Marketplace ad management is full of signals. ACOS moves. TACoS drifts. CPCs rise. Stock cover falls. A competitor cuts price. bol visibility improves. MediaMarkt retail media clicks get more expensive. Amazon search terms suddenly start converting again after weeks of silence.
The hard part is not seeing those signals. Most €5K+ ad accounts have enough dashboards to spot them. The hard part is deciding which signal is allowed to move money.
The named mistake I see is treating every fresh signal as a bid instruction. A campaign shows a 19% ACOS for three days, so the operator raises bids. Nobody checks whether the margin file is current, whether the SKU has 14 days of stock left, whether last week’s promotion is still influencing conversion, whether the search term already had enough learning budget, or whether bol and MediaMarkt need the same product more urgently. The signal was real. The decision was under-governed.
My stance: every managed Advertentie Service account above roughly €5K monthly spend needs a profit signal ledger. Not another report. A working ledger that says which signals matter, how fresh they are, who owns them, what action they permit, and when they expire. Without that layer, operators end up optimising the loudest metric instead of the most profitable constraint.
This guide is for brands advertising on Amazon, bol and MediaMarkt in the Netherlands and Belgium. The goal is simple: stop ACOS, ROAS, stock, margin and competitor signals from fighting each other in Slack, and turn them into one clean spend-permission system.
What competitor advice gets right
The public guidance around marketplace advertising has improved. BidX explains ACOS, break-even ACOS and target ACOS clearly, including why sellers must understand whether a campaign is below the product’s margin ceiling before calling it profitable. Its budget automation content also makes a useful point: distributing budget by cost can accidentally reward expensive campaigns, while distribution by ACOS can strengthen campaigns that already look efficient.
BidX’s PPC controlling guide goes further into max CPC math. The formula is practical: price multiplied by target ACOS multiplied by conversion rate gives a rough maximum CPC. If a €30 product has a 20% target ACOS and 10% conversion rate, the max CPC is €0.60. That is useful operator math.
Podean’s 2026 Marketplace Index makes a different point: brands do not win on Amazon purely because they spend more. They win because fundamentals such as Prime eligibility, content quality, store execution and AI-readable product context are done consistently. That matters because advertising cannot compensate forever for weak retail readiness.
Seller communities say the same thing in rougher language. Reddit threads and seller roundups repeat the warning that Amazon Ads can happily spend whatever budget you hand it, that Sponsored Products often works best when auto campaigns are harvested into manual exact terms, and that ACOS alone can lead operators into cutting useful demand or scaling useless revenue.
What most advice still misses is the operating layer between insight and action. It explains the metrics. It explains the platform levers. It rarely asks: which signal is allowed to overrule which other signal today? That is where profit leaks.
The profit signal ledger: five fields every signal needs
A profit signal ledger is not complicated. It is a shared decision table for the ad operator, commercial lead and finance owner. Every signal that might move bids, budgets or campaign roles needs five fields.
- Signal type — ACOS, TACoS, contribution margin, stock cover, conversion rate, CPC, search-term evidence, competitor pressure, price position, review movement, Buy Box strength or fulfilment promise.
- Freshness — when the signal was last updated and how long it remains valid. A search-term conversion signal from yesterday is not the same as a margin number updated three weeks ago.
- Decision permission — what the signal can actually trigger: raise bid, cap budget, pause, launch a test, reallocate spend, escalate to finance or do nothing.
- Override rules — which signals can block the action. Stock cover should block a bid increase. Negative contribution margin should block a ROAS celebration. Weak offer readiness should block competitor conquesting.
- Owner — who approves the decision if signals conflict: ad operator, marketplace lead, finance, supply chain or founder.
The point is not bureaucracy. The point is speed with memory. When an Amazon campaign improves from 31% to 18% ACOS, the operator should not start a debate from scratch. The ledger should already say: “ACOS improvement can release 20% more budget only if contribution margin is above €6.50, stock cover is above 28 days and TACoS is not rising faster than total revenue.”
FiveX is useful here because ad data, product profitability, stock cover, marketplace fees and campaign logs can sit in the same operating view. The ledger should not live as a heroic spreadsheet that one person updates after coffee. It should be connected to the same data the operator uses before changing spend.
Example 1: NorthSea Home should not scale the cleanest ACOS
NorthSea Home sells a kitchen organiser on Amazon.nl, bol.com and MediaMarkt. Amazon Sponsored Products looks excellent on Monday morning: €820 spend, €4,320 attributed revenue and 19.0% ACOS over the last seven days. The old rule would say “scale”. The profit signal ledger says “check permission”.
The SKU’s selling price is €39.95. After marketplace fees, fulfilment, packaging, expected returns and purchase cost, contribution margin is €7.80. At a 7.5% paid conversion rate, the break-even CPC is roughly €0.59. Amazon’s average CPC has climbed to €0.54, so there is only €0.05 of headroom. Stock cover is 19 days because a replenishment shipment is delayed. bol.com has 41 days of LVB stock and a stronger contribution margin of €8.90 because returns are lower.
The decision is not “Amazon ACOS is good, raise bids”. The decision is: cap Amazon harvesting budget at €140 per day, move €60 per day to bol Sponsored Products for the same SKU, and set an Amazon bid ceiling of €0.58 until stock cover returns above 28 days. The signal that won was not ACOS. It was profit capacity.
This is exactly the kind of cross-marketplace trade-off a managed Advertentie Service should make. FiveX’s contribution-margin reporting and inventory insights make the argument visible: the next euro belongs where the SKU can still fulfil profitably, not where the platform dashboard is greenest.
Example 2: StudioBright should renew learning, not call it failure
StudioBright launches a rechargeable desk lamp on bol.com with a €5,500 monthly ad budget across Amazon, bol and MediaMarkt. The bol launch campaign spends €640 in the first ten days, gets 1,280 clicks, 28 orders and 36% ACOS. A platform-only review would call the campaign inefficient next to Amazon’s 22% ACOS.
The ledger adds context. The product has 74 days of LVB stock, a contribution margin of €11.20, and the top non-brand query has moved from position 38 to position 17 in organic rank. Conversion on the exact term is still only 2.2%, but review count moved from 6 to 14 after the first order wave, and the product page is now eligible for a category promotion next week.
The mistake would be to pause the campaign because the launch ACOS looks ugly. The better decision is to renew the learning role for another €450, but narrow it: exact terms with at least 80 clicks stay live, broad discovery is capped at €25 per day, and any term above €95 spend with zero orders is quarantined. FiveX Ads AI can support this by proposing bid cuts and increases, but the ledger decides which recommendations are commercially allowed before they go live.
That distinction matters. Automation is strongest when it knows the role of the spend. A launch campaign with improving rank and healthy stock deserves different treatment from a harvest campaign with mature evidence and shrinking margin.
Example 3: VoltPeak should ignore the competitor panic signal
VoltPeak sells a €249 air purifier bundle on MediaMarkt and Amazon. A competitor drops price by 12% for the weekend, and the MediaMarkt retail media dashboard shows impression share falling from 42% to 27%. The sales team wants to fight back with a €900 weekend budget increase.
The ledger slows the room down. VoltPeak’s bundle contribution margin was €38.40 before the competitor move, but a planned coupon already reduces it to €30.10. Average CPC on the MediaMarkt placement has risen from €0.72 to €0.96. At the current 3.1% conversion rate, each order needs about 32 clicks, or €30.72 of media cost. The ad order is basically at break-even before returns.
Amazon, meanwhile, has a branded defence campaign spending €85 per day at 11% ACOS and 52 days of FBA stock. bol has weaker conversion but no direct price attack. The ledger decision: do not fight the MediaMarkt auction this weekend. Hold the placement, protect Amazon branded defence, and review competitor pressure after the promotion ends. If MediaMarkt share remains below 30% for five more days and margin returns above €34, reopen the conquest test with a €300 cap.
That is operator voice in practice: not every competitor move deserves a reaction. Some signals are panic, not permission.
How to run the ledger weekly
The weekly routine should be short enough to survive busy weeks. I like a 45-minute board with four passes.
1. Freeze stale signals
Start by marking any margin, stock, pricing or promotion signal that is older than its decision window. For €5K+ accounts, margin should usually be refreshed weekly, stock daily for hero SKUs, and promotion context whenever prices or coupons change. A stale signal cannot approve more spend. It can only trigger a review.
2. Separate signal from action
Write the signal in one column and the proposed action in another. “ACOS fell to 18%” is not an action. “Increase exact-match bid by 12%” is an action. This separation prevents dashboards from pretending they made the decision.
3. Apply blockers before optimisers
Before anyone discusses bid increases, check blockers: negative contribution margin, low stock cover, Buy Box loss, weak delivery promise, content issue, promotion overlap, return spike or cash constraint. If a blocker is active, optimisation becomes secondary. You either cap, pause, isolate or escalate.
4. Log the decision and expiry
Every material spend move needs a reason and an expiry date. “Raised bid from €0.42 to €0.48 because exact term hit 92 clicks, 9 orders, €8.70 contribution margin and 44 days stock. Review after €250 extra spend or seven days.” That sentence is not admin. It is future-proofing.
FiveX campaign logs, product strategies and advertising reports help turn that routine into an audit trail. When the next QBR asks why a budget moved, the team should not reconstruct the decision from memory. The proof should already exist.
The signals that should almost never act alone
Some signals are useful but dangerous on their own.
- ACOS should not raise budget without margin, stock and TACoS context.
- ROAS should not approve scale if the SKU has high return exposure or weak contribution margin.
- Search-term conversions should not trigger permanent bids until enough click volume exists.
- Competitor price changes should not force a response unless the SKU has enough margin to fight.
- Stock cover should not kill every campaign automatically; some brand defence may remain necessary during replenishment.
- Automation recommendations should not execute without checking role, evidence and commercial blockers.
The operator’s job is to turn conflicting signals into explicit trade-offs. If you scale Amazon, what does bol lose? If you defend MediaMarkt, which SKU stops learning? If you protect stock, which ranking momentum are you willing to sacrifice? Good ad management is not pressing the right button. It is choosing the right constraint.
Final thought: dashboards show signals, operators grant permission
The best marketplace ad accounts I see are not the ones with the most metrics. They are the ones where metrics have decision rights.
ACOS can suggest. Margin can veto. Stock can delay. Search evidence can renew learning. Competitor pressure can escalate. Finance can cap. The ledger makes those rights visible before money moves.
If your Amazon, bol and MediaMarkt spend is above €5K per month, this is where a specialist Advertentie Service should earn its fee. Not by reacting to every dashboard movement, and not by hiding behind generic ROAS targets. By building the operating layer that decides which signals deserve budget, which signals need more evidence, and which signals should be politely ignored.
That is how marketplace advertising becomes profit work instead of metric chasing. Slightly less glamorous than a green dashboard, yes. Much better for the bank account.