Managing one marketplace ad campaign is optimization. Managing fifty is air traffic control.
That sounds dramatic until the account reaches the normal mess: twelve Amazon Sponsored Products campaigns, three Sponsored Brands tests, eight bol Sponsored Products campaigns split across NL and BE, four MediaMarkt retail media pushes, a handful of competitor-targeting experiments, paused launch campaigns that nobody trusts anymore, and a few “temporary” budget caps that have been temporary for six months.
The named mistake I see is campaign tab management. The operator opens the platform, sorts by spend or ACOS, fixes whatever looks loudest, exports a report, then repeats the same routine next week. The account looks active. Bids move. Keywords get paused. Budgets get nudged. But nobody can answer the question that matters: which campaigns are part of the working profit engine, which are still learning, which are protecting shelf space, and which are just old decisions with fresh spend?
My stance: once an Amazon, bol or MediaMarkt account has more than roughly 25 active campaigns, it needs a campaign fleet map before it needs more bid tweaks. If monthly spend is above about €5K, the problem is rarely that the team cannot find a CPC to change. The problem is that the campaign estate has no operating map, so every change competes for budget, margin and attention without a shared rulebook.
What the existing advice gets right
The public advice around Amazon PPC campaign structure is useful. BidX recommends a controlled setup with separate Auto, Broad, Exact and ASIN/product-targeting campaigns so budget can be managed at campaign level and search terms can be graduated cleanly. That is a strong foundation. Their point about using negatives to prevent the same term from being served across multiple campaigns is especially important when an account starts to scale.
Podean takes the broader agency view. Their marketplace media positioning connects advertising with non-media signals such as out-of-stock data, low stock levels, price changes and merchandising tactics. That is exactly the direction mature retail media needs to move: media decisions should not live in a little PPC bubble.
Reddit threads from Amazon sellers add the operator pain. Sellers ask how to manage 100+ campaigns, how to recover from messy accounts, and why one campaign or ASIN keeps swallowing the budget. The practical answers usually come back to consistent structure, tags, bulk actions, clear budgets and not making emotional decisions from three days of data.
All of that is helpful. But it still stops before the hardest question for a managed Advertentie Service account: how do you govern the campaign fleet as a portfolio?
The missing layer: campaign structure is not campaign governance
Campaign structure decides where keywords, ASINs, match types and budgets live. Campaign governance decides what those campaigns are allowed to do.
That difference matters. A perfectly structured exact-match campaign can still be dangerous if it is defending a low-margin SKU with 11 days of stock left. A clean auto campaign can still be wasteful if it keeps rediscovering search terms that another campaign already proved unprofitable. A MediaMarkt placement can look like a great retail media test until the promoted variant sells out and the clicks land on weaker colours.
A campaign fleet map adds five fields that native ad platforms rarely force you to define:
- Role: defend, harvest, learn, launch, recover, clear stock or test placement.
- Permission: active, capped, learning, frozen, restart-only or awaiting commercial approval.
- Economic owner: SKU, product group, marketplace country or promotion.
- Constraint: margin floor, stock cover, offer eligibility, price position, return reserve or content readiness.
- Review date: the moment this campaign must re-earn its right to spend.
FiveX’s advertising view is built around that operating reality. Campaign performance is useful, but it becomes commercially useful when it sits next to SKU profitability, inventory runway, product strategy and change logs. That is the difference between “ACOS is 24%” and “this campaign may keep spending because the SKU still has a 38% contribution margin, 42 days of stock and a harvest role that expires next Monday.”
The 50-campaign failure pattern
Most accounts do not become messy because one person made a terrible decision. They become messy because many reasonable decisions never expire.
A launch campaign stays live after the launch. A defensive brand campaign keeps its old budget even after organic rank improves. A broad-match campaign is supposed to learn, but nobody defines what learning is worth. A bol BE test is copied from NL because the product is the same, even though the margin is not. A MediaMarkt campaign is left active after a promotion because nobody owns the retail-media aftercare.
The result is not chaos in the obvious sense. The dashboards still load. The names may even look tidy. The problem is silent: the account becomes a museum of old decisions.
The fix is to stop asking only “which campaign performed?” and start asking “which campaign still has permission?”
Example 1: the Amazon kitchen account with 52 campaigns
Imagine a kitchenware brand spending €9,800 per month on Amazon Ads. It has 52 active campaigns across 18 SKUs. The platform view says the account is healthy: blended ACOS is 23%, ROAS is 4.3 and sales are growing.
The fleet map tells a sharper story. Seven campaigns drive 64% of spend. Three of those campaigns are defending a silicone lunch box set with a selling price of €29.95, landed cost of €8.10, FBA and referral costs of €8.40, and a return reserve of €1.20. Before advertising, the SKU has €12.25 contribution margin. At a 4.8% conversion rate, the real max CPC is roughly €0.59 if the brand wants to keep €3.00 contribution after ads.
Two exact-match campaigns are averaging €0.54 CPC and can stay in harvest. One broad campaign is averaging €0.71 CPC and has not produced a new graduating term in 21 days. In the old workflow, the operator might lower bids by 10% and move on. In the fleet-map workflow, that campaign changes role from learn to capped-maintenance, receives a €12 daily ceiling, and gets a review date seven days later. The saved budget moves into a product-targeting test for a higher-margin glass container SKU with 58 days of stock.
The account did not need fewer campaigns. It needed role expiry.
Example 2: the bol NL/BE copy-paste trap
A home-office brand spends €6,400 per month on bol Sponsored Products. The team has 31 campaigns: NL search, BE search, category targeting and a few seasonal pushes. NL performance looks strong, so the agency copies the same structure to Belgium.
That sounds efficient. It is also the classic copy-paste trap.
The best-selling desk lamp has a €49.95 selling price in NL and a €47.95 effective price in BE after local shipping and promo pressure. Pick-pack and fulfilment are €1.40 higher in Belgium. Returns are also 3 percentage points higher. The NL campaign can tolerate a 27% ACOS. The BE campaign breaks even closer to 19%.
Without a fleet map, the BE campaigns inherit NL bids and budgets because the product is “the same”. With a fleet map, country is part of the economic owner. NL exact campaigns get a harvest role. BE campaigns get a learn role with a €350 monthly evidence wallet, a lower CPC ceiling and a rule that no term graduates until it proves contribution margin after local fulfilment. FiveX’s channel and product reporting makes that split visible, so the team does not hide BE leakage inside a blended bol result.
The campaign structure can be copied. The permission cannot.
Example 3: the MediaMarkt electronics launch that should not scale yet
Now take an electronics accessory brand testing MediaMarkt retail media with a €2,000 launch allocation inside a wider €14,000 marketplace budget. The hero product is a USB-C charging dock. Early numbers look promising: €620 spend, €3,400 attributed revenue and a 5.5 ROAS.
A normal dashboard says scale. The fleet map says wait.
Why? The black variant has 16 days of stock, the silver variant has 43 days, and the retail media placement is sending most clicks to the parent listing where shoppers choose black first. The black SKU has only €7.80 contribution margin before ads because of a temporary component-cost increase. The silver SKU has €12.40. Scaling the campaign without variant-level stock and margin control would buy demand for the weaker operational path.
The fleet-map decision is specific: keep the campaign active, cap spend at €70 per day, shift creative and product emphasis toward silver, and do not release the second €1,000 tranche until black stock is above 30 days or the listing can route demand more cleanly. FiveX inventory insights and SKU profitability are the natural guardrails here; the ad platform alone cannot see the full constraint.
How to build the campaign fleet map
You do not need a 40-column spreadsheet to start. You need a map that every operator, founder and finance lead can understand in five minutes.
1. Group campaigns by job, not by platform menu
Native platforms group by campaign type: Sponsored Products, Sponsored Brands, auto, manual, exact, broad, product targeting. Keep those fields, but do not make them the main decision layer. The main layer should be the job: defend brand demand, harvest proven non-brand demand, learn new queries, launch a SKU, recover after stockout, clear aged inventory or test a marketplace placement.
2. Attach every campaign to an economic owner
A campaign without an economic owner becomes hard to judge. Is it responsible for one SKU, a product family, a country, a promotion or a strategic marketplace test? If nobody knows, the campaign will be evaluated on blended ACOS. That is how weak products hide behind strong ones.
3. Give each role a budget rule
Defend and harvest campaigns can earn stable budget if margin and stock stay healthy. Learn campaigns need a fixed evidence wallet. Launch campaigns need tranche releases. Recovery campaigns need restart rules. Clear-stock campaigns need a margin exception and an end date. The budget rule should be written before the bid changes.
4. Add constraint vetoes
Every campaign should know what can stop it: stock under 21 days, contribution margin below the floor, Buy Box or offer loss, price position worse than agreed, return rate above threshold, content not ready, or promotion ended. FiveX’s Advertentie Service uses these vetoes to stop “good-looking” platform performance from overruling business reality.
5. Review the map weekly
The weekly review should not start with the biggest ACOS swing. Start with permissions. Which campaigns expire this week? Which learning campaigns bought enough evidence? Which harvest campaigns lost margin? Which restart campaigns can graduate? Which old campaigns should be archived because their job no longer exists?
The operator voice: fewer heroic optimizations, more boring permissions
There is a slightly uncomfortable truth here: a campaign fleet map makes advertising feel less heroic. It turns “I found a bid opportunity” into “this campaign has the right to spend because its role, margin, stock and evidence still match.”
Good. Marketplace advertising at €5K+ monthly spend should be a little boring. Boring permissions protect the budget. Boring review dates prevent zombie campaigns. Boring change logs explain why the team moved money. Boring SKU profitability checks stop the account from celebrating revenue that finance will later dislike.
That is where FiveX fits naturally. Our platform brings marketplace sales, ads, SKU profitability, inventory and product performance into one operating layer. Our Ads AI and automation can recommend bid changes, but the important part is that those recommendations can be reviewed against product strategy, stock and profit context. And for brands using FiveX Advertentie Service, the managed process turns that context into weekly budget decisions across Amazon, bol and MediaMarkt instead of leaving each platform tab to fight for itself.
Final thought
If you manage five campaigns, campaign structure may be enough. If you manage fifty, structure is only the skeleton. You also need a nervous system.
The campaign fleet map is that nervous system. It tells the team what every campaign is for, what it may spend, what can stop it, when it expires and which business constraint it must respect. Without it, a growing marketplace ad account becomes a pile of plausible optimizations. With it, Amazon, bol and MediaMarkt campaigns become a portfolio of controlled profit decisions.
And yes, it is less exciting than another “10 PPC hacks” list. Lovely. Profit usually is.