Replatform migrations destroy more organic traffic than any other commerce work. The mechanism is mostly preventable: URL changes that lose link equity, meta tag changes that disrupt rankings, structural changes that confuse search engines, and 404 explosions on pages that used to rank. The Magento 1 to Adobe Commerce migration has every one of these risks built in, and the migration teams that do not treat SEO preservation as a parallel workstream consistently leak 20-40% of their organic traffic in the launch window. The fix is a disciplined checklist and the operational rhythm to execute it.
This piece walks through the SEO preservation checklist Bemeir’s migration team runs on every Adobe Commerce migration. The categories below are organized by what to protect, what to test, and the failure modes that catch merchants most often.
Step 1: Capture the pre-migration baseline
Before any migration work touches production, capture an honest baseline. The data points that matter:
Full URL inventory. A crawl of the M1 site producing every indexed URL. Screaming Frog, Sitebulb, or a similar crawler does this efficiently. Output: every URL with its current title, meta description, canonical, status code, and link count.
Organic traffic by URL for the last 12 months. From Google Analytics or Google Search Console, the URLs that received organic traffic and roughly how much. This identifies the pages whose preservation matters most.
Keyword rankings for top organic landing pages. The keywords that drive organic traffic on the top 200 URLs, captured via Google Search Console or a third-party tool. This is your “what are we ranking for” snapshot.
Backlink profile. External backlinks pointing to the M1 site, captured via Ahrefs, Moz, or similar. The backlink targets are the URLs whose preservation has the highest direct equity value.
Indexation status. The current state of indexation in Google Search Console: total indexed, excluded, errors. This is the baseline you compare against post-migration.
The baseline is the reference point for everything that follows. Without it, you cannot prove what migrated cleanly and what did not.
Step 2: Plan URL preservation
The single most important SEO decision in the migration is whether URL structures are preserved or changed. The honest options:
Preserve M1 URLs exactly. Configure Adobe Commerce to generate URLs in the same pattern as M1, with the same suffixes, prefixes, and category path inclusion. This is the lowest-risk option and the one Bemeir recommends by default for stores with meaningful organic traffic.
Change URL structure with comprehensive 301 redirects. Adopt Adobe Commerce’s default URL patterns (or any other planned structure) and build a complete redirect map from old URLs to new URLs. This is acceptable when there is a strong reason to change URLs (e.g., the M1 URLs are objectively bad), but it incurs the temporary ranking volatility that comes with any URL change.
Mix of both. Preserve most URLs, change a defined subset. This is the version most stores end up at: keep product and category URLs stable, accept changes to CMS pages or admin-driven URLs where the new structure is clearly better.
The decision needs to be made before development starts, not during. Changing URL strategy mid-migration is expensive and SEO-damaging.
Step 3: Build the redirect map
For every URL that changes, a 301 redirect from the old URL to the new URL needs to be in place at launch. The redirect map needs to be:
Complete. Every URL in the pre-migration inventory must have a corresponding redirect or a same-URL match. Missing entries become 404s on launch.
One-hop. The redirect goes directly from old to new in one step. Chained redirects (old → intermediate → new) work but they accumulate latency and search engines treat them as weaker than one-hop.
Permanent. 301 redirects, not 302 or 307. 301 communicates “moved permanently” and transfers ranking signal. Anything else does not.
Configured at the right layer. Adobe Commerce can store URL rewrites in its url_rewrite table. The CDN or web server can also handle redirects at a faster layer. For high-volume sites, the CDN-level approach is faster and reduces application load. The decision depends on volume and infrastructure.
Tested at scale. A 50,000-URL redirect map cannot be tested by spot-checking. Either crawl every URL post-migration or use the redirect map as input to a targeted crawl that verifies each redirect resolves correctly.
Step 4: Preserve metadata
Every page on the M1 site has metadata that drives ranking and click-through. Preservation tasks:
Title tags. The product and category title tag values from M1 should be migrated to Adobe Commerce verbatim where they are working well. If the migration is also an opportunity to improve titles, the changes should be a deliberate decision per page or per template, not an accidental side effect of the migration tool.
Meta descriptions. Same treatment: preserve where working, improve deliberately where not.
Canonical tags. The canonical configuration on Adobe Commerce should match the M1 configuration. The default Adobe Commerce behavior (canonicals to category-specific URLs vs. canonical product URLs) can differ from M1 and affect indexation.
Structured data. M1 stores often have structured data (Product, Breadcrumb, FAQ, Review schemas) implemented via extensions or theme code. The structured data implementation needs to be replicated on Adobe Commerce; missing structured data is a meaningful organic CTR loss.
Open Graph and Twitter Card tags. Social sharing metadata should be preserved or improved deliberately.
Robots meta and X-Robots-Tag. Per-page indexation controls (noindex on staging, noindex on internal search pages) need to be replicated on Adobe Commerce.
Step 5: Preserve site structure signals
Beyond URLs and metadata, structural signals that affect SEO:
XML sitemap. Adobe Commerce generates sitemaps natively. The configuration should produce a sitemap with the same scope as the M1 sitemap (and ideally cleaner). Submit the new sitemap to Search Console at launch.
robots.txt. The robots.txt rules that worked on M1 should be replicated, with adjustments only for new URL patterns specific to Adobe Commerce.
Internal linking. Header navigation, footer links, breadcrumb trails, related product blocks, and category-product cross-links all carry ranking signal. The Adobe Commerce implementation should match the M1 information architecture or improve on it deliberately.
hreflang for international stores. Multi-language stores need to maintain hreflang implementation on Adobe Commerce. The pattern is different from M1’s typical implementation; verify it works correctly.
Pagination signals. rel=”next” and rel=”prev” on category pagination, if present on M1, should be replicated where useful (note that Google deprecated direct use of these but they still serve as soft signals).
Step 6: Test and validate
The launch-window testing checklist:
| Test | Tool | Pass criteria |
|---|---|---|
| Full crawl of new site | Screaming Frog or similar | All expected URLs respond 200, all changed URLs respond 301 |
| Redirect map validation | Crawler with custom URL list | 100% of pre-migration URLs reach the correct new URL via 301 |
| Title and meta description audit | Crawler comparison | Match or deliberate improvement on every URL |
| Canonical validation | Crawler | Canonicals match the new URL structure correctly |
| Structured data validation | Google Rich Results Test | Product, breadcrumb, and other schemas valid |
| Sitemap submission | Google Search Console | New sitemap submitted, processed, no errors |
| Mobile usability | Search Console | No regressions vs. M1 baseline |
| Core Web Vitals | Search Console + CrUX | Same or better than M1 baseline |
| Robots.txt verification | Direct fetch | Rules match intended configuration |
Every item passes before launch. Items that do not pass become launch blockers.
Step 7: Monitor the post-launch window
The first 30 days post-launch are the high-risk window. The monitoring discipline:
Daily crawl of top 1,000 URLs. Verify they continue to respond 200, retain titles and canonicals, and are not regressing.
Daily Search Console review. Indexation count, crawl errors, manual actions, mobile usability issues, Core Web Vitals.
Weekly traffic analysis. Compare week-over-week traffic by landing page; investigate any landing page dropping by more than 30%.
Bi-weekly ranking check. The keywords from the baseline; investigate any losing significant position.
Backlink check. Are linked URLs still resolving correctly? Any 404s on backlink targets become urgent fixes.
The first three weeks are when temporary ranking volatility happens even on clean migrations. The fourth week is when permanent issues become visible if they exist. Be patient with the volatility, ruthless with the permanent issues.
Failure modes that consistently cause traffic loss
Five patterns appear across migrations that lose meaningful organic traffic:
Incomplete redirect map. Several thousand URLs were missed in the inventory and have no redirects. Search engines find 404s and de-index the affected pages.
Canonical regression. Adobe Commerce’s default canonical behavior differs from M1, and product variants or category-path duplicates now compete with their canonicals.
Structured data not replicated. Rich result eligibility for products, FAQs, or recipes is lost because the schema implementation did not migrate.
Internal linking thinned. The new theme has fewer footer links, fewer breadcrumbs, or fewer related-product blocks than the M1 theme, concentrating ranking signal on fewer pages.
Robots or noindex misconfiguration. A staging environment configuration accidentally goes to production, or an Adobe Commerce default rule (e.g., on internal search results) is more aggressive than the M1 equivalent.
The merchants who execute the checklist above avoid all five categories. The merchants who skip it experience some combination of all five.
What good post-migration SEO looks like
A clean Magento 1 to Adobe Commerce migration should produce:
- Less than 5% temporary organic traffic dip in weeks 1-3
- Recovery to baseline by week 6-8
- Improvement above baseline by month 3-6 from the Core Web Vitals gains
- Structured data eligibility preserved on day one
- Indexation count within 5% of pre-migration baseline
- No manual actions, no security issues, no crawl error spikes
Bemeir treats SEO preservation as a launch-blocker, not a launch-nice-to-have. The migrations that go well are the ones with this discipline; the migrations that lose traffic are the ones where SEO was treated as a downstream effect of the application work. The Adobe Commerce SEO documentation covers the platform-level controls; the checklist above is what bridges those controls to actual preservation in the launch window. The work is well-understood; what remains is the operational rigor to execute it on every page.
