Marketplace ad accounts do not only need campaigns. They need campaigns with a job.
A brand campaign might defend demand you already earned. A generic search campaign might harvest existing category intent. A competitor campaign might test whether a rival’s shoppers are worth chasing. A bol Sponsored Products push might launch a SKU that needs early order velocity. A MediaMarkt retail media placement might clear stock before a new model arrives. Same ad platform, very different commercial jobs.
The named mistake I see is letting yesterday’s campaign role become today’s spending permission. The operator labels an Amazon campaign as DEFEND, a bol campaign as LAUNCH and a MediaMarkt placement as CLEARANCE. Those labels make sense when the plan is approved. Then price moves, stock cover changes, a competitor promotion ends, margin tightens, a review issue appears or the learning window simply finishes. The label stays alive anyway, so the budget keeps behaving as if the original reason still exists.
My stance: every Advertentie Service managing more than roughly €5K per month across Amazon, bol and MediaMarkt needs a channel-role expiry board. Not another naming convention. A weekly decision layer that says what job each campaign is allowed to do, what evidence keeps that job valid, and when the role expires unless someone deliberately renews it.
This is where marketplace ad management becomes operator work instead of platform babysitting. Bids matter. Search terms matter. Budget pacing matters. But none of those decisions are safe until the team knows whether the campaign is defending, harvesting, learning, launching, clearing stock, recovering after a stockout or deliberately losing money to buy evidence.
What existing advice explains well
The public advice on campaign structure and marketplace advertising is useful. BidX explains why Amazon PPC campaigns need clean structures, match-type separation and better budget control. Their budget automation content also makes a fair point: if budgets sit in the wrong campaigns, strong demand can run out of spend while weak campaigns keep money available.
PPC Alchemy’s beginner campaign-structure guidance makes the operational case for separating match types and keeping accounts manageable. Olifant Digital’s Amazon PPC guide goes deeper into campaign management, ACoS, TACoS, launches, audits and continuous optimisation. Podean’s Amazon content rightly pushes brands to think beyond media clicks and include retail metrics such as traffic, conversion and inventory.
Reddit and seller-community discussions are messier, but they reveal the real pain: sellers worry that lowering ACoS kills growth, that launch campaigns never stop spending, that inventory runs out before ads slow down, and that agencies report “optimisation” without explaining why a campaign still deserves money.
What most of this advice misses is role expiry. It tells you how to structure campaigns, move budgets and read performance. It rarely forces the uncomfortable question: does this campaign still have the same job it had when we opened it?
The missing layer: roles are not permanent
A campaign role is a temporary permission slip. It explains why a campaign is allowed to spend even when a narrow metric looks imperfect.
A LAUNCH campaign may accept 45% ACOS for a few weeks if it is buying ranking data and early conversion evidence. A DEFEND campaign may accept lower incrementality if a competitor is actively bidding on your brand. A CLEARANCE campaign may tolerate lower margin because stock is about to become obsolete. A LEARN campaign may spend without immediate orders because the team needs search-term evidence before building exact-match lanes.
Those permissions should not last forever.
The operator mistake is treating campaign roles like folder names. Once the campaign is called “brand defence” or “launch NL” or “MediaMarkt Q4 push”, the role becomes invisible. Nobody checks whether brand defence is still needed, whether launch has become harvest, whether the Q4 push should have ended, or whether learning has bought enough evidence.
A channel-role expiry board fixes that. Every role receives three things:
- A commercial reason: why this campaign is allowed to spend.
- A proof condition: what must stay true for the role to continue.
- An expiry date or trigger: when the role must be reviewed, downgraded or stopped.
That sounds simple because it is. The hard part is being disciplined when a campaign looks “fine” in the platform but no longer has a valid business reason.
The seven campaign roles a €5K+ account should define
You do not need twenty labels. You need enough role clarity to stop budget drifting.
1. Defend
Defend campaigns protect demand you already own: brand searches, branded product targets, hero ASIN defence, or a retailer placement that keeps a rival from intercepting high-intent shoppers. The expiry trigger is competitor pressure. If no meaningful competitor pressure exists for two review cycles, defence budget should be capped, tested down or moved to a cheaper protection lane.
2. Harvest
Harvest campaigns capture proven demand at acceptable contribution margin. These campaigns should have stricter CPC ceilings, tighter target ACOS and less tolerance for “maybe later” traffic. The role expires when margin, conversion rate, price position or stock cover breaks.
3. Learn
Learn campaigns buy evidence: auto campaigns, broad match, category tests, competitor exploration and MediaMarkt placement tests. The mistake is letting learning become a lifestyle. A Learn role needs a click, spend or time limit. After that, terms graduate, get blocked or stay on a watchlist with a smaller budget.
4. Launch
Launch campaigns support new products or new marketplace entries. They can carry higher ACoS, but only while retail readiness is strong enough: stock cover, content, reviews, price position and fulfilment promise. Launch expires when the product either graduates to Harvest, needs Fix work, or fails its evidence window.
5. Fix
Fix is not a normal scaling role. It exists when paid traffic is used carefully while content, offer, price or fulfilment issues are being repaired. If the fix does not happen by the expiry date, the campaign should not keep spending as if patience were a strategy.
6. Clear
Clear campaigns sell through old, seasonal or overstocked inventory. They must be connected to stock units, discount depth and remaining margin after returns. The role expires when inventory falls below the planned floor or when discount-adjusted margin no longer justifies paid demand.
7. Recover
Recover campaigns rebuild momentum after a stockout, Buy Box loss, content issue or delivery-promise break. They should start small, watch conversion evidence closely and graduate only after the marketplace proves shoppers are responding again.
Example 1: NorthSea Kitchen stops defending a brand term that no longer needs a wall
NorthSea Kitchen sells a premium stainless-steel pan set on Amazon NL and bol.com. The Amazon brand campaign was labelled DEFEND because two competitors had been bidding on “NorthSea pan set” during a spring promotion. The campaign spent €1,200 per month, reported 6.8% ACOS and looked lovely in the agency report.
The expiry board asked a different question: what competitor pressure still justifies this defence role?
The team checked two weeks of search results. Competitor ads appeared on the brand term only 8% of checks. Organic position was stable, the Storefront was first, and branded conversion was 24.6%. At the same time, the pan set’s contribution margin had fallen from €8.40 to €5.10 per order after a steel-cost increase and a temporary €6 coupon.
The old label said “defend”. The current evidence said “cap and test down”. NorthSea reduced the brand defence budget from €1,200 to €450, kept exact brand protection live, and moved €750 into a Harvest campaign for “induction pan set” where max CPC still fit the new margin. FiveX would surface that decision naturally by connecting campaign spend, contribution margin, coupon impact and search-term role in one view instead of leaving the operator to reconcile four exports.
Example 2: StudioLamp graduates bol launch spend before LVB stock becomes the constraint
StudioLamp launches a rechargeable desk lamp on bol.com with LVB fulfilment. The ad service opens a LAUNCH lane with €900 reserved for 21 days. The permission is explicit: accept up to 38% ACOS while the SKU builds search evidence, but only if LVB stock cover stays above 30 days and product-page conversion stays above 7%.
By day 21 the campaign has 162 clicks, 11 orders, €312 ad spend, €1,105 attributed revenue and 28.2% ACOS. On the surface, the launch looks acceptable. But stock cover has dropped from 46 days to 18 days because organic orders accelerated faster than expected. The role has expired.
Without expiry discipline, the campaign would keep “launching” until stockout panic forced a messy pause. With the expiry board, StudioLamp graduates only the winning exact terms into a small Harvest lane, cuts broad exploration by 70%, and shifts the rest of the learning budget to an Amazon variant with 54 days of FBA cover. The launch did its job. Keeping the same role alive would have turned success into a stock problem.
This is exactly why FiveX inventory insights belong in ad management. A bol campaign cannot be judged only by ACOS when LVB stock cover is the real constraint.
Example 3: MediaMarkt clearance gets a hard stop before discount margin disappears
NovaClean has 84 units left of an older cordless vacuum bundle and a newer model arriving in five weeks. MediaMarkt is the strongest channel for the electronics audience, so the team books a CLEAR role: €1,500 maximum retail media spend over two weekends, €32 contribution margin per unit after the planned discount, and a hard stop if sell-through reaches 65 units or return-adjusted margin drops below €18.
The first weekend sells 39 units with €610 spend. Good. The second weekend starts with 45 units left, but support tickets show that buyers are comparing the old and new model because the product content is not clear enough. Return risk rises. The operator does not need to debate whether the campaign has a nice ROAS. The CLEAR role had conditions, and one condition is now shaky.
The team pauses the second-weekend placement for 24 hours, updates the comparison copy, reduces the remaining spend cap to €520 and keeps a stop at 20 units. Campaign logs matter here. In FiveX, the reasoning behind that change should live next to the budget move: not “reduced spend”, but “CLEAR role renewed with lower cap after content-risk review”. That creates memory for the next clearance push.
How to run the weekly channel-role expiry board
The board should be short enough to run every week and strict enough to change money.
Step 1: assign every active campaign one primary role
If a campaign has three jobs, split it or choose the job that controls budget permission. A campaign cannot be Launch, Learn and Harvest at the same time without creating messy rules. Mixed roles create polite reporting and weak decisions.
Step 2: define the role’s proof condition
Proof conditions should be specific. “Good performance” is not a condition. “ACOS below 24% with stock cover above 28 days and contribution margin above €7 per order” is a condition. “At least 80 clicks or €250 spend before search-term graduation” is a condition.
Step 3: set the expiry trigger
Use dates for temporary roles and triggers for commercial risk. Launch might expire after 21 days. Learn might expire after 120 clicks. Clear might expire after 65 units sold. Harvest might expire if price changes by more than 5%, stock cover drops below 21 days or return rate rises above the SKU’s reserve.
Step 4: review exceptions before optimisations
Do not start with bid changes. Start with roles that expired. A campaign with an expired role should not receive normal optimisation until the team decides whether to renew, graduate, cap, shrink, pause or move budget.
Step 5: log the decision in human language
“Bid down 12%” is not enough. The useful log says: “LAUNCH expired after 21 days; exact terms graduated to Harvest; broad test capped because bol LVB stock cover fell to 18 days.” That sentence is gold in the next QBR.
Where automation helps — and where it should wait
Automation is useful once roles are clear. FiveX Ads AI can recommend bid changes, product strategies can shape future targeting, and campaign logs can preserve why a change happened. But role expiry should not be fully delegated too early.
The system can flag expired roles, margin drift, stock-cover breaks and overspend against a learning cap. The operator still needs to decide whether the campaign’s job changes. That is the trade-off: automate detection, keep commercial permission human until the rules have proven themselves.
For €5K+ accounts, this is usually the sweet spot. Let software do the watching. Let the ad service run the budget court. Then, once the role logic is stable, automate the safe moves: cap expired Learn campaigns, reduce Harvest bids when margin drops, pause Clear campaigns when stock floors are reached, and queue approval for cross-marketplace reallocations.
The practical template
Your channel-role board can start with ten columns:
- Marketplace: Amazon, bol or MediaMarkt
- Campaign or placement
- SKU or product group
- Current role
- Monthly or weekly spend permission
- Margin condition
- Stock or fulfilment condition
- Evidence condition
- Expiry trigger
- Decision: renew, graduate, cap, shrink, pause or reallocate
That template is not glamorous. Good. It is not supposed to be. Its job is to stop an ad account from spending because an old label still sounds reasonable.
Final thought: campaigns should earn the right to keep their job
The best marketplace ad services are not the ones that touch the most buttons. They are the ones that know when a campaign’s original reason has expired.
Defence can become overprotection. Launch can become waste. Learning can become avoidance. Clearance can become margin leakage. Recovery can become wishful thinking. Harvest can quietly break when price, stock or returns change.
Give every campaign a job. Give every job proof. Give every proof an expiry rule. Then let Amazon, bol and MediaMarkt spend only where the role still deserves permission.