Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Playbooks

Technical SEO Maintenance Playbook

An operating playbook for keeping a site technically healthy for search: release checks, Search Console triage, crawl and speed reviews, audits and a fix queue.
Playbookintermediate11 min readUpdated

What this playbook covers and when to run it

The playbook covers the standing work that keeps a site findable, crawlable and indexable between projects: a baseline and requirements engineering can build from, checks around every release, the weekly reading of Search Console, monthly reviews of crawling, indexing and page experience, rules for pages that end or change, the quarterly audit and log review, a single queue of fixes, and the point at which a change is too big for the routine. It does not teach any of those methods. The technical SEO hub explains what technical SEO covers and the order of priorities that usually pays, and each phase below hands off to the guide, procedure, checklist or audit that owns the method.

It sits beside the SEO content operations playbook, which runs research, briefs, publishing and refreshes; this one keeps the site those pages live on healthy. The reporting cadence playbook carries organic results to the people who read them, and the measurement governance playbook keeps the analytics side of those numbers honest.

Run it when any of these is true: a release has broken indexing or search traffic before; Search Console sends alerts that nobody reads; the technical audit happens once and its findings sit unshipped; engineering hears about SEO only when something is already broken; or a large catalogue, listings section or archive keeps generating URLs nobody planned. Like every playbook in the playbooks library, it is a starting shape: a small site on a hosted platform may run it in an hour a week, while a large site with daily releases will spread it across SEO, engineering and release management.

Roles and responsibilities

  • SEO lead. Owns the list of SEO-critical settings, the golden URLs and their expected values, the Search Console triage, the audit calendar and the fix queue. Decides what each finding is worth and argues its place in the queue.
  • Engineering owner. Owns the pipeline checks, the monitoring job, rollbacks and the fixes to templates, server configuration, redirects and page performance. Agrees with the SEO lead how each fix is ranked against other work.
  • Release manager. Confirms that the pre-release comparison and the post-release checks ran for every release that touches templates, routing, headers or the robots file, and that the change log entry exists.
  • Content editors. Apply the rules for pages that expire, change or are generated by the content system, and route content and metadata fixes through the normal publishing procedure.
  • Analytics owner. Keeps release dates and incidents annotated in analytics and the weekly reports, so a change in traffic can be lined up against a change in the site.

The release checks guide already splits ownership this way: engineering owns the pipeline, monitoring and rollback, the SEO lead owns the critical settings and alerts, and the release manager confirms the checks ran. The playbook adds the calendar around those roles. In a small team one person may hold the SEO and release roles; the engineering side still needs a named owner, because a finding with no engineer attached never ships.

Cadence, inputs and outputs

The rhythm below is a suggested starting point. Adjust it to how often the site releases, how large it is and how quickly its content changes; a site that ships daily needs its release checks automated, while a small site may fold the monthly review into the weekly one.

CadenceOwnerInputsOutputs
Per releaseEngineering owner, release managerGolden URLs, expected values, the staging buildA passed comparison, post-release checks, a change log entry, or a rollback
WeeklySEO leadSearch Console reports and alerts, monitoring alertsEach issue closed, ticketed with an owner or escalated
MonthlySEO lead, analytics ownerPage Indexing report trend, crawl stats or log summary, field page experience data, the last reviewA short review against last month and new items for the fix queue
QuarterlySEO lead, engineering ownerA fresh crawl, Search Console exports, server logs, the baseline scoreA rescored audit, a log review and a reranked fix queue
AnnuallySEO lead, engineering owner, product or platform ownerThe year's audits, incidents and queue historyDecisions on templates, architecture and platform, and any migration handed to its own plan

Each cadence feeds the next. Release checks and weekly triage catch regressions while they are small; the monthly review shows patterns individual alerts hide; the quarterly audit judges the pattern against the baseline; the yearly review decides what the routine cannot fix.

The playbook in phases

  1. Set the baseline and the owners

    Start with a full picture. The technical SEO audit tutorial gathers crawl, Search Console and log data and runs the checks in order, the technical SEO audit checklist holds the pass-or-fail list, and the technical SEO audit scores crawlability, indexation, canonicals, architecture, structured data, page speed, mobile and international set-up so the next run has something to compare with. The first score is the baseline. Then agree how findings reach engineering: the tutorial on SEO requirements for developers writes needs as acceptance criteria that are tested before a feature is built, which is the form every later fix and template takes. Name the SEO lead, the engineering owner and the release manager before the phase closes.

  2. Check every release

    The release checks guide keeps staging out of search with access control rather than robots rules, compares the SEO-critical settings on a set of golden URLs before each release, repeats the checks in production afterwards and adds monitoring that alerts on change. It explains, from Google's robots.txt documentation, why a robots rule manages crawling rather than keeping a page out of results. Its post-release list includes the robots file, directives, the canonical, status codes, redirects and the XML Sitemap; the SEO page checker reads the head and directive basics of a single URL for spot checks. Single pages published or changed by editors follow the procedure for publishing a web page. Every release that touches templates, routing or headers gets an entry in the change log both teams read.

  3. Triage Search Console weekly

    The Search Console error triage procedure runs this phase: a weekly review of the indexing, experience and enhancement reports for new or growing issues, the same-day handling of alert emails, sizing the impact, finding the cause, assigning the fix and validating it, with manual actions and security issues escalated at once. The Search Console tutorial explains what each report actually says and sets the weekly and monthly routine the playbook adopts. Each issue leaves triage closed with a note, ticketed with an owner, or escalated. The SEO troubleshooting collection holds the diagnoses for the commonest symptoms.

  4. Review crawling, indexing and page experience monthly

    Once a month, as a starting rhythm, read the trend rather than the latest alert. The crawling and indexing guide explains how discovery, crawling, rendering and indexing work and what each control does, and is the reference when the Page Indexing report shows a reason growing. Read field page experience data by template with the Core Web Vitals guide, which draws on Google Search Central's explanation of how Core Web Vitals relate to search results, and the page experience guide for the questions beyond them. Check that generated pages still follow the index rules for archive pages, and that rendering has not changed where the site depends on scripts, using the JavaScript SEO guide. Write the review as a short comparison with last month and add new findings to the fix queue.

  5. Handle pages that expire and change

    Some pages have an end date, some multiply through parameters and some carry markup that must change with the content. The guide to expired pages sets a rule per state for offers, events, listings and jobs, applied by the content system rather than by memory. The URL parameter audit keeps a register of every parameter with a rule for crawling, indexing and links, and makes the register part of release review. The structured data guide keeps markup accurate as the page it describes changes. Content editors own applying these rules; the SEO lead owns the rules themselves and checks their effect in the monthly review.

  6. Run the quarterly audit and log review

    Once a quarter, as a starting rhythm, rerun the technical SEO audit against the baseline, using the checklist and rescoring the audit scorecard, so the band shows whether the site is improving. Read the server's own record of crawling with the log file analysis tutorial: crawler requests by template and status code, where crawl goes to parameters, archives or redirects, and which important pages crawlers neglect. The log guide suggests a lighter monthly summary between audits, which fits the monthly review. The output is a rescored audit, a short log note and a reranked fix queue.

  7. Prioritise and ship fixes

    Findings from every phase go into one queue, written as the requirements tutorial describes: the finding, the affected template, sample URLs, the expected behaviour and how to verify it. The SEO lead and the engineering owner rank the queue together, using the hub's priority order and an impact-effort matrix where several fixes compete. Fixes are made at the template level wherever possible. After a fix ships, the tutorial on measuring SEO changes judges it against comparable pages left unchanged over matched windows, so the team learns which kinds of fix pay on this site.

  8. Hand large changes to a migration plan

    A new domain, a platform or framework change, a new URL structure, a redesign or a large consolidation is a migration, not maintenance. The site migrations guide defines what counts and runs the project from baseline and redirect map to post-launch monitoring, and the website migration checklist owns the before-and-during steps. The site architecture guide informs any structural change. The routine pauses for the affected sections while the migration runs, and resumes with a fresh audit afterwards, as the audit checklist advises.

The fix queue: fix, queue, defer or escalate

Every finding, from an alert, a review, an audit or a log read, leaves with one of four decisions. Writing the decision down is what turns findings into maintenance rather than a growing list.

DecisionWhen it appliesWho decidesWhat happens next
Fix nowImportant pages blocked from indexing, wrong canonicals or server errors on key templates, a release regressionSEO lead raises; engineering owner rolls back or hotfixesReverse first, diagnose second, verify in production and log the incident
QueueA real problem that is not blocking, such as duplicates, buried pages or a slow templateSEO lead and engineering owner togetherA ticket with acceptance criteria, ranked in the queue, with a date to measure the result
Defer with a reasonLow impact, high effort, or waiting on a planned template or platform changeSEO lead, recorded in the queueThe reason and a review date, so the item returns at the next audit rather than vanishing
Escalate to a migrationThe fix changes domains, platforms, URL structure or many templates at onceSEO lead with the product or platform ownerThe item leaves the queue for a migration plan with its own owner

The release checks guide sets the rule for the first row: when a release adds a noindex, a disallow or a wrong canonical, reverse the change first and diagnose second. Everything else is ranked on what it blocks, not on how easy it is. A queue item deferred audit after audit is either re-argued with fresh evidence or closed with a note, because a queue nobody believes stops being read.

Illustrative example — not a real result

A recipe publisher keeps its site healthy

Ixlowen Recipes, an invented recipe publisher with a small SEO team and fortnightly template releases, starts the playbook after a release ships a noindex header on its dessert section and 1,200 recipes drop out of the index over the following days. The baseline audit scores amber. Engineering adds a golden URL list of 18 pages and a pipeline check that fails any production build with a blanket disallow or a noindex on an indexable template; in the next quarter the check stops 2 releases. Weekly Search Console triage turns 9 alerts into 4 tickets and closes the rest with notes. The first quarterly log review finds that tag archives and sort parameters take about 40% of crawler requests while new recipes wait; the index rules for archives and a parameter register become the top two items in the fix queue, ranked with engineering. Six months on, the audit scores green on indexation, and every deferred item carries a reason and a date.

Measuring the playbook

The playbook works when technical problems are found by the routine rather than by a traffic drop. Operating signals show whether the rhythm holds: whether every release that touches templates has its checks recorded, whether weekly triage happens and every issue leaves with a decision, whether the monthly review is written, and whether the quarterly audit and log review run on schedule.

Outcome signals show whether the site is getting healthier: how long regressions stay live before they are found, how many are caught before release rather than after, the audit band from one quarter to the next, the trend of growing reasons in the Page Indexing report, the share of crawler requests reaching pages the site wants indexed, and how many queued fixes ship and are measured. Report them in the team's existing monthly and quarterly packs, beside organic results, rather than in a separate technical report nobody outside SEO reads.

Common mistakes

  • Treating the audit as the programme. A yearly audit without release checks and weekly triage finds the same problems again. Run the routine between audits.
  • Release checks that depend on memory. Manual checks catch what someone remembers. Automate the comparison in the pipeline and alert on change.
  • Protecting staging with robots rules. A robots file manages crawling and travels with the build. Use access control on non-production hosts.
  • Alerts with no owner. Search Console emails sent to a shared inbox are read by nobody. Name the person who triages and the day they do it.
  • A fix list engineering never agreed. Findings ranked only by the SEO team wait behind everything else. Rank the queue together and write each item as a testable requirement.
  • Fixing page by page. Most technical problems live in templates. Fix the template, and measure the result against pages left unchanged.
  • Running a migration as maintenance. A domain, platform or URL change needs its own plan, redirect map and monitoring. Hand it over rather than squeezing it into the queue.

Frequently asked questions

What does ongoing technical SEO maintenance involve?

Checking every release before and after it ships, reading Search Console reports and alerts each week, reviewing crawling, indexing and page experience each month, rerunning the technical audit and a server log review each quarter, and keeping one ranked queue of fixes agreed with engineering. Large changes such as a new domain, platform or URL structure leave the routine and become a migration project.

Who should own technical SEO on a site?

The SEO lead owns the list of critical settings, the Search Console triage, the audits and the fix queue. Engineering owns the pipeline checks, monitoring, rollbacks and the fixes themselves, and a release manager confirms the checks ran. Content editors apply the rules for pages that expire or change, and an analytics owner keeps the change dates visible in reporting.

How often should a technical SEO audit be run?

This playbook suggests a quarterly audit with a log review, a yearly look at architecture and templates, and an extra run after any migration, redesign or new template. That is a starting rhythm, not a rule: a large site that releases daily may review more often, and a small, stable site can review less, as long as releases are still checked.

Terms used on this page

Tools to try

Frameworks

Where this page is used

Sources

  1. Page Indexing report — Google Search Console Help
  2. Understanding Core Web Vitals and Google search results — Google Search Central
  3. Robots.txt introduction and guide — Google Search Central

Related