Skip to main content
    Platform Migration10 min read

    The Replatforming Performance Dip Nobody Budgets For

    Every platform migration plan includes a timeline, a feature checklist, and a launch date. Almost none of them budget for the 4-8 week Google Ads performance dip that follows. It's not a risk - it's a certainty. And the brands that don't plan for it panic into bad decisions.

    The Dip Nobody Plans For

    You've spent six months migrating from Magento to Shopify Plus. The dev team is proud. The new site is faster. The checkout converts better in UAT. You launch on a Tuesday.

    By Thursday, your Google Ads CPA has doubled. By the following Monday, your ROAS has dropped 40%. By the end of week two, the MD is asking whether the migration was a mistake.

    It wasn't a mistake. But the Google Ads performance dip was entirely predictable - and nobody planned for it. Not the dev agency, not the platform consultancy, and often not even the PPC agency managing the account.

    Why It Happens - Every Time

    Platform migrations don't just change your website. They change every signal Google uses to make bidding decisions. Four systems break simultaneously:

    1. Conversion tracking - pixel implementations change, tags fire differently, enhanced conversions need reconfiguring
    2. Smart Bidding models - historical conversion data becomes irrelevant because the new site converts at different rates with different user flows
    3. Product feed URLs - every product URL changes, meaning Shopping ads need to reindex and Quality Score resets
    4. Site speed and UX - even improvements create change, and change disrupts the algorithm's learned patterns

    Tracking Breaks First - And Hardest

    The most immediate impact is tracking. A new platform means a new checkout flow, new thank-you page structure, and often a new GTM container. Even with careful implementation, conversion tags behave differently:

    • Tag firing gaps - the new checkout might redirect differently, causing tags to miss 10-30% of conversions in the first week
    • Enhanced conversions mismatch - hashed email parameters might not pass correctly on the new platform
    • Server-side tracking reset - if you were using CAPI or server-side GTM, the integration needs rebuilding from scratch
    • Consent mode reconfiguration - cookie consent implementations differ between platforms, changing modelled conversion rates

    If Smart Bidding sees 30% fewer conversions in week one - even if real conversions haven't changed - it responds by reducing bids. Lower bids mean less traffic. Less traffic means fewer conversions. The death spiral begins.

    Smart Bidding Loses Its Memory

    Smart Bidding learns from historical conversion patterns - who converts, when, on which devices, at what price points. A platform migration makes all of that history unreliable.

    The new site has a different checkout funnel. Pages load at different speeds. Mobile behaviour changes. The user signals that Smart Bidding learned over months no longer predict conversion probability accurately.

    Google recommends 2-4 weeks of learning phase for any significant change to Smart Bidding inputs. A platform migration is every change simultaneously. Expect the learning phase to be 4-6 weeks for Shopping campaigns and up to 8 weeks for PMax, which has more signal dependencies.

    URL Migration Disrupts Your Entire Feed

    When your product URLs change - and they almost always do in a migration - Google Shopping has to re-crawl and reindex every product. This means:

    • Temporary disapprovals - products may be disapproved until Merchant Centre verifies the new URLs
    • Quality Score resets - landing page quality metrics reset when the destination URL changes
    • Feed attribute mismatches - different platforms structure product data differently, causing feed validation errors
    • Image recrawling - if image URLs change too, Google needs to reprocess all product images

    For a catalogue of 5,000+ SKUs, full reindexing can take 7-14 days. During this period, your Shopping coverage is incomplete. Products that were converting yesterday aren't showing today.

    The Real Recovery Timeline

    Based on managing accounts through dozens of migrations, here's what actually happens:

    Week 1: Performance drops 30-60%. Tracking fires intermittently. Shopping coverage drops as URLs reindex. Smart Bidding enters learning mode.

    Week 2-3: Tracking stabilises. Shopping coverage recovers to ~80%. Performance improves but remains 20-40% below pre-migration.

    Week 4-6: Smart Bidding recalibrates with fresh data. Performance reaches 85-95% of pre-migration levels. The new site's UX improvements start to show.

    Week 7-10: Full recovery. If the new platform genuinely performs better, performance should exceed pre-migration by week 8-10.

    Total impact: 6-10 weeks of below-baseline performance. For a brand spending £30k/month, that's £30k-£80k in lost profit compared to business-as-usual. This is a real cost that belongs in the migration business case.

    How to Mitigate the Damage

    You can't eliminate the dip. But you can reduce its severity and duration:

    Before migration

    • Implement parallel tracking - set up conversion tags on the staging environment and validate before launch
    • Preserve URL structures - if possible, maintain the same product URL paths. Every URL that stays the same is one less disruption
    • Pre-build the feed - have your new product feed validated in Merchant Centre before launch day
    • Time it right - migrate during your lowest-spend month, not before peak season
    • Budget for the dip - plan for 30-50% lower Google Ads contribution for 6 weeks. Build this into the migration P&L

    After migration

    • Monitor tag firing rates daily - compare conversion counts against backend orders. If they diverge, fix tracking immediately
    • Consider manual bidding temporarily - if tracking is unreliable, Smart Bidding will make bad decisions. Manual CPC or portfolio strategies give you control while data stabilises
    • Don't panic-optimise - the worst response is making campaign structure changes during the recovery window. Let Smart Bidding learn on the new site before changing anything else
    • Audit Shopping coverage daily - check Merchant Centre for disapprovals, feed errors, and indexing status until coverage returns to 95%+

    Next Steps

    If you're planning a platform migration and spending £10k+ per month on Google Ads, the performance dip isn't optional - it's physics. What's optional is whether you're prepared for it. We build migration-specific Google Ads strategies that minimise disruption and accelerate recovery.