Marketplace ad retainers usually sound sensible at the moment they are signed.
The brand wants someone reliable on the account. The agency promises weekly optimisation, search-term reviews, bid management, reporting and strategy calls. Everyone likes the idea of continuity. Nobody wants the Amazon, bol or MediaMarkt account to depend on a founder remembering to check CPCs between finance meetings.
That is not the problem.
The problem is that “maintenance” can quietly become a comfortable word for activity without enough commercial permission. Bids get tuned. Budgets get nudged. Search terms get harvested. A report gets sent. The retainer feels busy, but the actual question is not answered often enough: which spend still deserves to exist this week?
The named mistake is treating marketplace ad maintenance as recurring optimisation instead of recurring profit control. A web maintenance plan can keep plugins updated, backups running and support available. Marketplace ads need something sharper. They need margin, stock, offer strength, campaign role, search evidence and decision ownership to be re-approved again and again, because the auction changes faster than a website maintenance checklist.
My stance: managed Amazon, bol and MediaMarkt accounts above roughly €5K monthly spend need an ad retainer maintenance room. Not a generic list of monthly tasks. A weekly operating layer that defines which campaigns are maintained, which are challenged, which are capped, which are rebuilt, and which are no longer allowed to consume retainer time or ad budget.
What the public advice gets right
The research was useful because it showed two very different worlds talking about the same word: maintenance.
Flywheel’s WordPress maintenance content explains why recurring maintenance plans are attractive for agencies and clients. The pitch is clear: predictable revenue for the agency, ongoing protection for the client, simpler billing and a natural way to keep a site healthy after launch. Their Maintainn partner page frames maintenance as safe hosting, website maintenance and support from a trusted team, with typical budgets that can range from under $5,000 to $50,000 depending on scope.
That model works for websites because many tasks are genuinely preventive. Updates, backups, security checks, uptime and support reduce risk even when traffic is calm.
Podean’s Amazon agency article adds the marketplace reality. It argues that Amazon specialists are useful because advertising cannot be separated from retail operations, forecasting, replenishment and product-page work. The most important line is that the person who owns the budget needs more than advertising to be successful on Amazon. That is exactly right.
BidX’s pricing page shows how the market packages support: platform access, managed platform, managed service, strategy calls, handled platform operations, assistance hours, consulting and automation. SalesDuo’s agency comparison pushes useful hiring criteria: campaign structure, search terms, TACoS, inventory, pricing, reporting transparency, raw data access and change logs. Daniks goes one level more numerical and warns that a $2,000 monthly retainer on $6,000 ad spend can be mathematically impossible to recover, even when the agency improves ACOS.
YouTube search results around Amazon PPC agencies repeat the same buyer fear in a less polished way: do not hire the wrong agency, ask better questions, understand the cleanup timeline, and compare agency, in-house, software and automatic campaigns before committing.
What everyone misses: maintenance has to expire permissions
Most agency selection content asks whether the agency is skilled enough. Most software content asks whether the platform can automate enough. Most retainer content asks what is included in the monthly fee.
Those questions matter, but they miss the maintenance failure I see in real marketplace ad accounts: permissions do not expire.
A campaign was created to defend a hero SKU during a competitor push. Three months later the competitor is gone, but the campaign still spends because branded ACOS looks nice. A bol Sponsored Products campaign was capped during a stock squeeze, then nobody reopened the cap when supply recovered. A MediaMarkt retail media test received a “pilot” budget in May and is still running in November, even though the promoted product now has lower margin after a supplier price increase.
The ad account is not messy because nobody maintained it. It is messy because maintenance kept old decisions alive.
That is the trade-off. A retainer should create continuity, but continuity without expiry becomes drift. The operator job is not only to optimise the account. It is to keep asking whether last month’s logic still has permission to spend this month’s money.
The ad retainer maintenance room
An ad retainer maintenance room is a weekly decision board for the managed account. It is boring on purpose. It forces every recurring optimisation task to connect to one of four commercial jobs:
- Protect: keep profitable demand stable when the SKU has margin, stock and offer permission.
- Repair: fix campaigns where structure, search terms, targets or budget routing are blocking profit.
- Challenge: make old winners prove they still deserve spend after margin, stock, price, reviews or competitor changes.
- Retire: stop maintaining campaigns, keywords, targets or reports that no longer move a decision.
This room is not a replacement for bid automation. FiveX Ads AI can still surface bid increases and decreases. Automation rules can still pause, activate or reprice rows. Product groups can still segment funded SKUs, margin bands or country splits. The room decides when those actions are allowed, when they need a human, and when the right answer is to stop touching the thing entirely.
Example 1: NovaNest and the €6,000 maintenance retainer that protected the wrong work
NovaNest sells home storage products on Amazon and bol. The brand spends €18,000 per month on marketplace ads and pays a €6,000 monthly management retainer. On paper, the service looks active: weekly bid checks, search-term harvesting, monthly reporting and a strategy call.
The problem sits in the workload. In one month, the agency spends 11 operator hours maintaining 43 low-spend campaigns that together use €2,100 of ad budget. Those campaigns produce €5,800 in attributed revenue at 36% ACOS. The same account has three hero SKUs spending €9,400 with unstable stock cover, but they receive only four hours of review because the reporting template treats every campaign family equally.
The maintenance room changes the rule. Any campaign below €150 monthly spend must show either a learning job, a defend job or a clear SKU margin reason. If it cannot, it moves to quarterly review instead of weekly maintenance. NovaNest frees seven hours per month and moves those hours into the three hero SKUs. One SKU gets capped because stock cover is 16 days. One gets scale permission because loaded break-even ACOS is 29% and current TACoS is stable. One moves to repair because bol NL is profitable but bol BE is below margin after fulfilment.
The point is not that small campaigns are bad. The point is that retainer hours need the same permission logic as ad budget. FiveX helps here by making product profitability, stock cover and campaign performance visible in the same operating layer, instead of asking the operator to reconcile three exports before the weekly call.
Example 2: BrewMate and the €5K account where software beats an agency until the brief changes
BrewMate sells coffee accessories on Amazon and bol and spends €5,200 per month on ads. A classic agency retainer offer of €1,750 per month looks reasonable until finance runs the ratio: the fee equals 34% of ad spend before a single click is bought.
The account has only 24 campaigns. The real problem is not strategic complexity. It is that bids are not tied to break-even ACOS and stock cover. In this case, the maintenance room says: do not buy a full agency retainer yet. Use software and a monthly operator review.
FiveX groups the products into three product groups: defend, launch and wait. Ads AI suggests bid changes inside max-CPC guardrails. Automation rules pause keywords that spend more than €45 without an order when the SKU has less than 22% contribution margin. A monthly review checks search terms, margin changes and stock risks.
The monthly management cost stays closer to €650 in internal/operator time rather than €1,750 in agency fee. That gives BrewMate roughly €1,100 more room for learning budget. The moment the account expands into MediaMarkt retail media, adds 80 SKUs or starts running country-specific bol NL/BE budgets, the decision may flip. Then the agency is not hired to “do PPC”. It is hired to run a larger maintenance room with explicit decision rights.
This is where many retainer conversations go wrong. The question is not “agency or software?” The question is “how much commercial ambiguity exists, and who is responsible for resolving it every week?”
Example 3: LumiClean and the MediaMarkt campaign nobody wanted to own
LumiClean sells small electronics across Amazon, bol and MediaMarkt. The brand spends €31,000 per month on marketplace ads. A MediaMarkt retail media campaign was launched as a three-month pilot with €4,500 monthly budget. Month one looked acceptable: €13,600 reported revenue and 33% platform ACOS. Month two looked similar. By month three, everyone had stopped questioning it.
Then finance updated the product cost. The landed cost rose by €6.40 per unit. The SKU also needed a €15 discount to hold price position against Amazon. The old break-even ACOS of 27% became 15%. Suddenly the “acceptable” campaign was spending outside profit permission.
The retainer maintenance room catches this because every campaign role has an expiry date and a margin version. The MediaMarkt campaign cannot remain in pilot status after 90 days without re-approval. It moves from maintain to challenge. The operator has three allowed actions: cap spend at €1,200 while margin is rechecked, redirect budget to Amazon branded defence where contribution margin is still 24%, or rebuild the MediaMarkt campaign around a bundle with higher average order value.
The team chooses a two-week cap and bundle rebuild. FiveX’s change logs and reporting exports preserve the decision trail: why budget was cut, which margin version triggered the change, who approved the bundle test and when the pilot must be reviewed again.
What should be inside a proper marketplace ad maintenance retainer?
If you are paying for Advertentie Service, the scope should not only say “campaign optimisation”. It should define the decisions the service maintains.
1. A weekly permission review
Every important SKU and campaign should receive a fresh label: scale, maintain, cap, repair, hold or retire. That label should be based on margin, stock, offer readiness, search evidence and marketplace role.
2. A campaign-role register
Each campaign needs a job. Defend branded demand. Harvest proven non-brand demand. Learn new search terms. Recover after stockout. Clear old stock. Support a promotion. If the job is unclear, the campaign should not receive routine maintenance as if it were important.
3. A maintenance-hour budget
Retainer time is finite. Decide how many hours go to protection, repair, learning, reporting and strategic review. If 60% of time goes to report production and low-impact bid edits, the retainer is probably serving the workflow more than the profit.
4. Expiry dates for old decisions
Bid ceilings, budget caps, negative keywords, campaign roles and paused targets all need review dates. A decision without an expiry date becomes account furniture. Everyone walks around it until it costs money.
5. A proof pack for every material move
When spend moves, the proof should be visible: SKU margin, current stock cover, search-term evidence, platform performance, expected role and rollback condition. This is where FiveX’s profitability dashboards, product groups, Ads AI recommendations, automation logs and exports become more than software features. They become the proof layer behind the retainer.
The simple retainer test
Before signing or renewing a marketplace advertising retainer, ask one question:
What will you stop maintaining next month if the evidence says it no longer deserves attention?
A strong operator can answer. They will talk about campaign roles, SKU margin, stock cover, search evidence, decision logs and review dates. They will explain which work is weekly, which work is monthly, which work is automated, and which work should disappear entirely.
A weak retainer hides behind volume: more bid changes, more reports, more calls, more exports, more “optimisation”. That can feel reassuring. It is also how €5K, €10K or €30K marketplace ad accounts become expensive to manage without becoming easier to trust.
The best retainer is not the one that keeps touching everything. It is the one that keeps the right decisions alive and lets the expired ones die.