Marketplace Selling & Growth Guide
A marketplace adds operational load immediately and revenue eventually. This guide covers how to work out whether a channel fits your business at all, what the fees actually do to your margin, how listings are ranked and advertised, and what running one properly asks of your team once the launch excitement has worn off.
CHAPTERS
8 chapters
READING TIME
38 min read
LAST UPDATED
19 September 2026
LEVEL
Intermediate
The decision that matters most is made before the first listing goes live: whether this channel should exist for you at all.
EXECUTIVE SUMMARY
What this guide argues, in five points
Marketplaces are usually sold as a growth channel and bought as a distribution decision, which is why so many launches stall in month four. They are a different business with a different margin, a different customer relationship and a different operational shape, running alongside the one you already have.
- Work out contribution after marketplace fees before anything else. A product that funds its own acquisition on your store can be loss-making on a marketplace at the same price.
- Operational load arrives on day one. Revenue arrives in month four, if the catalogue and the ranking work were done properly.
- On a marketplace the listing is the landing page, the search engine and the advert all at once. Listing quality caps everything downstream.
- Paid and organic are coupled: advertised sales feed ranking, which is why launch periods justify targets you would never accept as a steady state.
- You do not own the customer. Plan the unit economics without the repeat value you get on your own store, and treat anything you do get as upside.
The eight chapters below are the detail behind those five points, in the order we work through them when a store asks us about a new channel.
- Contents
Eight chapters
Read it in order the first time. After that, the chapter index is the useful part.
01
Five tests, answered in a spreadsheet, before anybody argues about which channel to open.
5 min
02
Six cost lines, not one. Model them per product, because the answer differs between two items on the same shelf.
5 min
03
Not a ranking by size. Where your category’s buyers already are, how crowded the shelf is, and what each option asks of operations.
4 min
04
The listing is the landing page, the search result and the advert at once. It caps everything downstream.
6 min
05
Marketplace search optimises for the probability that this shopper buys now. That narrows what works considerably.
5 min
06
Advertised sales feed rank, rank feeds organic sales. Paid and organic cannot sensibly be judged apart on a marketplace.
4 min
07
The load nobody quotes for. Any one of these can stop a channel earning while every marketing number still looks fine.
5 min
08
Gross revenue is the most misleading number on a marketplace. Contribution after fees is the only one worth reporting.
4 min
- Before you start
Who this guide is for
Written for the people who will carry the channel after the launch excitement wears off.
Founders and heads of eCommerce
You are deciding whether to open a marketplace at all, and want the margin answer before the strategy conversation.
Marketplace and channel managers
You are running one or more channels and want a structure for listings, advertising and account health that holds up.
Operations and supply chain leads
Chapter seven is your half, and it decides whether any of the rest of the guide matters.
NOT WRITTEN FOR
People looking for a quick revenue channel. A marketplace adds operational load on day one and revenue in month four at the earliest. If the base business is not stable, this is not the fix.
CHAPTER 01
5 min
Whether a marketplace fits you at all
Five tests, answered in a spreadsheet, before anybody argues about which channel to open.
The first question is not which marketplace. It is whether a marketplace suits what you sell, at the margin you sell it on, with the operation you currently have. Several genuinely good businesses should not open one, and finding that out in a spreadsheet is considerably cheaper than finding it out in month six.
Five tests, before the channel question
Test
Passes when
Fails when
Margin headroom
Contribution survives referral and fulfilment fees with room left
The product only works at full price on your own store
Demand exists on the channel
People already search for this category inside the marketplace
You would be creating demand on somebody else’s platform
Catalogue data is ready
Attributes, images and identifiers complete
Half the range has no dimensions and one photo
Operations can absorb it
Stock, dispatch and returns can take another channel
You are already missing dispatch cut-offs on one channel
Brand control is acceptable
You can live with the presentation and the competition next to you
Price presentation and adjacency matter commercially
Fail two of these and the launch will consume a quarter and produce a support queue. Fail margin alone and it does not matter how well everything else scores.
KEY TAKEAWAY
A marketplace multiplies operational load immediately and revenue eventually. If the base business is not stable, a second channel makes an existing problem more expensive rather than diversifying it.
What a marketplace is actually good at
- Reaching buyers who start their search inside it, which in several categories is now most of them.
- Removing the trust barrier for a brand nobody has heard of, at the cost of the relationship afterwards.
- Testing demand cheaply, in a new country or a new category, without building anything.
- Absorbing surplus stock without discounting your own storefront in public.
What it is not good at is building a brand you own, or producing customers you can market to again. Both of those stay on your own store, which is why the two channels should be planned together rather than traded off against each other.
CHAPTER 02
5 min
The economics: fees, margin and price
Six cost lines, not one. Model them per product, because the answer differs between two items on the same shelf.
Marketplace economics are not your economics with a commission on top. Several costs change at once, and some of them do not appear on an invoice at all. Model the whole stack per product before committing, because the answer frequently differs between two products in the same range.
The cost stack, in full
Referral or commission fee
Percentage of sale price
Varies by category, not by seller
Fulfilment fee
Per unit, by size and weight
Where bulky low-value products die
Storage
Per unit per month, often seasonal
Punishes slow-moving stock hard
Returns handling
Per return, plus unsellable units
Category-dependent and routinely underestimated
Advertising
Effectively mandatory for visibility
Treat as a cost of sale, not a growth budget
Account and software
Subscription, integration, listing tools
Small, but real, and usually forgotten
Six lines, not one. Model them per product: the answer often differs between two items on the same shelf.
KEY TAKEAWAY
Recalculate break-even ROAS using contribution after marketplace fees. It is a materially different number from the one on your own store, and using the wrong one is how advertising budgets get approved that cannot pay back.
Price, and the trap underneath it
- Pricing to protect your own store usually means pricing above the marketplace norm, which means low visibility. Decide deliberately which channel is the price leader.
- Automated repricing races to the floor unless it is bounded by a contribution figure rather than a competitor’s number.
- Bundles and multipacks are the honest way to change the comparison, and they also improve fulfilment economics on small items.
- Watch the margin after promotions and coupons, which stack with fees and are the most common cause of a profitable-looking channel losing money.
The output of this chapter is a short list: which of your products should be on a marketplace at all. On most catalogues it is a minority, and that is a good outcome rather than a disappointing one.
CHAPTER 03
4 min
Choosing the channel
Not a ranking by size. Where your category’s buyers already are, how crowded the shelf is, and what each option asks of operations.
Once the economics say yes, the channel question becomes answerable. It is not a ranking of marketplaces by size; it is a question of where your category’s buyers already are, how crowded that shelf is, and how much operational difference each option introduces.
The four questions that decide it
1
Where does this category’s demand already sit? Search inside each candidate channel for your own products before assuming the largest is the right one.
2
How crowded is the shelf, and by whom? A category dominated by the marketplace’s own brand is a different proposition from one dominated by small sellers.
3
What does this channel ask of operations? Dispatch cut-offs, service level requirements, label formats and returns policy vary considerably.
4
What is the cost of being wrong? A channel you can close in a month is a cheaper experiment than one requiring inventory to be shipped into its network first.
KEY TAKEAWAY
Open one channel properly before opening a second. Two half-run marketplaces reliably produce less than one run well, and they produce twice the support load while doing it.
Fulfilment: theirs or yours
Model
Argues for itself when
Costs you
Marketplace fulfilment
Speed badges matter and units are small and fast-moving
Margin, stock visibility, and storage fees on slow lines
Seller fulfilment
Products are bulky, slow-moving or need care in packing
Ranking advantage, and you must hit their service metrics
Split by product
Your range genuinely has both shapes
Complexity — but usually the right answer at scale
Model this per product alongside chapter two. Fulfilment choice changes contribution more than price does on many items.
Our Amazon versus Walmart Marketplace comparison works through where the demand actually is and what each channel asks of a team.
CHAPTER 04
6 min
Catalogue and listing quality
The listing is the landing page, the search result and the advert at once. It caps everything downstream.
On a marketplace the listing is the landing page, the search result and the advert at the same time. It caps everything downstream: ranking, advertising efficiency, conversion and returns. More marketplace programmes fail on listing quality than on any other single thing, and it is the cheapest part to fix.
Catalogue data is not admin. On a marketplace it is the product page, the targeting and the shelf position.
What a listing has to carry
Element
What it decides
Common failure
Title
Which searches you are eligible for
The internal product name, missing the attributes buyers search by
Main image
Click-through rate from the search grid
A lifestyle crop where the category convention is product on white
Supporting images
Conversion and returns rate
Too few, no scale reference, no detail or in-use shot
Attributes and category
Filter eligibility and classification
Left blank, so the listing never appears in a refined search
Bullets and description
Objection handling at the decision point
Adjectives where dimensions and compatibility should be
Identifiers and variation family
Matching, and whether reviews pool correctly
Variants listed separately, splitting reviews across children
A title and image rewrite across a catalogue routinely moves marketplace revenue more than any bid change.
KEY TAKEAWAY
Write titles in the order buyers search: brand, product type, then the attributes that distinguish it. Then check how the title truncates on a phone, because that is where most of it will be read.
Reviews, and the thing nobody tells you
- Reviews are the ranking and conversion asset you least control and the one that most decides both. Earn them through the product and the packing insert, within the channel’s rules.
- Never buy or incentivise reviews. Suspension risk aside, it corrupts the only honest signal you have about the product.
- Read the one-star reviews as a product brief. They name the expectation your listing is setting wrongly, which is a returns cost and a conversion cost at once.
- Structure variations correctly from the start. Merging families later is painful, and splitting reviews across children is a self-inflicted ranking problem.
Our marketplace listing template holds titles, bullets, attributes and character limits per channel in one working file.
CHAPTER 05
5 min
Ranking inside a marketplace
Marketplace search optimises for the probability that this shopper buys now. That narrows what works considerably.
Every marketplace runs a search engine, and it optimises for something narrower than a web search engine does: the probability that this shopper buys this item now. That single difference explains most of what does and does not work when trying to rank inside one.
What marketplace ranking actually rewards
- Relevance from structured data. Title, attributes and category, which is why chapter four comes first. You cannot rank for a term you are not eligible for.
- Conversion rate on the listing. Images, price, reviews and availability. A listing that converts gets shown more, which is a compounding loop in both directions.
- Sales velocity, recent more than lifetime. This is why launches need a deliberate push and why going out of stock is more damaging here than on your own store.
- Fulfilment promise. Speed and reliability badges are a ranking input, not just a trust signal.
- Account health. Late dispatch, defects and cancellations suppress visibility before they ever trigger a warning.
KEY TAKEAWAY
Going out of stock costs ranking that took months to build and does not come back when the stock does. Stock planning on a marketplace is a marketing decision.
The launch problem, and how to handle it
A new listing has no sales history, so it ranks nowhere, so it makes no sales. Breaking that loop is what a launch budget is for, and it is the one period where a deliberately unprofitable target is a defensible decision rather than a mistake.
1
Get the listing right before spending anything. Advertising a bad listing buys expensive proof that it converts badly.
2
Fund a defined launch window, with a budget and an end date agreed in advance, treated as an investment in rank rather than as performance media.
3
Drive external traffic carefully where the channel rewards it, and only to a listing that is ready to convert.
4
Hold stock through the window. Running out mid-launch wastes the entire spend.
5
Judge the window on rank and organic share afterwards, not on the ROAS inside it.
CHAPTER 06
4 min
Retail media, coupled to ranking
Advertised sales feed rank, rank feeds organic sales. Paid and organic cannot sensibly be judged apart on a marketplace.
Marketplace advertising is not a growth channel bolted onto a shelf. It is part of how the shelf works. Advertised sales feed ranking, ranking feeds organic sales, and organic sales reduce how much advertising you need — which means paid and organic cannot sensibly be judged separately.
A structure that survives
- Separate by intent. Defending your own brand terms, competing on category terms and discovery are three different jobs with three different acceptable costs.
- Separate by contribution tier, exactly as on any other paid channel. One target across a mixed-margin range overspends on the thin products.
- Separate launch from steady state. A launch campaign with a deliberately loose target should never sit in the same reporting line as a mature one.
- Keep a harvesting loop. Search terms that convert become their own targeted campaigns; terms that do not become negatives. This is still manual and still where the waste is.
KEY TAKEAWAY
Judge marketplace advertising on total channel contribution, not on advertised cost of sale alone. An account can improve its advertising metric while the channel earns less, simply by advertising less on listings that were ranking anyway.
The number that matters
Advertised cost of sale describes the advertising, not the business. The useful figure is total advertising cost divided by total channel revenue — paid and organic together — because that is the one that tells you whether the channel is becoming less dependent on spend over time. On a healthy account it falls as ranking builds.
Our article on cutting Amazon ACoS without losing volume covers the campaign architecture behind this chapter in detail.
CHAPTER 07
5 min
Operations, returns and account health
The load nobody quotes for. Any one of these can stop a channel earning while every marketing number still looks fine.
The part of marketplace selling that nobody quotes for is the operational load. Stock has to be in the right place, dispatch has to hit somebody else’s cut-off, returns arrive against a policy you did not write, and account health metrics decide whether any of the previous chapters matter at all.
A marketplace holds you to its service levels, not yours. That is an operations decision before it is a marketing one.
The four operational realities
Area
What changes
What it breaks if ignored
Inventory
Stock is committed per channel, and often physically moved
Overselling on one channel, dead stock on another
Dispatch
Their cut-off, their carrier requirements, their tracking
Late dispatch metrics, which suppress ranking quietly
Returns
Their policy applies, not yours, and often at your cost
Margin, and a returns rate that flags the listing
Account health
Defect, cancellation and response-time metrics
Visibility first, then suspension — usually without warning
Any one of these can stop a channel earning while every marketing number still looks fine.
KEY TAKEAWAY
Account health is a marketing asset. Suppression from a late-dispatch metric costs more visibility than any advertising budget will buy back, and it is invisible in the advertising report.
Keeping one catalogue, not three
- Hold one source of truth for product data and push to channels from it. Three hand-maintained catalogues diverge within a quarter, every time.
- Automate stock sync before launch, not after the first oversell. Overselling costs an account metric as well as a customer.
- Keep pricing rules in one place, bounded by contribution rather than by a competitor’s price.
- Decide who owns the channel internally. A marketplace with no named owner degrades slowly and invisibly, which is the most common way we find them.
Our marketplace launch checklist covers the five readiness gates and the ninety-day sequence before a channel opens.
CHAPTER 08
4 min
Measuring a marketplace honestly
Gross revenue is the most misleading number on a marketplace. Contribution after fees is the only one worth reporting.
Marketplace reporting flatters in a specific way: the channel reports revenue gross of fees, advertising reports its own cost of sale, and neither of them tells you what the channel contributed after everything. Building the honest view takes an afternoon and changes the decisions that follow it.
Report channel contribution after fees. Gross revenue on a marketplace is a particularly misleading number.
The four views worth building
View
Question it answers
Cadence
Channel contribution after all fees
Did this channel make money?
Monthly
Total ad cost over total channel revenue
Is dependence on advertising falling?
Monthly
Organic share of sales by product
Is ranking actually building?
Monthly
Contribution per product after fees
Which items should still be listed?
Quarterly
The fourth view usually shortens the catalogue, which is a good outcome. Delisting a loss-making product is a margin improvement disguised as a revenue cut.
KEY TAKEAWAY
Never sum marketplace revenue with storefront revenue when judging marketing performance. Different fee structures and different customer ownership make the combined number meaningless.
What to expect, and when
- Month one: listings live, operational problems surface, almost no revenue. This is normal and is not a signal.
- Months two to three: early reviews, first ranking movement, advertising doing most of the work.
- Months four to six: organic share becomes visible; this is the first honest read on whether the channel works.
- Beyond six months: either advertising dependence is falling and contribution is positive, or the channel is a fee-heavy way to sell at a loss. Both answers are useful.
EXPERT VIEW
“The most valuable thing a marketplace programme produces in its first year is not revenue. It is a defensible answer to whether this channel should be part of the business at all — and we would rather deliver that in month five than in year two.”
Dazzle Commerce · Marketplace team
- Worked examples
Three decisions, before and after
The three arguments that come up on almost every marketplace account we take over.
CATALOGUE SCOPE
Listing the whole range
Before
Full catalogue uploaded, including bulky low-value lines
After
Range cut to the products that clear contribution after fulfilment and storage fees
The bulky lines were paying more in fulfilment than they made. Removing them raised channel contribution while revenue fell, which is the right trade.
LISTINGS
Bidding harder on a weak listing
Before
Advertising budget raised on listings with one image and empty attributes
After
Titles, images and attributes rebuilt first; budget held flat through the rebuild
The listings were not losing auctions, they were converting badly once shown. Advertising a bad listing buys expensive proof that it is a bad listing.
REPORTING
Judging the channel on gross revenue
Before
Marketplace reported alongside storefront revenue in one growth number
After
Channel contribution after referral, fulfilment, storage, returns and ad cost
Gross revenue made the channel look like the fastest-growing part of the business. After fees it was the least profitable, and the budget moved accordingly.
Anonymised from accounts we manage. The shapes recur; the numbers will not match yours exactly.
- Take these with you
Three checklists
Written to be worked through rather than read. Nothing here needs a tool you do not already have.
Before you open a channel
Half a day with a spreadsheet.
- Contribution modelled per product after all six fee lines
- Break-even ROAS recalculated on post-fee contribution
- Demand checked by searching your own category inside the channel
- Catalogue data completeness audited against the channel requirements
- Fulfilment model chosen per product, not for the whole range
- A named internal owner for the channel agreed
Every listing
Run this on the top twenty by revenue.
- Title ordered brand, product type, then distinguishing attributes
- Title checked for truncation on a phone
- Main image matches the category convention
- Supporting images include scale, detail and in-use shots
- Attributes and category fully populated
- Variations grouped into one family so reviews pool
Monthly account review
An hour with the seller dashboard open.
- Account health metrics inside threshold, causes fixed not excused
- Stock cover checked against velocity on ranked listings
- Returns reasons read and fed back into listings
- Search terms harvested into campaigns and negatives
- Organic share of sales tracked per product
- Channel contribution after all fees reported to the business
Copy these into your own tracker. There is no download form and no email wall.
- Frameworks
Two frameworks worth keeping
The two decisions that get reargued every quarter, written down so they stop being reargued.
FRAMEWORK ONE
The should-this-product-be-listed test
Most catalogues should not be listed in full. Score each product on four tests and list only what passes all four. On a typical range this shortens the catalogue considerably, and raises channel contribution while doing it.
Test
Pass condition
If it fails
Contribution after fees
Positive with room for advertising
Do not list it
Fulfilment economics
Size and weight do not eat the margin
List it seller-fulfilled, or not at all
Demand on the channel
People search for this category here
Park it; you are not going to create demand here
Data completeness
Attributes, images and identifiers ready
Fix the data first — it is cheaper than advertising around it
Delisting a loss-making product is a margin improvement that shows up in the reporting as a revenue cut. Say so in advance.
Re-run this quarterly. Fee changes and fulfilment rate changes move products across the line without warning.
FRAMEWORK TWO
The marketplace priority order
When a channel needs everything doing, the order matters more than the list. Each step assumes the one above it has cleared, because advertising at step six is wasted while step one is wrong.
1
Economics. Contribution per product after all six fee lines. This decides what is listed at all.
2
Operations. Stock sync, dispatch against their cut-off, returns handling. A channel with broken operations cannot be marketed out of trouble.
3
Account health. Metrics inside threshold, because suppression removes every other advantage silently.
4
Listing quality. Titles, images, attributes and variation structure. Caps ranking, advertising and conversion together.
5
Ranking. Velocity, reviews and availability, with a funded launch window where one is needed.
6
Retail media. Structured by intent and contribution tier, judged on total channel contribution.
7
Expansion. A second channel or a second country, only once the first is stable.
Most failing marketplace accounts we inherit started at step six and have never been through steps one to three.
- In summary
Key takeaways
Eight sentences you could hand to somebody who is not going to read the guide.
1
Model contribution after all six fee lines before choosing a channel. A product that works on your store can be loss-making on a marketplace at the same price.
2
Operational load arrives on day one; revenue arrives in month four at the earliest. Plan the cash and the staffing for that gap.
3
Most catalogues should not be listed in full. Score each product and list only what passes all four tests.
4
The listing is the landing page, the search result and the advert at once. Fix it before spending anything on advertising it.
5
Marketplace search optimises for the probability of a purchase now, which is why velocity, reviews and availability outrank clever copy.
6
Going out of stock costs ranking that took months to build. Stock planning on a marketplace is a marketing decision.
7
Judge advertising on total channel contribution, and watch total ad cost over total channel revenue fall as ranking builds.
8
Never sum marketplace and storefront revenue when judging marketing. Different fees and different customer ownership make the total meaningless.
If you want a single first action: take your three best-selling products and calculate what they would contribute after referral, fulfilment, storage and returns fees.
- Go deeper
Where this work sits in the practice
Three services that pick up the chapters above.
Listings, pricing, retail media and account health run as one programme across every channel you sell on.
The catalogue, stock and fulfilment work chapter seven puts ahead of any advertising budget.
When the next channel is a new country rather than a new marketplace, and landed cost decides it.
- Keep reading
Related guides
The rest of the guide library, all published in full — plus the FBA profit and Amazon TACoS calculators if you want this guide’s arithmetic done for you.
Complete guide · 38 min
When the next channel is a new country: landed cost, operating model and localisation.
Complete guide · 38 min
Retail media in the context of every other paid channel, with the profit maths that sets the targets.
Complete guide · 37 min
Where a second sales channel sits in the order of work, and when it is the wrong next move.
Want this run on your account rather than read about it?
Our marketplace engagements open by modelling contribution after fees per product, then fixing the catalogue before anybody buys an advert.