the Flattening

Competing articles reduce this to "about 15% on Amazon". The published schedules are nothing like that tidy, and the spread is wide enough to change which channel makes sense for a given product.

Amazon: 5% to 45%, and Not Only on the Item Price

Amazon's referral fee runs from 5% to 45% of the total sales price depending on category, with most categories landing between 8% and 15%, plus a minimum of $0.30 per item in most categories (Amazon, operator schedule, retrieved 2026). The extremes are real products, not edge cases: Amazon Device Accessories sit at 45%, Gift Cards at 20%, while Computers and Consumer Electronics and Video Game Consoles are both 8%.

Several categories are tiered by price rather than flat. Watches take 16% up to a total sales price of $1,500 and 3% on the portion above it, which means the effective rate on a $3,000 watch is nothing like the headline. Media items carry an additional $1.80 per-item closing fee on top of the referral fee. And the base the percentage is applied to is the item price plus shipping and gift wrap, not the item price alone, so a low-priced item with real shipping cost is charged on a bigger number than the one on the price tag.

Walmart: 3% to 20%, Tiered by Price

Walmart Marketplace publishes a range of 3% to 20%, also charged on item price plus shipping and gift wrap (Walmart, operator schedule, retrieved 2026). Personal Computers are 6%, Books and Music are 15%, Jewelry and Precious Metals are 20% up to $250, and Watches drop to 3% above $1,500.

Apparel and Accessories shows the tiering most clearly: 5% at $15 or less, 10% between $15 and $20, and 15% above $20. A $19.99 shirt and a $20.99 shirt are not in the same fee bracket, which is the kind of detail that quietly decides whether a price point works. eBay publishes its own category-based schedule; read its current fee page directly rather than trusting a figure quoted second-hand.

The Comparison Is Fee Versus Acquisition Cost, Not Fee Versus Margin

Here is the framing that makes the numbers usable. A $50 sale in a 15% Amazon category nets roughly $42.50 before fulfilment. The same $50 on your own store nets around $48.50 after payment processing. Read that as a $6 advantage for your own store and you have stopped the calculation one step too early, because the marketplace sent you the buyer and you did not.

The real comparison is the referral fee against your customer acquisition cost on that unit. If it costs you $12 in paid traffic to sell that $50 item on your own site, the marketplace is the cheaper channel on that sale even at 15%, and the gap widens on categories at 8%. If your own-store buyers arrive through email, organic search or repeat purchase at close to zero marginal cost, the marketplace is expensive and the 45% categories are unworkable. Two stores with identical products and identical fees can reach opposite conclusions, and both are right, because the variable that decides it lives on their side of the equation rather than on the platform's. Where each channel sits in a wider pricing plan is the subject of the competitive pricing strategy guide.

One asymmetry worth naming: the fee is a variable cost you pay only when you sell, while acquisition cost is largely a fixed spend you commit before you know whether it works. That difference in risk profile is often what makes marketplaces attractive to smaller sellers even at rates that look punitive on a spreadsheet.

Price Parity Rules Are Contested and Moving

Sellers on both structures eventually hit the same question: can a marketplace tell you what to charge somewhere else? The honest answer in 2026 is that the rules are contested and moving, and anyone giving you a clean yes or no is overstating what is settled.

There are live threads on several fronts. The EU's Digital Markets Act is one live thread, and how it applies to marketplace pricing terms is still being worked out in practice. Read the regulation itself, not a summary of it. In the United States, the Federal Trade Commission's case against Amazon is another. Check its current status on ftc.gov rather than relying on any article's description of it, including this one. Marketplace seller policies themselves sit behind seller-account logins and get revised without announcement.

Which leaves you with an unsatisfying but correct instruction: read the agreement you are actually signed to, check it again periodically because it changes, and take legal advice before you design a strategy that depends on a particular answer. Do not confuse this with minimum advertised price policy, which is a manufacturer's rule about how resellers advertise a product and a separate mechanism with separate law behind it.

What You Watch Changes With the Structure

If the two structures differ, then monitoring the same way in both is a mistake, and it is the mistake most sellers make.

On a marketplace, the competitive set is handed to you. It is the other offers on the listing you are already in, it is fully visible, and it changes hour to hour because the sellers in it are running software. The monitoring problem is frequency and reaction: knowing within minutes rather than days that the offer beneath yours moved, and knowing your own floor well enough that the answer is sometimes to let the unit go. There is very little discovery work, because the platform already decided who your competitors are.

On your own store nobody hands you a list. The competitive set is a judgement call you make and revise: which four or five products a shopper would realistically hold against yours, which may sit on four different domains, on four different ecommerce platforms, at four different price points, and none of which is the identical SKU. Building that set is most of the work, and it changes slowly enough that checking it every few hours adds little. What matters is that the set is honest, and that you notice a structural move rather than a weekend promotion.

Sellers who do both need two watch lists with different logic, not one list with a single cadence applied to everything.

Running Both Watch Lists in One Place

Respot reads public product URLs, which is what makes a mixed list possible: you paste a competitor's product URL, it detects the product details automatically, and it tracks price and stock per variation rather than only at parent-product level. Extraction is browser-free and works across many platforms, so marketplace listings and independent Shopify, WooCommerce, BigCommerce or Magento stores can sit in the same list.

Alerts fire on price change and on stock change, and the check cadence is adaptive rather than an interval you set by hand, with floors of two hours on Starter, one hour on Pro and thirty minutes on Agency. The free plan covers 5 trackers with 7 days of price history and no credit card, which is enough to put three rival marketplace offers and two substitute products from your own-store set side by side and see how differently they behave over a week. Paid plans run from 100 trackers up to 2,000, with longer history at each step.

Decide Which Game Each Product Is Playing

Choosing a channel is the easy part, and plenty of businesses should be in both. The harder discipline is refusing to price them with one set of instincts.

Take your ten best sellers and sort them into two columns. For anything sold inside a shared marketplace listing, write down your floor per unit and accept that price is doing all the competitive work, so everything else you can control has to happen upstream in cost, in the fee bracket you land in, and in stock. For anything sold only on your own store, write down the four products a shopper would compare it against, and accept that your problem is more likely to be that nobody is comparing you than that someone is undercutting you.

Sellers who mix the two instincts end up matching marketplace prices on their own store for no reason, or holding a proud margin inside a listing where nobody can see the reason for it. Start tracking the prices that actually compete with yours once you know which of the two you are looking at.