Website migration guides are usually written for the technical handover: move the WordPress site, point DNS, check redirects, confirm analytics, fix a few broken links and celebrate when the new environment loads. Flywheel’s WP Engine Site Migration guidance is a good example of that practical mindset. It explains how to move a site from another host or live server with a migration plugin, which is exactly what a web team needs.
Paid search migration checklists add another useful warning. Seer Interactive’s PPC migration advice is blunt: if developers underestimate the advertising side, ad URLs can redirect to the wrong pages, conversion tracking can break and paid traffic can drop. Also true. Also still not enough for a marketplace ad account spending €5K, €15K or €40K a month across Amazon, bol and MediaMarkt.
The named mistake I see is letting the website migration team own marketplace ad risk. The Shopify theme, WordPress host or brand-site URL changes. Everyone checks Google Ads and SEO. Meanwhile Amazon Sponsored Brands still points to an old Store page, bol campaign traffic keeps pushing a SKU whose product content was changed for the new brand positioning, and a MediaMarkt retail media placement links to a campaign page that now shows a different bundle. The migration was technically successful. The ad account lost its commercial map.
My stance: every Advertentie Service managing more than roughly €5K monthly spend needs a marketplace ad site-migration freeze. Not because Amazon Sponsored Products always send shoppers to your own site. Most do not. The freeze matters because a site migration rarely changes only the site. It changes URLs, tracking, product feeds, creative claims, bundles, landing-page logic, brand-store routing, stock allocation, analytics attribution and sometimes the product economics behind the campaigns. If spend keeps moving while those things are unsettled, the account can buy traffic into a version of the business that no longer exists.
This guide is for brands in the Netherlands and Belgium using Amazon Ads, bol Sponsored Products and MediaMarkt retail media with an agency or specialist operator. It borrows the discipline of web migration checklists, but translates it into the missing layer: which marketplace ad decisions must pause, shrink, validate or reopen when the digital shelf around them changes.
What the existing migration advice gets right
The public advice is useful in its own lane.
Flywheel and WP Engine-style migration guides focus on moving the site cleanly: source environment, destination environment, plugin connection, database and media transfer, DNS, SSL and post-launch checks. The core lesson is operational: do not improvise a migration when a repeatable flow exists.
Paid search migration guides add the performance layer. Seer’s article highlights the danger of wrong redirects, broken conversion tracking and paid campaigns sending traffic to pages that no longer work. That is the right fear for Google Ads, Microsoft Ads, Meta or any channel where the landing page sits on your own domain.
Amazon PPC handoff content adds another relevant point. Ad Badger’s “hand off your Amazon PPC without blowing up your brand” discussion argues for a slow handoff ladder instead of dumping full control on a new operator immediately. That idea is very useful here: migration risk should be released in stages, not handed back to full spend in one confident Monday morning.
What these sources usually miss is the marketplace-specific dependency map. Amazon, bol and MediaMarkt ads are not only linked to URLs. They are linked to offer strength, stock, product content, retail-media placements, attribution windows, search-term history, brand-store paths and margin assumptions. A website migration can disturb all of those without creating a classic 404 error.
The operator rule: freeze decisions, not the whole account
A site migration freeze does not mean “turn off all ads”. That is the panicked version. It can protect margin, but it can also kill ranking, lose learning and hand competitors a quiet week.
The better rule is: freeze decisions whose evidence just became unreliable.
If an Amazon Sponsored Products campaign points to a stable PDP, stock is healthy, price is unchanged and the product economics did not move, it may keep running with normal guardrails. If a Sponsored Brands campaign routes to a Store page that was rebuilt to match the new site navigation, it needs a validation window. If MediaMarkt retail media promotes a bundle whose landing-page copy, stock allocation or price logic changed during the migration, it should not scale until the new page proves conversion and margin again.
That is where FiveX fits naturally. FiveX connects advertising performance with product profitability, stock, pricing and channel data. During a migration, that matters because the ad platform alone will only tell you clicks, spend and attributed sales. It will not always tell you that a €42 average order now contributes €4.10 less because the bundle mix changed, or that the promoted SKU has only nine days of stock after inventory was reallocated to the new webshop launch.
The five ledgers every migration freeze needs
1. The route ledger
The route ledger records where paid marketplace demand lands after the migration. For Amazon, that means Store pages, Sponsored Brands landing pages, Sponsored Display destinations and any external attribution links. For bol, it means product detail pages, brand pages where relevant, promotion pages and campaign-specific product selections. For MediaMarkt, it means retailer placements, category pages, brand shelves, product pages and any custom retail-media destination.
The rule is simple: if the destination changed, the old performance evidence becomes reference evidence, not permission evidence. Useful, yes. Still not enough to scale.
Named example: NorthSea Homeware migrates its brand site and rebuilds the Amazon Store to mirror the new category structure. Before migration, Sponsored Brands sent traffic to a “Coffee & Tea” Store page with 14 products and converted at 9.8% with a 23% ACOS. After migration, the new Store page leads with gift bundles. The first 340 clicks convert at 6.1% and average contribution margin drops from €11.40 to €7.80 because shoppers choose the lower-margin bundle. The freeze rule keeps daily Sponsored Brands spend capped at €85 until the new route clears 8% conversion and €9 contribution per order for three consecutive days.
2. The tracking ledger
Marketplace teams often assume tracking is a webshop problem. That is too narrow. Amazon Attribution links, DSP pixels, Store Insights interpretation, bol campaign reporting exports, MediaMarkt retail media reports and CRM or BI imports can all be affected by naming, UTM, landing-page, feed or analytics changes around a migration.
The tracking ledger asks: can we still trust the source, campaign, SKU and order mapping enough to move budget?
FiveX helps here by turning reporting into a cross-marketplace view instead of leaving the operator to reconcile platform screenshots. If spend appears in Amazon, sales appear in a marketplace report, stock moves in the catalog and margin changes in finance, the migration period needs one operating truth. Otherwise every channel looks “fine” in isolation while the combined P&L gets fuzzy.
3. The feed and content ledger
Site migrations often trigger product-content work: new titles, new image crops, rewritten benefits, changed bundles, updated variant names, different category navigation. That content may then be copied into Amazon, bol or MediaMarkt because everyone wants consistency. Sensible. Dangerous if advertising keeps the old keyword and conversion assumptions.
Named example: BrightNest Lighting changes its webshop category language from “desk lamps” to “home office lighting” during migration. The marketplace team copies the language into bol titles and bullets for six SKUs. One bol Sponsored Products campaign keeps bidding on “bureaulamp zwart” at €0.46 CPC because it was historically profitable at 19% ACOS. After the copy shift, click-through improves, but conversion falls from 11.2% to 7.4% because shoppers looking for a simple desk lamp land on more lifestyle-led content. The freeze ledger lowers the bid to €0.31 and moves €450 of weekly test budget into exact terms that still match purchase intent.
4. The stock and allocation ledger
A migration can be a launch moment. Teams send email, update social, refresh the brand story and sometimes push traffic to the webshop. That may pull stock away from marketplace channels. If Amazon, bol and MediaMarkt ads keep running from last month’s stock assumptions, they can create a stockout that looks like “good demand” until ranking and margin suffer.
The stock ledger sets minimum cover before spend reopens. For example: no scaling if stock cover is below 21 days for replenishable hero SKUs, below 14 days for seasonal SKUs, or below the event-specific threshold agreed with operations. FiveX inventory insights make this practical because the ad decision can see the product’s stock runway before budget moves.
Named example: DeltaFit Accessories launches a new site and sends a 25,000-subscriber email to the same hero fitness strap that Amazon and MediaMarkt are advertising. Amazon stock cover drops from 32 days to 13 days in four days. The Amazon campaign still shows 31% ACOS against a 35% target, so the ad interface says “fine”. The stock ledger says “not fine”: cap Amazon at €60/day, pause MediaMarkt prospecting and keep only branded defence live until inbound stock is confirmed.
5. The margin ledger
Migration projects love new bundles, new price architecture and new promotional mechanics. The ad account must not inherit those changes blindly. A page can convert better and still be worse for profit if the order mix shifts toward lower-margin products, delivery costs rise or a launch discount sticks around too long.
The margin ledger updates break-even ACOS, target ACOS and max CPC after the migration. Not next month. Before spend reopens. FiveX profitability dashboards make this less theatrical: SKU margin, ad spend, marketplace fees and operational costs can be reviewed together instead of turning the migration review into a debate between marketing and finance.
A practical 10-day migration freeze calendar
Day -5 to Day -3: map every paid route. Export campaigns, destinations, top SKUs, campaign roles, daily budgets, target ACOS, current CPC ceilings, stock cover and contribution margin. Label each campaign as safe, route-sensitive, tracking-sensitive, content-sensitive, stock-sensitive or margin-sensitive.
Day -2: lock risky decisions. Freeze bid increases, new broad-match expansion, new competitor targets, new MediaMarkt placements and new Sponsored Brands routes unless they are part of the migration plan. Do not let automation discover new demand while the commercial map is moving.
Day 0: migrate, but keep the account boring. Maintain essential defence and proven harvest campaigns. Keep learning budgets small. No heroic scaling because the new site “looks better”. Looking better is not evidence.
Day +1 to Day +3: validate routes and tracking. Check that campaign destinations resolve, Store pages show the intended products, bol product content did not break intent, MediaMarkt placements still match the promoted SKU, and reporting joins spend to product and margin correctly.
Day +4 to Day +7: reopen by evidence tier. Safe harvest campaigns can return first. Route-sensitive campaigns need conversion and margin evidence. Content-sensitive campaigns need search-term review. Stock-sensitive campaigns need runway confirmation. Margin-sensitive campaigns need updated break-even ACOS.
Day +10: expire the freeze. A freeze should not become a permanent excuse. Every campaign gets one of four labels: reopen, cap, rebuild or retire. FiveX automation logs and Ads AI recommendations can support that review, but the operator still owns the commercial decision.
What your agency should show you
If an Advertentie Service manages your account, ask for a migration freeze pack. It does not need to be pretty. It needs to be decisive.
- Which campaigns were frozen, capped or left open?
- Which destinations changed?
- Which SKUs had margin, stock, price or content changes?
- Which target ACOS or CPC ceilings were recalculated?
- Which automation rules were paused or put in shadow mode?
- Which campaigns can reopen by evidence, and which require commercial approval?
The answer should name products and numbers. “We monitored performance” is not enough. “We capped BrightNest BL-240 on bol from €120/day to €55/day until conversion returns above 9% and stock cover is back above 21 days” is an operating decision.
The FiveX view: migration is a permission reset
A migration is not just a technical project. For marketplace advertising, it is a permission reset. The account needs to ask again: do we still trust this route, this content, this stock level, this margin and this evidence?
That is the operator voice I would want in the room. Not anti-growth. Not afraid of change. Just unwilling to let a successful site migration quietly rewrite the rules of a €5K+ ad account.
FiveX helps marketplace teams run that reset with connected ad, product, inventory and profitability data. Your agency can still optimise bids. Your brand team can still improve the site. Finance can still protect margin. The difference is that every post-migration euro has to earn its permission again before Amazon, bol or MediaMarkt spend returns to full speed.
That is how you migrate the site without migrating yesterday’s assumptions into tomorrow’s budget.