A marketplace ad management SLA sounds painfully corporate. Very procurement. Very “please define response times in business days”. Not exactly the kind of thing that makes an Amazon, bol or MediaMarkt operator reach for coffee with excitement.
But once a brand spends more than roughly €5K per month on marketplace ads, the SLA becomes one of the most commercial documents in the account. Not because it tells the agency when to reply. Because it decides who is allowed to move money, when, and under which profit conditions.
The named mistake I see is the vague retainer handoff. A brand hires an Advertentie Service provider, gives access to Amazon Ads, bol Sponsored Products and the MediaMarkt retail media portal, agrees on a target ACOS, and assumes the specialist will “optimize”. Two weeks later the operator has improved keyword bids, but the Amazon hero SKU is nearly out of stock, the bol Belgium campaign is funding a variant with half the Dutch margin, and MediaMarkt spend is waiting for a product page fix nobody owned. Everyone worked hard. Nobody had decision rights.
My stance: a marketplace ad management SLA should not be a customer-service promise. It should be a profit operating agreement. If it only says what the agency does, it is too weak. It must say what the agency may not do when margin, stock, Buy Box, price position or data quality says “not today”.
This guide is for brands in the Netherlands and Belgium that sell on Amazon, bol and MediaMarkt, usually from around €5K monthly ad spend. At this level, ad management stops being a list of campaign tasks. It becomes a weekly system for deciding which SKUs deserve capital.
What the current agency advice gets right
The research landscape is useful, especially from the stronger marketplace specialists. Directive frames marketplace advertising as a margin decision before it becomes a campaign plan. Adverio is sharp on profit leaks: incrementality loss, query mismatch, conversion readiness, keyword cannibalisation, rank decay and Buy Box instability. ROI Revolution explains the value of flexible retail media strategy across Amazon, Walmart, Home Depot, Target, Best Buy and other retailers, with automation and human monitoring working together. Podean positions the work as full-service marketplace growth across retail operations, media, content and analytics, and even names waste of ad spend as a scorecard problem. BidX shows the software side clearly: faster campaign creation, automated keyword, bid and budget optimisation, DSP and AMC analytics.
Reddit and YouTube seller discussions add the operator reality. Sellers are not only asking “what is a good ACOS?” They are asking whether an agency should be hired before break-even ACOS by SKU is known, whether branded and non-branded campaigns are separated, whether TACOS is moving in the right direction, and whether the listing can convert the extra traffic. That is the right anxiety. It is less glamorous than “full-funnel retail media”, but it saves money.
What most advice still misses is the service boundary. It tells you what a good agency can manage, but not exactly which decisions the agency is allowed to make without waiting for the client, and which decisions must be blocked because finance, operations or marketplace readiness has not given permission. That missing layer is where €5K accounts leak.
The SLA should define decision rights, not just deliverables
A classic agency SLA says things like weekly optimisation, monthly reporting, campaign creation, search term reviews and response time within 24 hours. Nice. Also incomplete.
A profit-first marketplace ad management SLA answers five harder questions:
- Budget rights: who may move spend between Amazon, bol and MediaMarkt?
- Bid rights: who may increase bids above the product’s break-even ACOS or break-even CPC?
- SKU rights: which products are eligible for scale, test, protect or pause?
- Operational vetoes: what happens when stock, Buy Box, price position, delivery promise or listing quality breaks?
- Evidence cadence: what must be checked weekly before budget moves?
That is the difference between “we manage your ads” and “we operate your ad spend as working capital”. The second version is what brands need once the account has enough spend to hurt.
The four layers every €5K SLA needs
1. The commercial baseline
Before the first bid changes, the SLA should force a baseline by SKU. At minimum: selling price, marketplace fees, fulfilment cost, purchase cost, expected return cost, gross margin, contribution margin before ads, break-even ACOS and stock cover. Without this, target ACOS becomes account folklore.
FiveX fits naturally here because product profitability is not a side spreadsheet. The platform connects marketplace revenue, costs, fees and ad spend so the operator can see which products can afford growth. If a SKU has 18% contribution margin before ads, a 25% ACOS target is not ambitious. It is mathematically rude.
2. The marketplace role map
Amazon, bol and MediaMarkt rarely play the same role. Amazon may be the high-volume search engine. bol may be the local trust and LVB conversion channel. MediaMarkt may be the electronics-intent shelf where price, availability and retail media placement decide whether the click is worth buying.
The SLA should label each marketplace by role: protect, grow, test, clear stock or defend share. A campaign with a protect role should not receive the same freedom as a launch campaign. A MediaMarkt Sponsored Brand Ad for visibility should not be judged the same way as a bol Sponsored Product campaign defending an already profitable bestseller.
3. The permission rules
This is the heart of the SLA. Permission rules translate messy operational signals into yes/no advertising decisions.
- No margin, no scale: bids may not increase when SKU margin is below the agreed threshold.
- No stock, no boost: discovery budget pauses when stock cover falls below, say, 21 days.
- No Buy Box, no defence: Amazon campaigns stop defending a listing if Buy Box ownership is unstable.
- No price parity, no conquest: competitor targeting waits when the advertised offer is visibly more expensive than comparable marketplace alternatives.
- No clean data, no automation: AI bid recommendations are reviewed manually when cost or attribution data is incomplete.
FiveX advertising automation and AI recommendations become much safer when these rules exist. The point is not to let automation move every bid. The point is to let automation suggest moves inside commercial boundaries the business has already approved.
4. The weekly decision cadence
A proper SLA should define the weekly rhythm. Not a 30-slide report. A decision meeting.
The best cadence I have seen is simple: Monday checks budget pacing and stock; Wednesday reviews search terms, wasted spend and bid exceptions; Friday decides whether the next week’s budget moves across marketplaces. The output is not “performance discussed”. The output is a list of actions with owners: increase, cap, pause, relaunch, fix listing, replenish stock, adjust price or wait.
FiveX marketplace analytics helps here because it can bring campaign performance, product economics, inventory and channel contribution into one operating view. That keeps the meeting from becoming a platform-tab safari. Nobody needs six browser tabs and one nervous spreadsheet before breakfast.
Named example 1: Amazon hero SKU with a fake green ACOS
Imagine a Dutch supplement brand spending €3,200 per month on Amazon Ads. Its hero SKU sells for €29.95. After referral fees, fulfilment, purchase cost and expected returns, contribution margin before ads is €8.10, or 27.0%. The agency target is 24% ACOS, so the campaign looks safe.
Then the SLA check adds two missing signals: stock cover is 13 days, and branded defence accounts for 48% of attributed ad sales. The reported 22% ACOS is not a scale signal. It is partly a defence signal, and it is pushing a SKU toward stockout.
A weak SLA says: “ACOS below target, increase bids 10%.” A profit operating agreement says: “Keep branded defence capped, move €450 from generic discovery into the second-best margin SKU, and hold broad-match expansion until stock cover is back above 28 days.”
The difference is not subtle. The first decision buys a prettier report and a stock problem. The second protects rank, cash and availability.
Named example 2: bol NL/BE split that hides margin difference
Now take a home organisation brand spending €1,400 per month on bol Sponsored Products. The same storage box sells in the Netherlands and Belgium, but the economics are not the same. In NL the product sells for €34.95 with €9.20 contribution margin before ads. In BE, lower conversion and higher fulfilment cost leave €5.80.
The combined campaign reports 19% ACOS. Fine at first glance. But the SLA requires marketplace and country splits. NL is at 16% ACOS with strong stock and 4.6% conversion. BE is at 31% ACOS with lower conversion and only 17 days of stock.
A normal optimisation might lower a few bids and move on. A useful SLA gives the operator permission to split the campaign, protect exact-match Dutch winners, cap Belgian discovery, and ask the client to improve the Belgian product page before more budget goes there. In FiveX, that decision can be tied back to product profitability and channel contribution instead of only campaign ACOS.
This is exactly why Dutch and Belgian marketplace accounts need more than “bol Ads management”. They need local economics by SKU and country.
Named example 3: MediaMarkt launch spend waiting on retail readiness
For the third example, picture an electronics accessory brand launching on MediaMarkt with €900 test budget. The product is a €79.99 USB-C docking station. The margin looks attractive: €18.40 contribution before ads. The team wants Sponsored Product Ads and Sponsored Brand Ads running in week one.
The SLA blocks full launch for three reasons: the product has only two reviews, the price is €6 higher than the strongest comparable offer, and inventory is split across channels with only 120 units reserved for MediaMarkt. The retail media plan is not wrong. It is early.
The profit operating agreement allows a narrow test: €25 per day on exact category-intent queries, no competitor conquesting yet, no Sponsored Brand budget until the product page reaches the agreed review and price-position threshold, and a stock alert at 45 units. If the product converts above 3.2% and ad ACOS stays below 28% after 14 days, the test can expand.
That is what good service looks like. Not slow. Disciplined.
What to put in your SLA before signing an Advertentie Service retainer
If you are about to outsource marketplace ad management, ask for the SLA to include these sections before access is handed over:
- Account scope: channels, countries, ad formats, campaign types and marketplaces covered.
- Commercial inputs: who maintains purchase cost, fees, shipping, return assumptions and margin thresholds.
- Budget authority: monthly caps, channel caps, SKU caps and rules for moving budget without approval.
- Bid authority: maximum bid increases, break-even ACOS rules and escalation thresholds.
- Operational blockers: stock cover, Buy Box, listing quality, price parity, review thresholds and delivery promise.
- Search term governance: how winners, blockers and watchlist queries are stored and reviewed.
- Automation governance: which actions may be automated, which need approval, and which are always blocked.
- Reporting output: weekly decision table, not just monthly performance slides.
- Client responsibilities: who fixes pricing, content, stock, product data and promotions when ads expose a non-media problem.
The last bullet is important. Many ad accounts underperform because the advertising team finds issues the client-side team does not fix. If nobody owns the product page, price, stock or product data correction, the agency can only keep paying for the same problem to appear in the data.
The trade-off: a stricter SLA may spend slower
A profit-first SLA will sometimes make the account spend slower than a pure media plan. That can feel uncomfortable, especially when the monthly budget is approved and everyone wants momentum.
But slow is not the enemy. Blind is the enemy.
If €5K ad spend is available, the operator’s job is not to spend €5K elegantly. The job is to decide whether Amazon should receive €2,800 this month, bol €1,600 and MediaMarkt €600 — or whether that mix should change because stock, margin and marketplace demand changed. Sometimes the best action is to hold €400 back from a campaign that cannot currently turn it into profit.
That is a mature advertising service. It treats budget as permission, not obligation.
How FiveX supports this operating model
FiveX helps an Advertentie Service team run this SLA without turning every week into manual spreadsheet archaeology.
- Marketplace analytics gives the operator a cross-channel view of Amazon, bol, MediaMarkt and other marketplaces instead of isolated platform screenshots.
- Product profitability shows whether each SKU can afford the proposed ACOS, CPC or budget increase after costs, fees and returns.
- Inventory insights prevent campaigns from scaling into weak stock positions or stealing stock from a stronger channel.
- Advertising automation and AI recommendations help turn approved rules into faster bid and budget actions, while keeping commercial vetoes in place.
That combination matters because the real value of an Advertentie Service provider is not “we changed bids for you”. It is “we made better capital-allocation decisions than your team could make from three ad dashboards and an export.”
The bottom line
If you are spending less than €1K per month, a simple campaign checklist may be enough. At €5K and above, it is not. The account has too many ways to look efficient while quietly leaking profit: branded sales disguised as growth, low-margin variants winning budget, stockouts after good weeks, MediaMarkt tests launching before retail readiness, and bol country splits hiding weak economics.
So before you hire or renew an Advertentie Service partner, do not only ask what they optimise. Ask what they are allowed to refuse.
The best marketplace ad management SLA is not the one with the fastest response time. It is the one that protects the business from bad spend before anyone has to explain the damage in a monthly report.