Most “world’s top marketplaces” articles are useful but dangerous for decisions. They rank Amazon, Walmart, eBay, Alibaba, Marketplaces Libre, Rakuten, Etsy and TikTok Shop by reach, category, geography or brand awareness. That explains the landscape. It does not tell an agency where to put the next 40 specialist hours.
For marketplace agencies, that distinction matters. A huge marketplace can still be wrong for one client. Its unit economics can still fall apart after fees, fulfillment, returns, discounts, creator commissions and ad spend.
My stance: agencies should stop presenting marketplace expansion as a ranked list of logos. Present it as a profit portfolio. Each marketplace needs a job. One channel may be the demand engine. Another may be the margin stabilizer. Another may be a testing channel. Another may be rejected, not because it is bad, but because the client cannot serve it profitably yet.
The named mistake is what I call logo-led expansion: adding a marketplace because it looks impressive in the QBR slide. The team celebrates “now live on six channels,” but nobody can answer which SKUs are allowed to scale, which marketplace is stealing inventory from a better-margin channel, or whether the agency fee is still covered after reporting time doubles.
This guide is written for marketplace agencies in Germany, the US and cross-border teams managing clients with enough complexity to need software: multiple marketplaces, paid media, inventory pressure and margin targets. We will use the world’s top marketplaces as the starting point, but the real goal is more practical: deciding where client budget, stock and agency capacity should go next.
What the competitor content gets right
The research pattern is clear. ChannelEngine’s marketplace overview is strong at mapping the global landscape. It explains the major platforms and naturally leads into multichannel operations. Channable’s marketplace selection content is helpful for matching products to marketplace fit and feed requirements. MerchantSpring and DataHawk focus more on the tool stack agencies need once channels multiply: unified reporting, analytics, profitability signals and automated dashboards. Rithum and Productsup emphasize operational scale: product listings, inventory, product feeds, supplier onboarding and channel-ready content. Pacvue brings the retail media and commerce operations view, with media, inventory and finance signals connected.
That is the gap FiveX should own. Agencies do not just need a list of marketplaces. They need a scoring system that combines marketplace opportunity with contribution margin, ad efficiency, stock availability, returns risk and reporting effort.
The agency version of “top marketplace” is different
A brand owner may ask, “Should we sell on Amazon, Walmart or TikTok Shop?” An agency should translate that into five sharper questions:
- Can this marketplace create profitable incremental demand, or will it cannibalize existing sales?
- Do the client’s hero SKUs have enough contribution margin after marketplace fees, fulfillment and ads?
- Can operations support the channel without stockouts, late delivery or manual exception work?
- Does the marketplace provide enough data to manage bids, pricing, inventory and reporting properly?
- Can the agency service this channel at a healthy internal margin?
That last point is easy to ignore and very agency-specific. A marketplace can be profitable for the client and unprofitable for the agency if it adds messy reporting, manual reconciliation and constant firefighting. FiveX helps here by pulling marketplace, advertising, inventory and profitability data into one environment, so the agency is not rebuilding the same spreadsheet for every client every Monday morning.
A better framework: the marketplace profit portfolio
Instead of ranking marketplaces from “best” to “worst,” group them by the role they play in a client portfolio.
1. Demand engines
These are channels with serious buyer intent and enough search volume to move revenue. Amazon is the obvious example in the US and many European categories. Walmart can be a demand engine in the US for selected categories. In Germany, Amazon and Otto may play different demand roles depending on category, brand position and logistics readiness.
The danger: demand engines can hide margin leakage because the sales volume looks good. If a client sells a kitchen appliance for €79 with €27 landed product cost, €11 marketplace and fulfillment cost, €6 expected returns/warranty provision and €9 ad cost, contribution is €26 before overhead. That looks workable. But if Prime Day discounting pushes price to €67 and ad cost rises to €13, contribution drops to €10. The channel is still “top.” The SKU may no longer be allowed to scale.
2. Margin stabilizers
Some marketplaces may produce less volume but better economics. A specialist marketplace, a local platform or a B2B-friendly channel can be valuable if referral fees are lower, competition is lighter or average order value is higher. Agencies should not dismiss these channels because they are less famous.
For one client with 600 SKUs, a secondary marketplace doing only €38,000 in monthly GMV can be more attractive than a larger marketplace doing €110,000 if the smaller channel retains 18% contribution margin and the larger one retains 6%. In that case, the smaller channel contributes €6,840 while the larger contributes €6,600, before agency time. Fame did not pay the bill. Margin did.
3. Discovery channels
TikTok Shop, creator-led commerce and social marketplaces are often discovery channels. They can create demand rather than simply harvest it. That makes them exciting, but also volatile. A creator video can produce 800 orders in 48 hours. Lovely, unless the client only had 420 sellable units, the next replenishment is three weeks away and Amazon Sponsored Products were still spending as if stock was safe.
For agencies, discovery channels need stock and margin guardrails before growth tactics. FiveX can flag when ad budget, creator activity or marketplace momentum is pushing a SKU below safe stock cover or below target contribution margin. That turns social commerce from “hope the spike is good” into a controlled commercial test.
4. Learning channels
Some marketplaces are worth testing because they reveal category demand, price sensitivity or content gaps. The goal is not immediate scale. The goal is learning quickly without creating operational drag.
A good learning test might be 40 SKUs, €3,000 in controlled ad spend, a target of 100 orders, and a rule that no SKU graduates unless it reaches at least 12% contribution margin after marketplace fees and ads. That is much more useful than “let’s launch the full catalog and see what happens.” Ah yes, the classic strategy of turning the warehouse into a roulette wheel. Fun in theory. Less fun in finance.
How to score a marketplace before recommending it
Use a simple 100-point scorecard. The numbers do not need to be perfect. They need to force the right conversation.
- Demand fit — 20 points: search volume, category traffic, marketplace audience and competitive intensity.
- Margin fit — 25 points: referral fees, fulfillment costs, payment costs, returns, discounts, ad costs and expected contribution margin.
- Operational fit — 20 points: feed requirements, fulfillment model, inventory synchronization, customer service burden and returns workflow.
- Advertising fit — 15 points: retail media maturity, campaign controls, reporting quality and ability to connect spend to SKU profit.
- Data fit — 10 points: availability of order, traffic, ad, inventory and profitability data.
- Agency fit — 10 points: onboarding effort, reporting effort, required expertise and service margin for the agency.
The agency fit score is the one most teams forget. If a channel needs manual exports, custom client slides, marketplace-specific troubleshooting and weekly reconciliation, it consumes capacity. FiveX’s agency dashboards, automated reporting and profitability views are useful because they reduce that hidden servicing cost. The client sees clearer decisions. The agency protects delivery margin.
Scenario 1: a US home brand choosing between Amazon, Walmart and TikTok Shop
Imagine a US home brand with 350 SKUs, $420,000 monthly marketplace GMV and a $65 average order value. Amazon currently generates $310,000, Walmart $70,000 and TikTok Shop $40,000. The client wants to “go bigger on TikTok because it is the future.” Maybe. But the profit portfolio tells a more useful story.
Amazon has a 14% contribution margin after fees, fulfillment and ads. That creates $43,400 contribution on $310,000 GMV. Walmart has 16% contribution, creating $11,200. TikTok Shop has 7% contribution after creator commission, subsidized shipping and promotional discounts, creating $2,800.
If the agency only looks at growth rate, TikTok wins the meeting. If it looks at contribution per operational hour, the answer changes. TikTok might still deserve a test budget, but not unlimited scaling. The better recommendation is:
- Protect Amazon budget on SKUs above 12% contribution and more than 30 days of stock cover.
- Expand Walmart for SKUs where competition is lower and contribution is above 15%.
- Run TikTok Shop as a controlled discovery channel: 25 creator-ready SKUs, max $8,000 promotional support, and automatic pause rules if contribution drops below 8% or stock cover falls under 21 days.
This is where FiveX fits naturally. The agency can show SKU-level contribution margin, ad spend, order velocity and stock cover in one view instead of arguing from Amazon Ads, Walmart Seller Center, TikTok exports and a finance spreadsheet.
Scenario 2: a German electronics client considering Amazon, Otto and Kaufland
Now take a German electronics accessory brand with 1,200 SKUs and €260,000 monthly GMV. Amazon drives €190,000, Otto €45,000 and Kaufland €25,000. The client wants to list the full catalog everywhere because “more availability should mean more revenue.” This is logo-led expansion wearing a tidy blazer.
The agency audits 150 representative SKUs. It finds that 48 SKUs have enough margin and stock depth for Amazon ads, 32 are suitable for Otto because of stronger basket economics, and only 18 should be tested on Kaufland because returns handling and price competition eat margin on low-priced accessories.
One USB-C hub sells for €39.99. Landed product cost is €12.40. Marketplace and fulfillment cost is €7.80 on Amazon, €6.90 on Otto and €6.50 on Kaufland. Expected return cost is €2.20. Average ad cost per order is €5.60 on Amazon, €3.80 on Otto and €4.40 on Kaufland. Contribution becomes €11.99 on Amazon, €14.69 on Otto and €14.49 on Kaufland before overhead. Otto is smaller, but for this SKU it deserves attention.
Another cable set sells for €12.99 and looks like an easy catalog filler. After costs and returns, it retains €1.10 on Amazon, €1.70 on Otto and €0.60 on Kaufland before ads. One sponsored campaign can wipe that out. The recommendation is not “launch everywhere.” It is “do not spend agency time scaling low-ticket SKUs unless bundled AOV improves.”
The metrics agencies should report in every marketplace portfolio review
A good QBR should not only show sales by marketplace. It should show the commercial role of each marketplace and whether that role is still valid. I would include:
- GMV and orders by marketplace.
- Contribution margin by marketplace and by top SKU.
- Ad spend, ACOS, TACOS and contribution after ads.
- Stock cover for promoted SKUs.
- Return rate and return cost by marketplace.
- Buy Box or offer availability for key SKUs.
- Agency servicing time by marketplace, especially during onboarding.
FiveX product hook number three: this is exactly the kind of reporting that should be automated. If an account manager spends four hours assembling a marketplace portfolio deck, that cost comes out of agency margin. If FiveX turns the same data into a repeatable dashboard with AI recommendations, the team can spend the time on decisions: pause this SKU, shift budget there, fix inventory before the next creator push.
What agencies should be careful with
Do not confuse integration with readiness
Being technically able to connect a marketplace does not mean the client is commercially ready. A feed tool can push products live. A marketplace integrator can sync orders. That is necessary, not sufficient. Readiness means pricing, stock, fulfillment, content, ads and reporting are commercially controlled.
Do not let retail media outrun operations
Retail media can accelerate the wrong SKU just as efficiently as the right one. If campaigns do not see margin and inventory, they will spend into stockouts and low-profit products. Agencies need rules that connect ad pacing to commercial reality.
Do not treat every marketplace as a full-service retainer
Some channels deserve deep management. Others deserve light monitoring. A learning channel with 40 SKUs should not require the same weekly reporting package as Amazon. Build service tiers per marketplace role, or your internal delivery model becomes heavier every time the client expands.
A practical 30-day agency workflow
Here is the workflow I would use before recommending a new marketplace or changing channel priorities.
Week 1: build the marketplace profit baseline
Pull the last 90 days of orders, fees, ad spend, returns and inventory. Calculate contribution margin by SKU and marketplace. If finance data is not available at SKU level, use conservative assumptions and mark the confidence level. Do not hide shaky inputs. Clients appreciate honesty more than beautiful nonsense.
Week 2: segment SKUs by role
Create four groups: scale, protect, test and exclude. Scale SKUs have margin, stock and demand. Protect SKUs are profitable but stock-constrained. Test SKUs have potential but need controlled spend. Exclude SKUs cannot profitably survive the marketplace economics.
Week 3: score marketplace roles
Apply the 100-point scorecard. Assign each marketplace a role: demand engine, margin stabilizer, discovery channel or learning channel. Then define budget, stock and reporting rules for each role.
Week 4: present the decision, not the data dump
The client does not need 46 screenshots. They need a recommendation: “Amazon remains the demand engine, Walmart gets incremental budget on 60 SKUs, TikTok Shop stays capped until creator economics improve, and Kaufland is a 30-SKU test only.” Show the numbers behind it, but lead with the decision.
Where FiveX helps marketplace agencies
FiveX is not another logo in the marketplace tool stack. For agencies, the value is connecting the signals that normally live in different systems:
- Marketplace analytics: see performance across Amazon, bol, Mirakl retailers, Walmart and other channels without rebuilding exports.
- Profitability dashboards: connect revenue, fees, fulfillment, ads, returns and margin at SKU level.
- Advertising automation: manage spend with guardrails based on contribution margin, TACOS, stock and campaign performance.
- AI recommendations: surface which SKUs need budget shifts, pricing review, inventory action or campaign pauses.
- Agency reporting: turn cross-marketplace data into repeatable client reporting so specialists spend less time collecting numbers and more time improving them.
That matters because the best agency advice is rarely “sell on the biggest marketplace.” It is: “sell the right SKUs on the right marketplace with the right budget while protecting margin and capacity.”
Final takeaway
The world’s top marketplaces are a menu, not a to-do list. Agencies create value by choosing what not to order.
Amazon, Walmart, eBay, Etsy, Otto, Kaufland, Marketplaces Libre, TikTok Shop and specialist marketplaces can all be excellent in the right portfolio role. But none of them are automatically right for every client or every SKU.
If you want a practical rule, use this: a marketplace is only “top” when it improves the client’s profit portfolio without breaking operations or delivery margin. Everything else is just a bigger logo on the slide.