Google Search Console
First-party search performance, indexing evidence, selected canonicals and supported generative-AI reporting.
SEO and AI visibility
Help buyers find your expertise in search and AI answers
See what may be holding back your visibility in search and AI answers, with a report that connects findings to specific pages, priorities and actions.
- Find the problems and opportunities that matter
- Know what to change on which pages
- Check progress against clear criteria
What you receive
- A scoped review of crawling, indexing and priority commercial pages.
- A prioritised list of technical and content corrections.
- A content plan based on buyer questions and evidence you can substantiate.
- Page-structure and internal-linking recommendations connecting services to useful resources.
- A measurement framework for relevant search visits, qualified enquiries and observed AI citations.
The problem to solve
Make your priority pages easier to find, understand and use. We review technical access and buyer-facing content, then prioritise improvements to your service pages and supporting resources. You receive a plan tied to useful enquiries and observable search performance, with AI-answer observations recorded separately.
A page can be published yet difficult to crawl, or attract visits without answering a buyer's questions. Evidence may be buried, internal links weak and service descriptions unclear. Adding AI terminology does not resolve those gaps. The work starts with access, relevance and a credible answer.
See what is holding visibility back and what to do next
Every recommendation should answer four questions: what is the evidence, why does it matter, what changes, and how will we verify it?
Start with the decisions that matter
A short priority list should explain what to fix first, what to investigate and what can wait.
Explore the checks and examples
Data in the report
The opening summary connects findings to your business goals and shows the affected pages, evidence source, confidence, likely consequence and dependencies. Missing access or incomplete data is visible rather than disguised as a score.
Next step
Agree the outcome you care about, then review the proposed order of work. A blocked buying page, an unclear product offer and an unverified backlink concern need different responses.
How we verify it
Each priority has supporting evidence, an accountable owner role, a realistic effort estimate and a completion check. Estimated effort and expected benefit remain separate from measured results.
Evidence and priorities, not an arbitrary overall score.
Data in the report
- Priority rationale
- Owner role
- Estimated effort
- Prerequisite
- Acceptance criterion
- Evidence date
What we check
- Business questions, target audiences, markets and the agreed audit scope.
- Evidence quality, affected templates and the distinction between a blocker and an improvement.
- Priorities based on relevance, risk, effort and dependencies rather than a composite score.
- A decision-ready summary linked to the detailed page-level evidence.
The opening summary connects findings to your business goals and shows the affected pages, evidence source, confidence, likely consequence and dependencies. Missing access or incomplete data is visible rather than disguised as a score.
Next step
Agree the outcome you care about, then review the proposed order of work. A blocked buying page, an unclear product offer and an unverified backlink concern need different responses.
How we verify it
Each priority has supporting evidence, an accountable owner role, a realistic effort estimate and a completion check. Estimated effort and expected benefit remain separate from measured results.
Official guidanceGoogle Search Central
Connect search demand to the page that can answer it
The keyword map should tell you which page to improve, create, keep or investigate.
Explore the checks and examples
One reader task, a clear page destination
- Understand
Guide
- Compare
Guide
- Buy
Category page
- Resolve a concern
Guide
Data in the report
- Query or topic
- Reader intent
- Target page
- Country or market
- Device
- Estimated search volume
- Observed or estimated position
- Search clicks
- Impressions
What we check
- Queries and topics grouped by the task: understand, compare, buy or resolve a concern.
- Existing landing pages compared with the intent they actually serve.
- First-party performance distinguished from third-party volume and position estimates.
- Potential gaps and overlaps, with a page-level decision and the evidence still needed.
The report can combine real query and page performance with topic research and clearly labelled third-party demand estimates. It records the reader’s task, market, device, relevant landing page and any competing page. A missing query row is not evidence of no demand.
Next step
Choose a primary destination for each important intent. Validate a content gap before commissioning a new page; investigate overlapping destinations before deciding to merge them.
How we verify it
Each proposed destination is backed by a defined reader task and a recorded source. Query-to-page overlap is a diagnostic lead, not proof that pages are harming one another’s rankings.
Official guidanceGoogle Search Central
Make important pages easy to reach
The link map should reveal missing routes and show where a useful connection belongs.
- HomepageCategory page
- HomepageGuide
- Category pageProduct A
- Category pageProduct B
- GuideProduct A
- GuidePage to investigate
Example existing linkExample proposed link
View as a table
| Source | Destination | Agreed scope |
|---|---|---|
| Homepage | Category page | Example existing link |
| Homepage | Guide | Example existing link |
| Category page | Product A | Example existing link |
| Category page | Product B | Example existing link |
| Guide | Product A | Example existing link |
| Guide | Page to investigate | Example proposed link |
Explore the checks and examples
Data in the report
- Source page
- Link anchor
- Target page
- Destination status
- Incoming page links
- Crawl depth
What we check
- Crawlable links, descriptive anchors and the destination’s final response.
- Incoming links and navigation depth for the pages that matter to the business.
- Pages missing from the observed link graph, cross-checked against known URLs.
- Contextual paths from useful guides to relevant products, services and next steps.
A crawl-based link table records the source page, destination, anchor text, response status and incoming links. Depth and disconnected-page findings are interpreted against the defined crawl scope and the full known inventory.
Next step
Connect related pages where the next step is useful to a reader. Fix broken or unnecessary redirecting links, and check the purpose of an isolated page before adding it everywhere.
How we verify it
Re-crawl the retained pages and verify the intended navigation paths. A sitemap entry, canonical or language annotation does not replace a crawlable HTML link.
Official guidanceGoogle Search Central
Make titles and summaries match the page’s purpose
A useful before-and-after explains the improvement, not just a character count.
Explain the page more precisely
Before
Discover our range of products.
Proposed after
Compare ceramic coffee drippers by capacity, filter shape and care needs. Find the format that suits your daily brew.
The proposed version names the product and useful selection criteria. Use it only when those criteria reflect the actual page; no unverified delivery claim or ranking promise is added.
Explore the checks and examples
Data in the report
- Target page
- Page title
- Meta description
- Search clicks
- Impressions
- Click-through rate
What we check
- Missing, repeated or misleading titles and descriptions.
- Consistency between the search message, heading and actual page content.
- Representative before-and-after examples with a clear editorial rationale.
- Real performance context when available, without a universal title-length rule.
The report lists titles, descriptions and relevant headings by URL, flags missing or repeated text, and compares the wording with the page’s intent. Where available, actual search appearances, clicks and impressions inform the review.
Next step
Rewrite the pages where the message is unclear or misleading. Keep the wording specific and truthful; resolve duplicated page purposes before polishing duplicate titles.
How we verify it
Check the rendered metadata and the page facts it describes. Search engines may choose different titles or snippets, so the example is not a promise of exact display or higher click-through rates.
Official guidanceGoogle Search Central
Find the exact technical obstacle before prescribing a fix
Crawl access, indexing instructions, canonical signals and language alternatives are different checks.
Explore the checks and examples
Find which instruction or signal needs attention
- Can the page be fetched?
- What do its index rules say?
- Which URL is preferred?
- Are language equivalents aligned?
- What did the engine actually select?
Use field benchmarks without confusing them with your results
LCP≤ 2.5 sGood field threshold
INP≤ 200 msGood field threshold
CLS≤ 0.1Good field threshold
A page can return 200 and still request noindex
On a page intended for search, an unintended noindex instruction is a specific problem to verify. The simplified before-and-after shows that instruction changing; it is not a recommendation to make every URL indexable.
Before
<meta name="robots" content="noindex">Proposed after
<meta name="robots" content="index, follow">First confirm the intended policy and check both the HTML and response headers. The default already allows indexing when no restrictive instruction applies, so an explicit index, follow tag is not always necessary.
How we verify it: Re-fetch the page, check all applicable directives and crawl access, then use authorized search-console evidence to inspect actual indexing. Removing noindex does not guarantee inclusion.
Data in the report
- HTTP response
- Crawl and indexing rules
- Canonical URL
- Language alternatives
- Sitemap inclusion
- Rendered content
- Largest Contentful Paint
- Interaction to Next Paint
- Cumulative Layout Shift
- Field period and coverage
What we check
- HTTP status, redirects, robots rules, meta robots and X-Robots-Tag instructions.
- Canonical consistency, sitemap coverage and reciprocal language alternatives when applicable.
- Important content and links available in the rendered page, not just assumed from a template.
- LCP, INP and CLS measured with clear field and lab sources and missing-data states.
The technical evidence table records HTTP responses, redirects, crawl permissions, index directives, rendered content, canonicals, sitemaps and multilingual annotations where relevant. Performance evidence states whether it comes from real users or a lab test, with device, date window and URL coverage.
Next step
Correct conflicting signals while preserving deliberate exclusions. Diagnose slow loading and layout movement on representative templates, then distinguish implementation checks from search-engine outcomes.
How we verify it
Verify the changed responses and rendered HTML, then use authorized search-console evidence for actual indexing and selected canonicals. A successful crawl does not prove indexing; a single lab run does not prove a field Core Web Vitals pass.
Show which questions the site answers well and which it leaves open
A content recommendation should make a page more useful, not merely longer.
Explore the checks and examples
| Reader intent | Guide | Category page | Product A |
|---|---|---|---|
| Understand | Example coverage | Example gap | To verify |
| Compare | To verify | Example coverage | Example gap |
| Buy | Example gap | Example coverage | Example coverage |
| Resolve a concern | Example coverage | To verify | To verify |
The sample matrix shows how reader tasks can be compared with page types. Its cells are illustrative; a real report links each gap to page-level evidence.
Data in the report
- Target page
- Page purpose
- Unanswered question or content gap
- Author or reviewer
- Evidence of possible overlap
What we check
- Coverage of discovery, comparison, purchase and reassurance questions.
- Accurate specifications, service details, limitations and commercial policies.
- Original evidence, useful sources and truthful author or reviewer information.
- Possible content overlap and conversion paths, with a reasoned keep, improve, merge or create decision.
The content matrix compares reader tasks with existing pages, factual completeness, supporting evidence and the route to a meaningful next step. It records potential overlap, authorship or review gaps and missing information that could prevent a confident decision.
Next step
Prioritize the facts and explanations people need to choose. Use real product or service records, identify genuine contributors and decide whether overlapping pages need a clearer distinction or careful consolidation.
How we verify it
Verify claims against their sources and compare the revised page with its intended task. A merge decision needs page-level evidence, a destination plan and checks for links and redirects.
Official guidanceGoogle Search Central
Judge external links in context
The report should distinguish a useful referral, a broken route and a concern that still needs evidence.
Explore the checks and examples
- Referring-domain and referring-page context, topical fit and editorial placement.
- Anchor text, link attributes, destination response and redirected or lost links.
- Dataset coverage, observation dates and material evidence gaps.
- Potential link manipulation and manual-action evidence before any exceptional disavow advice.
Data in the report
- Referring domain
- Referring page
- Link placement
- Link attributes
- Link anchor
- Final destination
- First observed date
What we check
- Referring-domain and referring-page context, topical fit and editorial placement.
- Anchor text, link attributes, destination response and redirected or lost links.
- Dataset coverage, observation dates and material evidence gaps.
- Potential link manipulation and manual-action evidence before any exceptional disavow advice.
A backlink review separates referring domains from individual links and records the referring page, context, anchor, attributes, destination and retrieval date. Third-party datasets provide a view of the web, not an exhaustive inventory.
Next step
Investigate important lost or broken referrals, mismatched destinations and evidence of manipulative link patterns. Preserve useful routes and avoid treating an unfamiliar domain as harmful by default.
How we verify it
Record the evidence behind each recommendation and the limits of the sample. A vendor toxicity label alone does not justify disavowing links; exceptional action needs the appropriate supporting evidence.
Official guidanceGoogle Search CentralGoogle Search Console
Check whether accurate information can be found, understood and cited
AI visibility starts with verifiable facts and platform-specific eligibility, not a guaranteed citation.
Explore the checks and examples
- Useful visible facts, clear sourcing, meaningful structure and supported commercial details.
- Search eligibility and the current generative-AI inclusion settings, including inherited controls.
- Search crawler access separated from model-training policy and user-triggered retrieval.
- Provider-specific impressions or citations, with scope, date and missing-data limits.
Make machine-readable information match the visible facts
This short Product excerpt demonstrates the relationship between a product name on the page and in the markup. It is deliberately incomplete and contains a placeholder, not production-ready product data.
{
"@context": "https://schema.org",
"@type": "Product",
"name": "[…]"
}Populate only verified facts that are also visible to the reader. Do not invent reviews, availability, prices or certifications to satisfy a validator.
How we verify it: Check the complete implementation against the relevant structured-data requirements and the product record. Valid syntax alone does not guarantee a rich result or an AI citation.
Data in the report
- Crawler policy
- AI inclusion control
- Visible verified facts
- Structured data
- Provider-reported citations
- Evidence date
What we check
- Useful visible facts, clear sourcing, meaningful structure and supported commercial details.
- Search eligibility and the current generative-AI inclusion settings, including inherited controls.
- Search crawler access separated from model-training policy and user-triggered retrieval.
- Provider-specific impressions or citations, with scope, date and missing-data limits.
The report checks accessible page content, search and snippet eligibility, applicable inclusion controls, crawler policies and structured data that reflects visible facts. It distinguishes provider evidence from limited manual observations and records what could not be verified.
Next step
Fix the underlying content and search issues first. Review search discovery and model-training preferences separately, and measure supported AI impressions or citations without confusing them with visits or sales.
How we verify it
Verify the current provider controls and actual account settings where access is authorized. No special AI schema or llms.txt ranking fix is assumed. Eligible content can still receive no citation.
Turn findings into work someone can own
Each next step needs a reason, a dependency and a definition of done.
Explore the checks and examples
| Priority | Lower estimated effort | Higher estimated effort |
|---|---|---|
| Blocks the intended outcome | Correct a verified blocker | Scope the blocking project |
| Does not block eligibility | Plan a useful improvement | Validate the value first |
This matrix is a starting point. Evidence quality, business value, risk and dependencies also affect the final order; it is not an ROI forecast.
Data in the report
- Priority rationale
- Owner role
- Estimated effort
- Prerequisite
- Acceptance criterion
What we check
- A clear distinction between verified fixes, investigations and optional improvements.
- Priority informed by user value, business relevance, risk and dependencies.
- An accountable owner role, estimated effort and the access or approval required.
- A testable completion criterion and a planned observation after implementation.
The action plan brings together the evidence, proposed change, priority, owner role, estimated effort, prerequisites and acceptance check. The matrix supports a conversation about blockers and complexity; it does not forecast return on investment.
Next step
Confirm the order with the people who will implement the changes. Establish the baseline, clear verified blockers, settle architecture decisions and then complete the dependent content and template work.
How we verify it
Close an action when its stated acceptance criteria are met. Measure subsequent outcomes separately; finishing a fix is not proof that it caused a change in traffic or revenue.
Official guidanceGoogle Search Central
Keep search visibility, AI exposure and business outcomes separate
A good baseline makes later decisions possible; missing data must remain unknown.
Explore the checks and examples
Follow the path from eligibility to a useful outcome
- Visibility: compare available impressions and citations, keeping them separate from visits.
- Visits: identify traffic sources and landing pages, noting attribution gaps.
- Enquiries: define relevant enquiries and check that they are recorded reliably.
- Clients: link enquiries to confirmed sales, keeping potential deals out of reported revenue.
Data in the report
- Source page
- Impressions
- Search clicks
- Provider-reported citations
- Identifiable referral sessions
- Defined conversion event
- Evidence date
What we check
- Baselines for each source, date windows, filters and reliable conversion definitions.
- Google search and generative-AI reporting kept distinct where their metrics differ.
- Bing citations and identifiable assistant referrals separated from conversions.
- Comparable periods, change annotations, other factors and an explicit next decision.
The measurement plan identifies the source, period, filters, dimensions and limitations for each metric. It separates indexing, search clicks and impressions, AI impressions or citations, identifiable assistant referrals and meaningful conversion events.
Next step
Record the baseline before changes and annotate releases. Compare suitable periods and check seasonality, changes in demand and other interventions before attributing an outcome to the work.
How we verify it
Use only metrics supported by the source. Google’s dedicated generative-AI report documents impressions, not separate AI clicks or CTR. Bing citations are not a universal AI ranking. A repeated prompt sample remains a limited observation.
Tools and data sources
Sources and tools are selected for the agreed scope and available authorized access. Logos do not imply a partnership or that every tool is included in every audit.
Google Analytics 4
Landing-page behaviour, identifiable referral traffic and defined conversion events to connect visits with business actions.
Ahrefs
External-link research, referring-page context, topic research and clearly labelled search estimates.
Semrush
Keyword and competitor research, estimated search visibility and supporting audit evidence.
Screaming Frog SEO Spider
A scoped crawl of URLs, links, titles, directives, canonicals and other technical page evidence.
PageSpeed Insights
Available field performance evidence and lab diagnostics to investigate loading, responsiveness and layout stability.
Bing Webmaster Tools
Bing search evidence and supported AI Performance data, including citations and available topic or intent views.
Rich Results Test
Check the supported structured-data implementation and diagnose feature-specific errors.
Semrush is a trademark or registered trademark of Semrush Inc. in the U.S. and other countries. All rights reserved. All other trademarks are the property of their respective owners.
Method and limitations
A useful audit connects search demand, site structure, content and technical evidence to clear decisions. Explore the checks, data tables and visual analyses the report can contain, then see how the findings become an ordered action plan.
This is a report blueprint, not an audit result or a client case study. Diagrams and before-and-after examples are illustrative. The scope and available data sources are agreed for your site.
Evidence and priorities, not an arbitrary overall score.
- Google Search Console
- Requires authorized access to the relevant verified property. Report availability, scope and missing data are recorded.
- Google Analytics 4
- Requires authorized property access, appropriate consent and a review of tracking quality and attribution limits.
- Ahrefs
- Used where the agreed scope and available access support it. Coverage is not exhaustive; a paid licence is not implied.
- Semrush
- Selected where relevant to the scope and available access. Estimates are distinguished from first-party results; no subscription is assumed.
- Screaming Frog SEO Spider
- Requires an agreed crawl scope and permitted access. Tool configuration and limits are recorded; private or blocked areas require appropriate permission.
- PageSpeed Insights
- Field coverage may be insufficient or available only at origin level. Lab diagnostics are reported separately from real-user results.
- Bing Webmaster Tools
- Requires authorized property access and feature availability. Supported surfaces, preview features and estimated values are disclosed.
- Rich Results Test
- Validation is combined with a check against the visible page. A passing test does not guarantee display or citation.
How the engagement works
- 01
Check discovery barriers
We examine technical access and available data before assuming content is the only reason a page is underperforming.
- 02
Strengthen the answer
We identify what buyers need to understand, compare and trust the offer, then map missing evidence and supporting links.
- 03
Sequence the improvements
We prioritise changes and define useful observations, keeping search traffic and AI citations distinct.
What we need from you
- Your priority services, markets and common prospect questions.
- Available search and analytics access, with existing performance records.
- Publishable expertise, product documentation and substantiated evidence.
Scope and responsibilities
Pages, markets and available data are agreed upfront. Implementing fixes is scoped separately. Rankings, indexation timing and inclusion in AI answers cannot be guaranteed.
How we assess progress
Where access is available, track indexation, relevant impressions, clicks and qualified enquiries. Record AI-answer observations with the query, date and context. A single citation is an observation, not proof of stable coverage or an attributable increase in revenue.
References behind the approach
Google says established SEO fundamentals apply to its AI features. No special AI file guarantees inclusion. Google Search Central · AI features and your website
Compare engagementsQuestions before you start
Can you guarantee a mention in ChatGPT or Google's AI answers?
No. We can improve access, clarity and evidence, then observe the results. Search and AI systems decide which sources to show, and their responses can vary.
Do we need to create more pages?
Not always. Improving an existing page for the same buyer need may be more useful than creating another. The review looks for gaps as well as duplication.
Review our search visibility
Prices are in USD, excluding applicable taxes. Your written proposal confirms scope, schedule, access requirements and working language. An enquiry does not commit you to a purchase.