Create the preferred URL inventory
A Technical SEO Audit can investigate a broader site; this worksheet follows the preferred address through specific evidence. Begin with a small, named URL set, the site origin and the responsible owner. In the educational example, a product address and a parameter variant may show equivalent material. Record that as a question to verify, not permission to remove either address.
Primary addresses and actual variants
List the protocol, host, language path, trailing slash and parameters actually observed. A similar-looking address may support another task. Record the page purpose and proposed preferred version before selecting technical treatment. Keep an export of the current state so later corrections can be compared with a dated baseline rather than memory.
Page purpose and indexing decision
Decide whether the page should be publicly indexable, intentionally excluded or still unresolved. A customer account page does not become a search destination merely because it returns successfully. Separate content ownership from the indexing decision. An unresolved field remains open, with an evidence owner; it is not automatically a defect or approval.
| Record | Required evidence | Decision boundary |
|---|---|---|
| Primary address | Origin and page purpose | No invented live URL |
| Variant | Content comparison and response | Similarity alone is insufficient |
| Private page | Documented access purpose | Preserve intentional restriction |
Read HTTP, robots and canonical signals together
The Technical SEO Audit context helps separate access from canonical preference. Capture the requested address, each redirect step, final response and declared canonical. Read robots instructions alongside those observations. A successful response does not establish Google indexing, and a canonical declaration does not establish the address Google will select. Keep these results in different fields.
Redirect destinations and status codes
Check that a proposed destination actually answers the original task and is accessible under the intended conditions. A redirecting address and its final destination are separate rows. A login-dependent browser result may not represent public access. Record loops, broken targets and contradictory exclusions without changing the live site merely to make the worksheet look complete.
Declared and Google-selected canonical
Store the live declaration separately from authorized Search Console evidence of Google’s selected canonical. Include the inspection date and version. Crawl restrictions and noindex serve different purposes, and a crawler needs access to read the instruction. Do not remove deliberate private or preview controls. Assign conflicts to the technical and content owners for a bounded decision.
Signal evidence
Declared preference
Keep the emitted declaration and target separate from Google’s observed choice.
Observed selection
Use dated authorized property evidence; missing access stays unknown.
Validate the sitemap index and each child
SEO Migration Services addresses changes across a site; this check verifies the sitemap index and every listed child file. Record GET response, content type, XML parsing and the addresses returned. Compare HEAD separately rather than treating it as proof that the body is usable. One working index does not establish that every child or destination works.
XML responses and preferred URLs
Trace the index to its children and record which file contains each sampled preferred URL. Check generation origin, encoding and URL construction. Use current official limits when designing automated file-size and URL-count controls. A template that generates an XML-looking document has not passed parsing or endpoint tests until those observations are recorded.
Routes that should be excluded
List redirected, broken or intentionally excluded entries as findings for sitemap membership, not instructions to delete customer pages. Investigate unexpected variants before removing them from the file. The desired set contains approved preferred indexable addresses. A successful submission supports discovery; it does not guarantee crawling, indexing, ranking or a rich result.
Record four pieces of evidence for each URL
In a technical SEO review, record the requested URL, its final destination after redirects, the declared canonical and the sitemap entry in one row from the same check. If an old URL redirects to its replacement but the sitemap still lists the old address, assign someone to trace how that entry was generated. When you inspect Google’s selected canonical, record it separately with the observation date.
Align language equivalents and internal links
International SEO Services helps define genuine language equivalents. Map each actual counterpart and its publication state before including reciprocal alternates or sitemap entries. A planning route is not an approved public page. Preserve the existing owner in each language; do not send every translated version to one Turkish canonical merely because the subject is shared.
TR, EN and approved FA versions
Compare the reader task and substantive content before calling two pages equivalents. Check return references and each approved version’s own preferred address. An untranslated placeholder or unrelated guide is not a language counterpart. Keep withheld Persian destinations as planning data until the required route and editorial approvals exist; this worksheet does not activate them.
Navigation and contextual links
Inspect menus and contextual links against the selected same-language address. Record obsolete links and their actual target before proposing a correction. Align discovery signals with the content decision instead of changing only the canonical field. When HTML rendering affects the evidence, assign that observation to its appropriate implementation review rather than infer it from a source file.
Review dates and generation rules
A Technical SEO Audit should also explain where dates originate. Record real initial publication and meaningful modification events, with their source history. A deployment or sitemap regeneration alone is not a content update. Avoid random backdating or refreshing every lastmod value on each build. Date consistency cannot be established by making several fields repeat an invented timestamp.
Real publication and meaningful updates
Identify the changed page material and the owner who accepted it. Preserve the original publication date when it remains factual. If a date is unknown, mark it unknown instead of reconstructing a convenient schedule. Keep infrastructure activity, editorial work and search observation as separate events so the review can explain why a modification date changed.
Lastmod, visible dates and schema
Compare the page’s visible date, source record, relevant structured data and sitemap lastmod. Resolve disagreement at the responsible source, then test its generated output. Do not use priority or changefreq as a ranking promise. Google’s documented treatment of these fields does not remove the need for accurate URLs or meaningful change provenance.
| Field | Evidence | Acceptance |
|---|---|---|
| Publication | Original editorial record | No fabricated first release |
| Modification | Meaningful change and owner | Not every build |
| Sitemap lastmod | Page update source | Consistent significant-change date |
| Visible/schema date | Actual emitted values | Explained or repaired discrepancy |
Verify fixes and hand over monitoring
The JavaScript SEO Release Checklist covers rendering-specific acceptance. Hand over this URL worksheet with the requested change, owner and test conditions, then recheck the affected responses and sitemap outputs after an authorized correction. Technical acceptance and later Google reassessment remain separate. Do not close a finding solely because a sitemap was submitted or a task ticket was marked done.
Live checks and later recrawl evidence
Before any approved implementation, retain the previous generator rule, URL decision and affected content version. Name a rollback owner and the condition that would trigger reversal. An unexpectedly missing useful page or wrong destination needs a clear recovery path. This document plans those checks; it neither modifies redirects nor removes live URLs.
Save the worksheet or inspect a sample
Use the fourteen rows to retain evidence, unresolved questions and the next observation. A limited sample review can begin with a public sitemap or redacted URL record, without customer or account secrets. Worksheet saving, email delivery and CRM acknowledgement are separate facilities requiring implementation. None is implied by reading the checklist or choosing a discussion CTA.
Handover state
Controlled acceptance
Retain response/XML tests, owner and previous version.
Open observation
Search reassessment and delivery facilities need separate evidence.
Checklist evidence worksheet
| Check | Input | Evidence | Owner | Acceptance |
|---|---|---|---|---|
| Record the site origin | Host,protocol,language path | Inventory and request record | Technical owner | Generated address uses the intended origin |
| List actual variants | Slash,parameter,host variants | Dated URL/content comparison | SEO and content | Distinct need is separated from equivalent duplicate |
| Record purpose and preference | Primary user job | Reasoned URL decision record | Content owner | Unrelated pages are not forced to one target |
| Separate access and indexing purpose | Login,robots,noindex | Response,instruction and restriction rationale | Technical and policy owner | Intentional privacy preserved;conflict explicit |
| Check final HTTP destination | Initial URL and redirect steps | Chain and destination response log | Developer | No loop/broken target or explicit open finding |
| Record canonical declaration | Emitted output and version | Declaration,target and test time | Developer and SEO | Accessible target consistent with preference |
| Keep Google selection separate | Authorized property and full URL | Dated URL Inspection or unknown | Authorized account owner | Declaration does not replace observation |
| Fetch sitemap index | Actual index address | GET,body type,XML;HEAD difference | Infrastructure owner | XML parses;HEAD difference explained |
| Verify every child file | Children listed by index | Response/parse evidence per file | Generator owner | No missed child;current limits checked |
| Decide URL membership | Preferred,broken,redirected,excluded URLs | URL-row to XML comparison | SEO and content owner | File correction is not page-deletion approval |
| Map genuine language counterparts | Content equivalence and publication state | Reciprocal mapping and access | Localization owner | Only approved equivalents;FA hold preserved |
| Check navigation and text targets | Incoming-link inventory | Same-language target/old-reference log | Content and development | Links match decision;old paths explained |
| Verify truthful date provenance | Publication,modification,lastmod | Content history and emitted dates | Editor and generator owner | No invented dates;build is not content update |
| Hand over tests,rollback and monitoring | Version,change owner,trigger | Before/after record and open tasks | Release and rollback owner | Technical acceptance is not a search guarantee |
Frequently asked questions
Does canonical force Google to select an address?
No. It expresses preference. Record Google’s observed choice separately and examine conflicting access, links and sitemap signals.
Does sitemap inclusion ensure indexing?
No. Discovery, page eligibility and Google’s indexing decision are separate. A parsed file or successful submission is not an index certificate.
Should all language pages use the Turkish canonical?
Do not collapse genuine language counterparts that way. Review each owner and map only accessible, approved equivalents; withheld FA plans are not public alternates.
Can lastmod change on every build?
Only a genuine meaningful page change supports the relevant update date. Preserve original publication evidence and distinguish infrastructure activity from editorial work.
Can this review start without Search Console access?
Public HTTP, HTML and XML can support part of the worksheet. Google-selected canonical and account submission evidence require an authorized owner; unknown values remain unknown.
Does this checklist delete pages or deliver an email?
No. It records decisions and future acceptance conditions. Live changes need authorization and testing; resource, lead and CRM delivery remain separate implemented gates.