Establish the baseline before changing the site
“SEO Migration Services” owns URL-move implementation; this checklist gathers pre-redesign acceptance inputs. Define the affected site sections and customer tasks first. Select representative products, services, categories and articles rather than inspecting only the homepage. For each example, retain its URL, response behaviour, owning purpose and important content. If organic landing-page and business data are available, state the source, period and definition. If access is missing, leave it missing. Without a baseline, a numerical claim about redesign impact is unsupported.
Priority pages and customer tasks
A heavily visited page may matter, but traffic is not the only priority criterion. Finding the right service, discovering a product or retaining support information can also justify protection. Ask the business owner to describe the task that must survive. If an existing URL owns a distinct question, the new design should not silently transfer that ownership to a different page with a broader purpose.
Versions and observation dates
The old content, design approval and technical review may come from different moments. Identify the version represented by each record. Original publication dates remain separate from the creation date of this new source. A later screenshot is not necessarily proof of the earlier state used in a decision. Shared examples should omit customer details, administrative sessions and values that grant access to private systems.
| Field | Required input | Evidence | Acceptance boundary |
|---|---|---|---|
| Priority URL | Existing owner and task | Dated content/response record | A sample does not certify the entire site |
| Organic entry | Period and source | Authorised export | Missing access is not zero traffic |
| Content revision | Publication/update distinction | Source version and dates | No arbitrary backdating |
| Contact path | Expected customer task | Controlled path test plan | A button is not lead or CRM delivery |
Separate visual change from a URL move
“Hreflang and Multilingual Website Checklist” supports multilingual address decisions; first establish whether a change is actually needed. Headings, navigation and layout can change while addresses remain stable. First decide which URL actually needs to change and why. Adding unchanged addresses to a redirect plan can create unnecessary confusion. In a multilingual site, keep each language owner separate; one language’s address decision should not automatically rewrite another. Equivalent content and approved publication are required before assuming that a new language alternative can be linked.
A genuine destination for each changed address
Choose a replacement that fits the old page’s meaning and customer task. An unrelated product or service is not a suitable match merely because it is reachable. Record the old URL, intended target, reason and expected response. The verification plan should distinguish the initial response, redirect chain and final content. This source applies no redirect: actual behaviour must be inspected in the relevant release environment.
Content without a replacement
Discuss why information is being withdrawn and whether a genuine alternative exists. Sending every removed address to the homepage does not answer the lost task. An appropriate removal response, explanation or real successor needs a case-specific decision. Do not invent equivalence between unrelated pages. The content owner should accept the intended behaviour; that decision cannot guarantee which address Google will select or when an old result will change.
| State | Customer need | Planned decision | Required evidence |
|---|---|---|---|
| Stable address | Reach the same task | Preserve owner and path | Old/candidate comparison |
| Changed address | Reach a genuine successor | Choose a justified target | Chain, final response and content acceptance |
| Withdrawn content | Accurate explanation or alternative | Choose a state-appropriate response | Business owner’s rationale |
| Language candidate | Equivalent customer need | Wait for publication acceptance | Actual language owner and public eligibility |
Two different changes
New appearance
Existing URLs and topic owners may remain.
New address
A genuine successor and response acceptance are needed.
Compare important content and paths in the new templates
“Technical SEO Audit” supports template inspection; a visual mockup alone is not acceptance. A design mockup does not prove that the real candidate page supplies the same information. Compare the main heading, explanation, important sections, product or service facts and contextual links in old and candidate versions. Include mobile and initial interaction states. Information existing only inside an image or closed panel needs attention. Do not label every difference a defect; explain which task or access condition the difference affects.
Preserve meaning and scope
Copy may become shorter or appear in a different order; identical wording is not the acceptance condition. A changed service boundary, exclusion, pricing condition or product property is more than a design variation. The relevant business owner must approve that substantive change. Preserve the original publication history and use a real material-update date. A fresh-looking date must not imply an editorial improvement that did not occur.
Links and contact tasks
Plan checks for menus, cards and paragraph links reaching the correct owning page in the correct language. Hidden, broken or misdirected paths can interrupt a customer task. A visible contact button does not establish server acknowledgement or delivery of a request. CTA access, actual lead capture and CRM delivery require different evidence. This text defines acceptance needs; it does not enable a form or activate tracking.
Review preview and production configuration together
“CI/CD Release Checklist” owns the relevant technical decision; preview and production values need joint verification. Name the hostname, environment and configuration used to inspect the candidate. Preview access or indexing restrictions can leak into production, while production values can be inappropriate in a private preview. Responses, robots controls, canonicals and sitemaps should describe the same page decision. A setting’s name or a configuration screenshot is insufficient. Acceptance needs the actual response and the version intended for publication.
Canonical, language and sitemap alignment
Record the real page identified by the canonical and whether its language equivalents have publication approval. Sitemaps should align with accepted eligible URLs rather than a list of proposed addresses. A canonical in source data may differ from the emitted response. Request comparable evidence across representative templates. A sitemap submission or successful response is not indexing acceptance; Google’s selected canonical and indexed state are separate observations.
Media and truthful dates
Covers, social images and other important media must work at their intended destinations. A preview-host URL, missing asset or unrelated image is a distinct finding. Include dimensions, alternative text and mobile presentation in the review plan. Do not backdate original publication or randomise dates across a batch. Material-update and sitemap timestamps should reflect meaningful real changes, not simply the time a build was produced.
Name the launch decision and recovery owner
“Web Design and Development” supports release responsibilities; blockers and recovery acceptance still need a record. A high completion percentage can coexist with a blocking defect. State the decision rule for an inaccessible service page, wrong destination or missing main answer. Identify the team providing evidence, the role accepting it and the person deciding whether to launch. Technical, content and business approval should not collapse into one unexplained tick. The remaining risk needs an explicit owner even when most ordinary checks have passed.
Stop and correction conditions
For each important issue, record expected behaviour, observed difference and required correction. Unknown critical states are not ready. An approved exception needs a reason, scope and recheck time. Where design, content, infrastructure and URL changes can be separated, staged acceptance may make the evidence clearer. Do not promise a platform’s easy migration or safe one-click recovery feature unless its actual implementation and conditions have been verified.
Returning to an earlier version
A recovery plan covers more than an older code commit. Content, redirects, environment settings, caches and measurement definitions need to remain compatible. Specify the version that can be restored and the customer tasks to recheck afterwards. This checklist performs no rollback. The release owner must establish access and real system behaviour before treating a plan as executable evidence; no fixed traffic recovery period follows from its existence.
Launch decision
Stop
Critical unknowns or customer-task blockers remain open.
Proceed
Requires authorised acceptance tied to the actual version.
Observe the same examples after release
“Traffic Loss After a Website Redesign” is the separate diagnosis owner when an actual loss occurs; first observe the same examples. Priority URLs, redirects and fundamental tasks need a fresh check on the published version. Local preview evidence cannot stand in for its production counterpart. Record version, date, environment and method with the observation. Search visibility, indexing and business outcomes require their own source access. This source-generation task did not crawl the website, inspect a Search Console account or trigger any Google processing.
Movement is not an explanation
When evidence is available, compare equivalent periods, channels and page types with the baseline. Campaigns, seasonality, pricing or measurement changes can affect the same chart. Do not interpret every fluctuation as a penalty or redesign failure. If a real issue is found, hand a supported task to the appropriate diagnosis owner. Missing evidence should lead to a request for a defined observation, not an invented cause.
The delivery boundary of this checklist
Complete the worksheet’s inputs, evidence, owners and acceptance conditions using actual records. These specifications do not establish a delivered download, lead acknowledgement or CRM flow. Native language, account, capacity, routing and release approvals remain separate; FA public routes stay held. The coordinator owns private previews and normal CI acceptance. A count of authored checklist items is not proof that the working lead-checklist requirement has been met.
What Does This Checklist Cover?
Start with page and query inventories and technical accessibility. Prioritize findings by business impact, implementation effort and ownership. Compare new content with existing pages serving the same intent. Include post-implementation checks and Search Console monitoring; rankings or results by a fixed date are not guaranteed.
Preparation: Record current traffic
Baseline information and ownership
- Record current traffic
- Inventory important URLs
- Map new page equivalents
Implementation and Test Checks
Record evidence for each check
- Check content loss
- Test redirects
- Review canonical changes
- Update internal links
Handover and Follow-Up Checks
Next actions and acceptance
- Check sitemap scope
- Inspect pre-launch noindex
- Monitor queries after release
How to Use the Findings
Mark each item complete, incomplete or not applicable. Attach a screen, URL or test result to incomplete items. Address findings that block critical user tasks first. Share the list with the team and assign ownership and a retest date.
Checklist evidence worksheet
| Check | Input | Evidence | Owner | Acceptance |
|---|---|---|---|---|
| Define redesign scope | Templates, components and customer tasks | Changed/retained area list | Business owner and project lead | Sample coverage limits recorded |
| Record the SEO baseline | Important URLs and existing content | Dated response and content record | SEO owner | Baseline and source gaps clearly separated |
| Match versions and periods | Content/code revision and data period | Revision and export identifiers | Analytics and technical team | Old/new examples tied to correct revisions |
| Choose genuine URL successors | Old address, target and reason | URL mapping and intended response | SEO and content owner | No unrelated or blanket homepage target |
| Review withdrawn content separately | Removal reason and real alternatives | Business decision and response plan | Business owner | Removal or equivalence is justified |
| Preserve language ownership | Language URLs and equivalent content | Approved language-owner mapping | Localisation owner | FA and new alternates not activated without acceptance |
| Compare main-content meaning | H1, scope and important explanation | Old/candidate content comparison | Content and business owner | Scope differences explicitly accepted |
| Test links and CTA paths | Menu, card and paragraph destinations | Customer-path test in correct language | UX and technical team | Correct owner reached; lead delivery separate |
| Check staging boundaries | Environment, robots and access settings | Actual response/environment comparison | Release owner | Preview restriction does not leak to production |
| Align canonical and sitemap decisions | Actual owner URL and eligible addresses | Representative head/response/XML record | SEO and release team | Preference consistent; Google selection separate |
| Verify media and truthful dates | Cover/OG, publication and material update | Actual asset and date source record | Editor and media owner | No missing environment asset or backdating |
| Record launch blockers | Critical difference, exception and owner | Reasoned go/no-go record | Authorised release decision owner | Unknown critical behaviour is not ready |
| Plan recovery feasibility | Code, content, redirects and settings | Actual version/access acceptance plan | Release and operations | No feature or recovery guarantee |
| Recheck production and delivery separately | Priority URLs, tasks and data sources | Actual version record; separate delivery server evidence | SEO, release and delivery owners | Local test is not indexing or CRM delivery |
Frequently asked questions
Does every redesign need a URL move?
No. Layout and content can change while existing addresses remain; genuine address changes need a separate decision and evidence.
Can old URLs all redirect to the homepage?
Not as an automatic rule. A genuine successor and the customer task determine a suitable target or removal response.
Is visual approval SEO acceptance?
No. Main content, links, metadata, responses and customer paths need inspection on the actual candidate version.
Does sitemap submission guarantee indexing?
No. It supplies a discovery input; access, Google’s canonical choice and indexing remain separate observations.
Does a recovery plan prevent traffic loss?
It does not guarantee that outcome. An executable plan and repeat checks help manage risk without a fixed recovery promise.
Does this source accept publication or lead delivery?
No. Production, native review, account and capacity, server acknowledgement, download and CRM delivery need separate acceptance.
What does an incomplete item mean?
An incomplete item alone is not proof of a failure. Review evidence and scope before choosing a remedy.
What should I prepare for an initial discussion?
Share your current system, priority problem, sample outputs and decision-making team. These help define an initial scope without adding unnecessary work.
How are costs and timing determined?
Assess scope, existing data, integrations, review rounds and continuing work together. A binding schedule and price follow initial technical or content assessment.
Can you work with our existing team?
Define a collaboration model with internal and external responsibilities, access and acceptance criteria. The proposal states who owns each task.