Hiring a marketplace ad service feels like the moment things should finally speed up. The agency or operator gets access, audits the account, cleans the campaign structure, finds wasted spend and starts improving ACOS. Nice and tidy.
Except the first two weeks often do the opposite. Everyone is waiting for exports. The Amazon account has three years of campaigns but no explanation of why half the exact terms are paused. The bol Sponsored Products campaigns are live, but nobody has the SKU margin file. MediaMarkt retail media has a seasonal placement booked, but stock for the hero product is already tight. Finance has landed cost in one sheet. Operations has inbound dates in another. The old agency knows why a keyword was blocked, but that knowledge lives in somebody's head.
The named mistake is access-first onboarding: giving a new operator console access before the business has handed over the commercial truth behind the account. It feels efficient because the ad team can start clicking. It is risky because the first optimisation sprint is then based on platform data, not profit permission.
My stance: before anyone touches bids, a marketplace ad service should receive a profit data room. Not a fancy enterprise portal. A practical, version-controlled onboarding pack that connects campaigns to SKU economics, stock, offer health, channel roles, decision rights and historical reasons. If the data room is missing, the operator may still improve the dashboard. They just cannot safely improve the business.
This guide is written for ecommerce brands in the Netherlands and Belgium spending from roughly €5K per month across Amazon, bol.com and MediaMarkt. At that level, onboarding is no longer an admin step. It is the control system that decides whether the first 30 days build profit or just create a prettier account.
What the market already explains well
The existing advice around Amazon PPC onboarding and marketplace advertising is useful, but it usually stops too early. Ecom Managers describes agency onboarding as more than account linking: a serious PPC partner needs to understand the business from the inside out. That is the right direction. General agency onboarding guides like Leadsie and Portaal are strong on access, goals, contacts, brand assets, timelines and client communication. That matters because messy onboarding creates slow delivery and avoidable trust issues.
BidX is strong on campaign mechanics. Its PPC structure guidance explains why separating auto, broad, exact and ASIN campaigns helps with budget control and evaluation speed. Its broader Amazon PPC material rightly pushes advertisers beyond bid tweaks into structure, retail readiness, creative, pacing and automation. Bol-focused guides from BAAT, Amazin and HYPD explain Sponsored Products, CPC bidding, ACOS, koopblok eligibility, product-page quality, stock and the importance of moving from broad visibility to controlled optimisation. SalesDuo and similar Amazon ad-type guides are useful for mapping Sponsored Products, Sponsored Brands, Sponsored Display and DSP by placement, pricing and use case.
All fair. All helpful. But most of this content focuses on how campaigns should be run once the operator is inside the account. The missing piece is the handover layer before that: what commercial evidence must exist before the first optimisation is allowed?
That is the angle FiveX can own: marketplace ad onboarding as a profit-permission data room, not a login checklist.
The onboarding data room: what it must contain
A good data room gives the new ad operator enough context to make the first five decisions safely: what to protect, what to reduce, what to test, what to pause and what to escalate. It should be small enough to maintain weekly and complete enough that nobody has to reverse-engineer last year's logic from campaign names.
Build it around seven folders.
1. SKU economics
This is the folder most teams under-build. Campaign reports show sales, cost and attributed revenue. They do not show whether a SKU can afford the next click.
For every advertised SKU, collect selling price, VAT treatment where relevant, marketplace commission, fulfilment cost, landed cost, packaging, expected return cost, coupons, warranty reserve, contribution margin in euros and contribution margin percentage. Then calculate the loaded break-even ACOS. If a €49.95 kitchen scale has €12.40 contribution margin after fees and expected returns, its loaded break-even ACOS is 24.8%. That does not mean the target ACOS should be 24.8%. It means 24.8% is the cliff edge before operating profit disappears.
FiveX fits naturally here because our profitability dashboards turn marketplace, advertising and financial data into SKU-level margin views. The operator should not have to ask finance for a new spreadsheet every time a campaign needs a bid decision.
2. Campaign history and search-term memory
Do not just export current campaigns. Export the reasons behind them. Which exact terms are proven winners? Which terms were blocked because they attracted bargain hunters, wrong variants or high-return buyers? Which competitor ASINs looked attractive but never converted after 80 clicks? Which campaigns exist only because a previous operator wanted cleaner reporting?
The data room should include at least 90 days of search term data, preferably 180 for accounts with seasonal demand. Add labels: winner, watch, blocked, seasonal, brand defence, competitor test. The label is more valuable than the export because it explains intent.
This is where FiveX advertising automation helps: rules and AI recommendations can act on search terms faster, but only if the human logic behind winners, blockers and tests is documented. Automation without memory is just a quick way to repeat old mistakes.
3. Stock and fulfilment permission
Marketplace ads do not fail only because bids are wrong. They fail because the shelf cannot receive demand. For Amazon, collect FBA stock, inbound shipment status, FBM fallback, Buy Box or Featured Offer stability and days of cover. For bol, collect LVB stock, own-warehouse delivery promise, koopblok status and Belgian availability where relevant. For MediaMarkt, collect stock allocation, delivery promise, retailer approval status and any retail media booking tied to a date.
Use simple permission labels: scale when stock cover is healthy, hold when stock is acceptable but not abundant, protect when ads should defend only high-intent demand, and freeze when spend would accelerate a stockout or route traffic to a weak offer.
FiveX inventory insights make this less painful because stock cover can sit next to ad spend and margin. That matters when a campaign has a beautiful ACOS but only 11 days of stock left.
4. Channel role and budget logic
Amazon, bol and MediaMarkt should not automatically receive budget because they exist. Each channel needs a job. Amazon may be the search-demand engine. Bol may be the NL/BE volume stabilizer. MediaMarkt may be the electronics credibility channel for selected SKUs. Or the roles may be completely different by category.
Document the intended role for each SKU by marketplace: protect, grow, test, harvest, clearance or do not fund. Then document the budget rule. A simple example: protect campaigns may spend until 70% of loaded break-even ACOS, grow campaigns may spend until 85% if stock cover exceeds 30 days, and test campaigns must cap at €250 until they produce at least 20 attributed orders or a clear search-term signal.
That sounds slightly strict. Good. Onboarding should make trade-offs visible before the first weekly call, not after €1,000 disappears into a vague “learning phase”.
5. Listing, offer and retail-readiness issues
A new operator should not have to discover on day nine that the main Amazon image was changed by a reseller, the bol listing lacks a key specification, or the MediaMarkt product page has a weak warranty message. Collect known listing issues before optimisation starts.
For each priority SKU, add image status, title quality, review count, rating, product-page conversion issues, compliance blockers, price position and offer eligibility. If the listing is not ready, the campaign should receive a smaller role. Do not use ads to compensate for a page that cannot convert profitably.
6. Decision rights and escalation rules
This is the folder that prevents Slack chaos. Decide who can approve budget increases, who can pause a campaign, who owns price changes, who approves coupon support, who confirms stock exceptions and who signs off channel allocation shifts.
For €5K+ monthly spend, the operator should not wait three days to reduce waste, but should not be allowed to double a bol budget during a promotion without margin and stock permission either. Use thresholds. For example: the operator may reduce bids or pause low-evidence spend immediately; may move up to 15% of weekly budget between approved campaigns; needs approval for new marketplace formats, total budget increases or SKU-level spend above the agreed ceiling.
7. Reporting baseline
Before optimisation begins, freeze the baseline. Not because baseline reports are glamorous. They are not. They are the only way to know whether the new service improved the account or merely changed the dashboard.
Record 30- and 90-day performance by marketplace, SKU and campaign role: ad spend, attributed revenue, ACOS, TACoS where available, orders, CPC, conversion rate, contribution profit after ads, stock cover, refund/return rate and organic sales trend. FiveX marketplace analytics helps here by connecting ad data with operational and financial context instead of leaving the baseline split across Seller Central, bol Partnerplatform, MediaMarkt exports and spreadsheets.
Three examples: what good onboarding changes
Example 1: BrewScale stops a beautiful ACOS from creating a stockout
BrewScale sells a €49.95 smart kitchen scale on Amazon.nl and bol.com. The Amazon campaign shows 18% ACOS, 3.8% conversion rate and 210 attributed orders in 30 days. A new operator without the data room would likely raise bids on the best exact terms.
The data room changes the decision. Contribution margin after fees, fulfilment and expected returns is €12.40, so loaded break-even ACOS is 24.8%. That looks fine. But FBA stock is 410 units, inbound stock arrives in 21 days, and the current run-rate is 14 units per day before any bid increase. Stock cover is 29 days on paper, but only 12 days above the minimum Amazon reserve the brand wants to keep. The SKU gets a protect label, not scale. FiveX inventory insights flag the stock cover next to the campaign, and the operator shifts €600 of planned growth budget to a bol bundle with 54 days of cover instead.
Example 2: SkinNova avoids funding the wrong bol winner
SkinNova sells a €27.50 serum on bol. Sponsored Products report a 16% ACOS on “vitamine c serum”, which looks like a winner. The previous agency planned to increase daily budget from €80 to €150.
The onboarding data room shows why that would be a mistake. The serum's contribution margin is only €5.10 after bol commission, fulfilment, packaging and a 7% return allowance. Loaded break-even ACOS is 18.5%, and the target ceiling is 13% because finance needs margin for influencer seeding. Worse, search-term notes show that the term attracts buyers comparing larger 50ml bottles, while SkinNova sells 30ml. Returns are 11% on that query cohort versus 5% account average. The operator keeps brand-defence terms live, moves “vitamine c serum” into watch mode, and tests a more precise “30ml vitamine c serum gevoelige huid” cluster with a €250 evidence cap.
Example 3: VoltNest saves a MediaMarkt launch from a channel-role mismatch
VoltNest is launching a €149 USB-C docking station through MediaMarkt and Amazon. MediaMarkt offers strong electronics context, but the ad booking is tied to a two-week retail media push. Amazon Sponsored Products already converts at 22% ACOS on high-intent terms.
The data room shows the trade-off. The docking station has €37 contribution margin on Amazon after FBA and €31 on MediaMarkt after retailer fees and a warranty reserve. MediaMarkt stock allocation is 180 units, while Amazon has 620 units and faster replenishment. The operator gives MediaMarkt a test credibility role with a €900 cap and strict price-position monitoring, while Amazon receives the capture budget for proven bottom-funnel terms. The result is not “MediaMarkt is bad”. The result is “MediaMarkt has a job, and that job is not unlimited conversion spend this month”.
The first 10 working days after handover
A good marketplace ad service should not spend the first 10 working days randomly “optimising”. Use a sequence.
- Days 1-2: access, tracking checks, billing checks, export validation and baseline freeze.
- Days 3-4: SKU economics review, loaded break-even ACOS calculation and margin-risk labels.
- Days 5-6: stock, offer, Buy Box, koopblok and listing-readiness checks.
- Days 7-8: campaign role mapping: protect, grow, test, harvest, clearance, do not fund.
- Days 9-10: first bid and budget actions, limited to the decisions that have clear profit permission.
This order matters. If you move bids before SKU economics and stock permission are clear, you are not onboarding. You are gambling with a nicer meeting agenda.
What to ask a marketplace ad service before you start
Use these questions before signing or during kickoff:
- Which SKU-level margin fields do you need before the first optimisation sprint?
- How do you calculate loaded break-even ACOS for Amazon, bol and MediaMarkt?
- What campaign actions are you allowed to take without approval?
- How do stock cover, Buy Box, koopblok and listing issues change bid decisions?
- How do you preserve search-term memory from the previous operator?
- When do you refuse to scale a campaign even if platform ROAS looks good?
- What will be different in the account after the first 10 working days?
If the answer is mostly “we will audit the account and optimise from there”, keep pushing. Audit is not enough. You want to hear how they connect platform data to profit, stock and decision rights.
Where FiveX fits
FiveX is built for this operating layer. The platform connects marketplace analytics, profitability dashboards, advertising automation, AI recommendations, repricing signals and inventory insights so the advertising service is not steering from isolated ad-console metrics.
That is especially important for NL/BE brands where Amazon, bol and MediaMarkt each tell only part of the truth. Amazon may show efficient Sponsored Products. Bol may have stronger local conversion but tighter koopblok rules. MediaMarkt may create premium electronics demand but need stricter stock and warranty controls. The job of an ad service is not to make each console look good separately. The job is to decide where the next euro has permission to work.
Onboarding is where that discipline starts. Build the profit data room first. Then touch the bids.
Bottom line
The first mistake in marketplace ad service onboarding is not a bad campaign structure. It is starting with campaign structure before the operator understands margin, stock, search memory, channel role and decision rights.
If you are spending from €5K per month on Amazon, bol and MediaMarkt, treat onboarding as a commercial control process. Give the operator a data room that says which SKUs deserve spend, which campaigns carry old evidence, which offers are not ready, and which decisions need approval.
Access lets an agency move buttons. A profit data room tells them which buttons should not be moved yet. That difference is where margin survives the handover.