A stockout does not end when inventory comes back. That is the bit most marketplace ad accounts get wrong. Amazon shows units available again, bol.com accepts the offer, MediaMarkt has the product live, and someone switches the campaigns back on because the problem is “fixed”. Lovely. Also a very efficient way to buy expensive recovery clicks before the SKU has earned the right to spend again.
The named mistake is the revenge restart. A hero SKU goes dark for six days, the team panics about lost ranking, then restarts every campaign at the old bids the moment stock returns. The dashboard looks active again. The P&L is less impressed: CPCs are higher, conversion is weaker because delivery promises are not fully stable, competitors used the gap to win placement, and the first €600 of renewed spend mostly pays for algorithmic warm-up rather than profitable growth.
My stance: after a stockout, marketplace ads should not be restarted. They should be re-admitted. That means the SKU passes a short commercial permission check before budget goes back to normal. Stock must be deep enough. The offer must be eligible. Price must be competitive. Reviews and content must still convert. The break-even ACOS must be recalculated with the recovery discount or expedited freight included. Only then should Amazon Sponsored Products, bol Sponsored Products or MediaMarkt retail media receive full permission again.
This guide is for brands spending from roughly €5K per month across Amazon, bol.com and MediaMarkt. At that level, one sloppy restart can burn a meaningful part of the month. The goal is not to “make up for lost time”. The goal is to rebuild profitable demand without teaching the ad account that every recovery window deserves unlimited budget.
What the existing advice gets right
The research is useful, especially around the first half of the problem. Optmyzr makes the practical case for pausing Amazon ads before inventory runs low. It points out that low inventory can create longer delivery windows, weaker conversion and rising ACOS, and it shows how inventory reports can feed rule-based campaign pauses. That is sensible operating hygiene.
Mayan explains the broader damage clearly: stockouts hurt sales, seller credibility, customer experience, organic visibility and PPC eligibility. Its key warning is simple: a campaign may appear “running” while the product is ineligible, which makes the account look healthier than it is.
Adbrew goes deeper on the hidden cost. Stockouts do not only create lost sales; they disrupt PPC history, weaken ranking momentum, hurt cash flow and force sellers to spend more when they restock. Sellerboard makes a similar point from the sales-velocity angle: when sales velocity drops, organic rank, Best Seller Rank, review flow and PPC effectiveness can all decline together.
The seller conversations are even more useful because they are messier. In Amazon seller forums and Reddit threads, operators worry about the trade-off: pause too early and you may lose momentum; keep spending and you accelerate the stockout; restart too aggressively and the recovery period becomes expensive. One Reddit seller described PPC still running while shoppers saw “in stock soon”, creating clicks that did not convert. Another seller worried that switching off a complete campaign could cause sales to drop for a week or two after stock returned.
BidX, Podean and Flywheel-style marketplace agencies mostly cover the adjacent principles: automation, budget control, localization, media efficiency and retail media maturity. BidX is strong on automated bids, budgets and scheduling. Podean’s Amazon Ads guidance stresses that media budget should not drive traffic to listings that cannot convert. Flywheel-type retail media thinking focuses on full-funnel marketplace growth and performance systems. The common gap is the same: most advice says when to pause or how to recover rank, but not how to decide which campaigns deserve budget on day one, day three and day seven after inventory returns.
The missing angle: restart permission, not restart speed
Most stockout playbooks treat recovery like a race. Get stock back. Push PPC. Run a coupon. Recover rank. Move on. That is sometimes right, but only for SKUs with enough margin and operational readiness to absorb the recovery cost.
For a managed advertising service, the better question is not “how fast can we restart?” It is: what level of spend is this SKU allowed to use while the marketplace relearns demand?
I use four restart permissions:
- Hold: ads stay off because the SKU is technically back but commercially unsafe.
- Trickle: low-budget exact and branded campaigns restart to test eligibility, conversion and delivery quality.
- Rebuild: proven keywords and product targets restart with controlled budgets and tighter ACOS caps.
- Scale: the SKU returns to normal or above-normal budget because stock, margin and conversion have recovered.
This is where FiveX fits naturally. In FiveX, the ad operator can see marketplace analytics, SKU profitability, inventory cover and advertising performance in one workspace instead of asking finance, operations and the agency for three separate files. The restart decision becomes a visible permission label, not a Slack message that says “stock is back, please restart”.
The five checks before any stockout restart
Before ads come back, run five checks.
1. Stock cover after the restart, not stock on hand today
“Back in stock” is not enough. A SKU with 280 units available and a normal sell-through of 35 units per day has eight days of cover before ads. If PPC recovery adds 20 units per day, cover falls below six days. That is not a restart. That is a second stockout with better lighting.
Use a restart threshold that includes paid demand. For hero SKUs, I like at least 21 to 30 days of stock cover after expected ad-driven velocity. For slower SKUs, 14 days may be enough. For MediaMarkt electronics or bulky bol.com products with long replenishment lead times, use a stricter threshold.
2. Offer eligibility and delivery promise
Amazon Sponsored Products may not serve properly if the offer is not eligible, the Buy Box is unstable or fulfilment is still settling. bol Sponsored Products can push traffic to a page where delivery promise, seller rating or price position is weaker than before. MediaMarkt retail media has the same basic truth: traffic does not fix a weak offer.
Check the live product page before restarting. Is the advertised offer visible? Is delivery competitive? Is the price still within the intended band? Did a competitor take the attractive slot while you were gone?
3. Loaded break-even ACOS
Recovery periods often carry extra costs: air freight, emergency fulfilment, a restart coupon, lower price, marketplace storage changes or higher return risk after delayed delivery. If the normal break-even ACOS was 28%, the recovery break-even may be 19% this week.
This is where FiveX profitability dashboards are useful. Instead of optimizing to the old campaign target, the operator can load updated COGS, marketplace fees, fulfilment cost, ad spend and return assumptions into the SKU view. The campaign does not get its old ACOS target back until the SKU economics justify it.
4. Conversion evidence from the first 48 to 72 hours
Do not judge the first recovery clicks like a normal week. The ad system, organic rank and shoppers are all re-learning. But do not ignore them either. Watch click-through rate, conversion rate, cost per order, delivery-related page changes and whether sales come from branded, exact generic or competitor targets.
The first 48 to 72 hours should answer: is the SKU merely available, or is it commercially back?
5. Campaign memory
Do not restart every campaign because the SKU used to be important. Separate the old winners from the old passengers. Branded defence, exact high-intent generic keywords, proven product targets and broad discovery should not receive the same recovery budget.
FiveX advertising automation and AI recommendations can help here by surfacing which campaigns previously converted within the SKU’s margin limits and which ones only looked good because stock was abundant or attribution was generous.
A practical 7-day restart model
Here is the restart model I would use for a €5K+ managed marketplace ad account.
Day 0: stock lands, ads do not automatically return
Confirm inventory is live in the marketplace, not just received in a warehouse file. Check product page availability, delivery promise, price, Buy Box or offer eligibility, content status and review rating. Recalculate loaded break-even ACOS. Then assign the SKU one of four labels: Hold, Trickle, Rebuild or Scale.
Days 1-2: trickle spend proves the shelf is alive
Restart branded defence and the top exact targets only. Keep budgets deliberately small. For Amazon, this may mean Sponsored Products exact campaigns and a modest Sponsored Brands defence if brand queries matter. For bol, restart the tightest Sponsored Products groups. For MediaMarkt, restart the placements closest to buying intent, not broad reach.
The goal is not volume. The goal is proof: impressions are eligible, clicks can convert, orders are not wildly above break-even and the delivery promise is not hurting conversion.
Days 3-4: rebuild only the proven lanes
If the first two days show acceptable conversion, restart the proven generic targets and product targets. Keep discovery, broad match and competitor conquesting capped. Recovery budget should lean toward demand you already understand.
This is also the moment to check TACoS. If ad sales recover but total sales do not, the restart may be buying demand that organic would have regained anyway. That does not mean stop. It means do not confuse recovery activity with incremental growth.
Days 5-7: scale only if stock and margin still agree
After a week, the SKU can return to normal budget only if four signals agree: stock cover remains above the threshold, conversion rate is near the pre-stockout baseline, loaded ACOS is below the recovery cap, and total contribution margin is improving. If one signal fails, stay in Rebuild or drop back to Trickle.
This sounds strict because it is. A stockout is not a holiday from commercial discipline.
Three named examples
Example 1: NorthSea Naturals on Amazon.de
NorthSea Naturals sells a magnesium spray on Amazon.de. Before the stockout, the SKU sold 42 units per day, had a €23.95 selling price, €8.40 contribution margin before ads and a normal break-even ACOS of 35%. Amazon Sponsored Products spent €3,200 per month at 24% ACOS.
Then the SKU stocked out for five days after a late inbound shipment. When inventory returned, only 620 units were immediately available. At the old paid velocity, that looked like roughly 15 days of cover. But the team planned a 10% recovery coupon and had paid €1.10 extra per unit for expedited freight, reducing contribution margin before ads to €4.91. The temporary break-even ACOS fell to about 20%.
A revenge restart would put the old €110 daily budget back into the account. The better move is Trickle for two days: €25 per day on branded and exact “magnesium spray” targets, no broad discovery, no competitor conquesting. If conversion returns within 15% of baseline and stock cover remains above 14 days, move to Rebuild at €60 per day. The SKU does not get its old budget until the coupon ends or margin recovers.
Example 2: BorealBaby on bol.com
BorealBaby sells a stroller organiser on bol.com in the Netherlands and Belgium. The product normally sells for €34.99, with €11.20 contribution margin before ads, 18% target ACOS and 1,100 units across fulfilment. bol Sponsored Products had been efficient at 13% ACOS, mostly on category and competitor terms.
A supplier delay caused an eight-day stockout. When stock returned, only the Dutch fulfilment lane was stable; Belgian delivery showed a longer promise for the first three days. The old campaign grouped NL and BE traffic together, so restarting it would hide the problem.
The restart permission is Rebuild for NL and Hold for BE. Split the campaign. Put €45 per day behind Dutch high-intent terms, keep Belgian campaigns paused until delivery is competitive again, and use FiveX inventory insights to monitor when Belgian stock cover crosses 21 days. This is not being timid. It is refusing to pay for clicks in a market where the offer is not ready.
Example 3: VoltEdge on MediaMarkt
VoltEdge sells a USB-C docking station through MediaMarkt marketplace and Amazon.nl. The MediaMarkt retail media campaign had a healthy 5.1 ROAS before stock ran out, but the SKU has tight economics: €79.99 selling price, €17.60 contribution margin before ads and a practical break-even ACOS around 22% after returns.
During the stockout, two competitors dropped price by €6 and collected the category traffic. When VoltEdge came back, the team wanted to spend €900 in the first week to recover share. The problem: matching the competitor price would reduce contribution margin to €11.60, cutting break-even ACOS to 14.5%.
The right restart is Hold for broad retail media and Trickle for exact product-intent placements. Spend €30 per day for three days, keep the price stable, and watch whether conversion survives the price gap. If conversion is weak, the decision is not “increase ads”. It is a commercial decision about price, bundle value or channel priority. FiveX marketplace analytics can show whether Amazon.nl is currently a better use of the same budget while MediaMarkt price position is weak.
What your agency should report after a stockout
If you use a marketplace advertising service, do not accept “campaigns restarted” as the update. Ask for a restart report with six lines:
- Stock cover today and stock cover after expected paid velocity.
- Offer status by marketplace: eligible, delivery promise, Buy Box or price position.
- Loaded break-even ACOS during the recovery window.
- Campaigns restarted, campaigns held and the reason for each hold.
- First 48-72 hour conversion evidence.
- Decision for the next seven days: Hold, Trickle, Rebuild or Scale.
This report is short on purpose. The longer the restart deck, the easier it is to hide the actual decision.
The operator’s checklist
- Never connect “inventory back” directly to “all ads back”. Add a permission step.
- Use paid-velocity stock cover, not static stock on hand.
- Recalculate break-even ACOS with recovery costs included.
- Restart exact and branded demand before broad discovery.
- Split markets when fulfilment or price position differs between NL and BE.
- Review 48-72 hour conversion evidence before adding budget.
- Scale only when stock, margin, offer and conversion all agree.
How FiveX helps
FiveX helps ecommerce teams treat stockout recovery as an operating decision, not a campaign toggle. Marketplace analytics show how Amazon, bol and MediaMarkt sales changed before, during and after the stockout. Profitability dashboards show the updated SKU margin after freight, fees, discounts, returns and ad spend. Inventory insights show whether the restart will create a second stockout. Advertising automation and AI recommendations help identify which campaigns deserve Trickle, Rebuild or Scale permission.
That is the difference between “we restarted ads” and “we rebuilt profitable demand”. The first is a task. The second is management.
If your brand spends more than €5K per month on marketplace ads, stockout restart rules should be part of the advertising service agreement. Not because stockouts are rare. Because they are common enough to deserve a system, and expensive enough to punish improvisation.