A mention is only the beginning of the story

Your agency sends a chart showing more AI mentions. The sales team has not noticed more enquiries. Does that mean the work failed?

You cannot tell from the chart alone. A mention shows that a captured answer named your business. It does not show whether the reader was a suitable buyer, visited the website or contacted you.

Consider a possible journey: someone sees your company in an AI recommendation, searches for your brand later and then submits an enquiry. Your website analytics may record the search visit without recording the earlier recommendation. That gap matters when you discuss where the customer came from.

This is an illustrative journey, not an observed sequence from our clients. A short “How did you hear about us?” question can add useful context, but a person’s answer is still self-reported evidence.

Can you reproduce the question sample?

A useful comparison starts with the same core questions. Changing them every month can improve the percentage without improving the answers buyers actually see.

Keep the exact question, platform, date, language and relevant location context with each capture. Note sign-in or personalisation where it affects the method. If a platform does not expose the model or another detail, label it unavailable.

Keep new questions in a separate group until you have a comparable starting result. “What is SEO?” and “which agency can diagnose our WordPress indexing problem before a redesign?” concern different decisions. A mention for one cannot stand in for the other.

You do not need identical answers on every run. You do need a consistent method that explains repeated checks, missing answers and changes in the platform.

What can Search Console currently show?

Check which Search Console report your property actually has, then read what its numbers count.

Google’s generative AI performance report documentation, checked on 4 October 2026, describes impressions for supported Search AI features. It provides page, country, date and device views. An impression is a displayed link event, rather than a qualified enquiry.

Save the date range and filters with the export. Without them, next month’s comparison can be misleading. Some chart and table totals use different aggregation, so a mismatch should be investigated before it becomes a performance claim. If a report is unavailable in your property, record that fact instead of presenting a guessed total.

This is Google evidence. It does not measure all answers on ChatGPT, Perplexity or other independent services. Ask the agency to label each platform’s evidence separately.

How do you check whether the agency did the work?

Compare the report with the agreed deliverables and a dated change log.

A content task should identify the URL, what was changed and why. A technical correction should show the affected response or setting and its verification. A profile task should show the authorised destination and completed change.

Ask which recommendations remain waiting on your team. A transparent report may show unfinished work because a developer or subject expert has not approved it yet. Hiding that dependency makes next month’s plan less useful.

For a published guide, check the public page, links and author information yourself. Your agency should also be able to explain which assertions came from your team, primary documentation or an explicitly hypothetical example.

Our AI SEO agency selection guide helps put those acceptance checks in the brief before work starts.

What makes a score difficult to trust?

Start with the calculation. Without the number of questions and the actual sample, you cannot tell whether a higher score reflects improved answers or simply a different set of tests.

Suppose a hypothetical report says visibility rose from four mentions in twenty questions to six mentions in twenty questions. That is a change from 20% to 30%: ten percentage points, or a 50% relative increase. Both calculations describe the same sample. Keep those labels clear. “AI authority improved by 40%” means little without the formula, the underlying events and the reason the change matters.

Keep score changes separate from content accuracy. A business mentioned more often with the wrong location or service description has a correction problem even if the graph rises.

Ask how repeated runs, missing answers and changed platform behaviour are handled. The method should make uncertainty visible rather than erase inconvenient results.

What should your renewal decision consider?

Before renewing, check what was finished, whether you trust the evidence and what your buyers still need.

• Were the agreed tasks delivered and verified?
• Did the research cover the buyers and questions you care about?
• Can your team inspect the evidence and access its own accounts?
• Are enquiries and qualification measured honestly?
• Does the next plan address a clear remaining problem?

You might renew a useful programme before sales impact is conclusive, especially where measurement is incomplete. State that decision as an investment in defined work, not proof of a return the report cannot establish.

Alternatively, pause the recurring content batch and fund the technical or catalogue task blocking progress. A renewal meeting should leave you with a choice, rather than another chart and an automatic invoice.

Buyer questions

Sources and editorial method

Technical claims link to primary documentation in the relevant section. Buying worksheets are Digital Magma’s editorial recommendations. Examples are illustrative, not client results. This guide was prepared with AI assistance; no measured search volume or ranking outcome is asserted. Send corrections through our contact page.

Related buying guides

• Should you hire an AI SEO agency for B2B leads or fix SEO first?
• Google Ads lead quality audit: what should you check before increasing budget?
• Investigate pages that are crawled but not indexed

The cheaper form submission can cost you more

Suppose Campaign A spends ₹20,000 and brings in 100 enquiries. Twenty meet your sales team’s qualification rules. Campaign B spends ₹15,000 and brings in 150 enquiries, but only ten qualify.

Campaign B looks cheaper in a basic lead report: ₹100 per enquiry versus ₹200. Once you compare qualified leads, the result reverses.

Divide spend by qualified leads: ₹15,000 ÷ 10 = ₹1,500. In this example, B costs 50% more per qualified lead even though its form submissions cost half as much.

This is a calculation, not a campaign result or a promise of improvement. It also does not tell you which campaign produces more sales. That needs a later sales outcome and a consistent way to record it.

Before increasing the budget, find out whether your report can make this comparison at all. If the sales team cannot connect its enquiries to the campaigns, start there.

Which actions are used as bidding goals?

Ask the agency to show the conversion actions and the goals actually applied to the campaign.

Google distinguishes primary and secondary conversion actions. Their role depends on how goals are configured, including custom-goal exceptions. Do not assume that a phone-button click and a qualified opportunity should be rewarded in the same way.

Request a list of action names, sources, counting settings, intended business meaning and campaign usage. Check for duplicate tracking as well as missing tracking. A form event firing twice can distort a result even when every visitor is real.

The useful question is: “What does this account treat as success?” A screenshot of a falling cost per conversion does not answer it.

How do you trace an enquiry through to a sale?

Pick a small sample of recent enquiries you are authorised to inspect. Follow each one from the form or call record into the sales system. Can you find the campaign, the qualification decision and what happened next?

Agree the qualification rule before counting. For a local service business, an illustrative rule might be: needs a service you sell, is in an area you serve and has provided enough contact information for a follow-up. Your sales team must decide its own rule and use it consistently across campaigns.

A duplicate submission should not become two opportunities. A person asking for a job should not be counted as a prospective customer simply because they completed the same form. Keep these reasons in the sales record so the marketing team can see the pattern.

Google’s enhanced conversions for leads guidance explains how first-party outcome information can support measurement. Any integration needs an account-specific review and appropriate data handling.

Test a record arriving once with the intended status. Then establish who handles duplicates and later status changes. A “CRM connected” badge does not answer either question. Keep personal lead details out of public reports.

What can you learn from the search and landing-page evidence?

Look for a mismatch between the buyer’s question and the offer they reach.

A company selling ongoing maintenance may receive requests for a free tutorial if its ads and landing page never make the commercial scope clear. The right correction may be wording and query targeting, rather than a larger daily budget.

A test is a hypothesis, not a promised improvement. Record what changed so you can interpret the next period without mixing several explanations together.

When is more budget a reasonable next step?

Consider increasing spend after the useful outcome is measured and your team can handle more of those enquiries.

Check what a qualified enquiry costs, whether sales can respond and how much data you have. A tiny number of sales cannot justify a confident forecast. Equally, a healthy campaign may be constrained by a slow response process rather than lack of demand.

Use a planning calculation with your own inputs: advertising cost divided by attributed qualified opportunities gives a cost per qualified opportunity for that defined period. It does not prove incremental sales or future performance.

Ask the agency what would invalidate its recommendation. A defensible budget proposal should explain the evidence, assumptions and review point, rather than making a revenue promise from clicks alone.

What should a paid lead-quality audit deliver?

The audit should explain what is wrong, what to change and who will make the change.

• An agreed qualification definition and a reconciliation sample.
• A conversion-action and campaign-goal review.
• Specific query, offer and landing-page findings.
• Implementation tasks with owners and acceptance checks.
• A reporting plan that separates submissions, qualified leads and sales.

Clarify whether implementation is included. Ask how sales feedback will reach the person managing the campaign, and whether your team can access the resulting records after the engagement ends.

If the enquiry form or page needs rebuilding, coordinate the audit with a website project scope review. Advertising and website work should use the same definition of a completed enquiry.

Buyer questions

Sources and editorial method

Technical claims link to primary documentation in the relevant section. Buying worksheets are Digital Magma’s editorial recommendations. Examples are illustrative, not client results. This guide was prepared with AI assistance; no measured search volume or ranking outcome is asserted. Send corrections through our contact page.

Related buying guides

• WordPress SEO audit or redesign: what should you buy when pages are not indexed?
• What should an AI SEO agency report before you renew its monthly package?
• Investigate pages that are crawled but not indexed

Is the problem a setting or the website itself?

Your WordPress site looks dated and enquiries have slowed down. A redesign sounds like a sensible next step. But if an important service page has an accidental noindex directive, new colours and a better hero image will leave the indexing problem in place.

That is one reason to diagnose the page before buying the project. A focused audit can separate a wrong setting from a template that cannot support the content you need.

These are illustrative cases. An audit should establish which one resembles your site, rather than turning every technical issue into a rebuild.

Before the sales call, choose a service page and a contact page. Open both while logged out, read them on a phone and complete a test enquiry. You now have a specific problem to discuss instead of “the site feels old”.

When is a focused technical fix enough?

A focused fix fits a specific fault with a limited area of impact.

These are decision examples. They do not establish that your site has those faults. Ask the supplier to connect the recommendation to a response, page capture or account observation.

When does a redesign make sense?

Choose a redesign when the existing templates cannot support the pages your customers need.

Perhaps important service information is scattered across templates, the mobile contact form is difficult to use, and the editor cannot maintain the content. Replacing those templates may be more useful than repairing each symptom independently.

Separate the visual wish list from the business requirements. Specify the pages that must exist, the tasks readers must complete and the controls your team needs in WordPress. Then ask what must change underneath the design.

Google’s hiring guidance recommends considering SEO early in a redesign. That is a planning point, not a claim that a new design automatically improves rankings.

What should happen to existing URLs?

Keep useful addresses when the redesign does not require them to change. A new layout is not, by itself, a reason to replace every URL.

Where a path must change, agree the destination before launch. Here is a hypothetical map a buyer could ask the developer to demonstrate:

The sitemap can miss useful addresses. Check analytics and existing links too, and ask sales which pages they still send to prospects. Save the old content.

Google’s site-move guidance covers URL mapping, redirects, sitemaps and monitoring. Use that process where URLs change; keep a same-address template update scoped to the work it actually needs.

Test the map first. Pick a service URL, an old article and a retired offer, and follow each redirect as a customer would before approving the full list.

How should JavaScript rendering be handled?

Make the main explanation and ordinary navigation accessible without requiring a visitor to trigger an animation or load-more button.

Google can render JavaScript, but its JavaScript SEO documentation describes processing limits and recommends considering server-side or pre-rendered content. For a WordPress service page, ask to see the actual document HTML and the rendered page.

Animation can improve feedback when a menu or accordion opens. It should not hide the service description until a script succeeds. Test an interrupted script load as well as the normal page, especially if the design uses a custom page builder.

Include readable text, meaningful links and accessible controls in the acceptance brief. “The animation looks good on my laptop” is a different test from “a buyer can read the offer and contact us.”

What should you approve before launch?

Who signs off? Agree a release checklist that your developer and SEO consultant can both use, with an owner for each check and a recovery plan if the deployment fails.

• Old and new URL behaviour, canonicals, robots directives and sitemap output.
• Mobile and desktop service-page readability, navigation and forms.
• Important content present in the anonymous response and rendered page.
• Tracking tested against the intended enquiry actions.
• Backups, rollback responsibility and a dated deployment record.

Keep delivery acceptance separate from indexing and traffic. The supplier can verify a deployed redirect or working form. Google decides subsequent crawling and index selection.

If your developer and SEO consultant are separate, ask them to share the same checklist. Otherwise, a task can be marked complete by one team while remaining untested by the other.

Buyer questions

Sources and editorial method

Technical claims link to primary documentation in the relevant section. Buying worksheets are Digital Magma’s editorial recommendations. Examples are illustrative, not client results. This guide was prepared with AI assistance; no measured search volume or ranking outcome is asserted. Send corrections through our contact page.

Related buying guides

• What should an ecommerce AI search audit include before you hire an agency?
• Should you hire an AI SEO agency for B2B leads or fix SEO first?
• Investigate pages that are crawled but not indexed

Follow one product before auditing the whole catalogue

Suppose your blue shirt in medium costs ₹1,499 and is out of stock. The product page shows both correctly. Its structured data agrees. Your shopping feed still lists ₹1,299 and “in stock”.

Now the catalogue team has something to investigate. A report that says “your shop needs more AI visibility” leaves them guessing.

In this hypothetical example, the next job is to trace the feed’s price and stock values to their source. Check the export time first. Then check whether the feed uses the parent product while the page shows the selected size. Those are questions to investigate, not conclusions to put in a report without checking.

Start the sample with products that expose these differences: a variant, an unavailable item, a recently changed price or an unusual shipping condition. A homepage-only audit will miss most of them.

Keep the SKU, size, colour, destination market and check time beside each finding. “Wrong stock” becomes a useful task only when someone else can reproduce it.

Do the page, schema and feed agree?

Compare the same offer. A medium blue shirt and a large blue shirt can have different prices or stock, so record the selected variant before comparing the page with its feed.

Google explains two routes for rich product data: product structured data and Merchant Center feeds. Providing both can broaden eligibility and help Google verify product information. That is Google’s guidance. Other assistants may use different sources.

Do not “fix” the audit by changing only the markup while leaving the price a shopper actually sees unchanged. Find the source of the mismatch.

What does an out-of-stock recommendation prove?

It proves that the captured answer and the checked availability disagree. It does not by itself identify the source or age of the assistant’s information.

Here is a hypothetical test: a shopper asks for a blue shirt in medium, and an assistant recommends the product while that variation is unavailable. Save the question, complete answer, visible source links, selected variant and time. Then check whether the page, schema and feed all showed the current state.

If the page, markup or feed was wrong, fix that record first. If the data agreed, inspect the cited source and any platform-specific retrieval documentation before blaming the feed. The assistant may have used another page or older information.

An audit should state the limit of the finding. “This answer was stale in this test” is defensible. “Every AI platform ignores your stock feed” requires much broader evidence.

What should the technical access check cover?

Check the exact product URL as an anonymous visitor and save what the page returned.

• A product URL that returns HTTP 200 and names the correct canonical.
• Robots directives and any CDN or firewall challenge affecting the agreed crawler.
• Product description, price and stock available in the delivered or rendered page.
• Useful internal links from category and related product pages.
• Structured data that agrees with the selected product.

A plugin setting is only a starting point. Open the product and compare what it shows with Google’s Merchant Center structured-data requirements, including the price and availability for the selected offer.

For an indexing exclusion, use the crawled currently not indexed guide before changing the whole catalogue.

How do you define acceptance before paying?

Another person on your team should be able to open the product, choose the recorded variant and see why the finding matters. If they cannot, the report needs more detail.

This follows the hypothetical example above. The cause still needs checking. A stale export and a plugin bug are possible explanations, but this mismatch alone cannot tell you which one happened.

Give product accuracy to the person who owns the catalogue or operations policy. A marketer can improve a returns explanation, but should not invent the return window. Valid markup can describe an incorrect policy just as easily as correct one.

For a large catalogue, test the proposed rule on a normal product and an awkward variant before applying it everywhere. Keep the old mapping so your team can reverse an unintended change. Treat this pilot as part of the scope, with its own approval.

What belongs outside the first audit?

A new checkout integration, an inventory-system replacement and a broad content programme are separate projects unless the agreement explicitly includes them.

Ask for these costs to be itemised. You may only need to correct a feed mapping or a product template. Paying for a full rebuild before tracing the mismatch makes the decision harder to assess.

Likewise, a presence report cannot prove sales attribution. The audit can capture what an assistant said, what it cited and which factual errors appeared in the sample. Orders and qualified enquiries require their own measurement evidence.

Use this scope to compare suppliers on the same job. The lowest quote is less helpful when one proposal includes fixes and another ends at recommendations.

Buyer questions

Sources and editorial method

Technical claims link to primary documentation in the relevant section. Buying worksheets are Digital Magma’s editorial recommendations. Examples are illustrative, not client results. This guide was prepared with AI assistance; no measured search volume or ranking outcome is asserted. Send corrections through our contact page.

Related buying guides

• WordPress SEO audit or redesign: what should you buy when pages are not indexed?
• What should an AI SEO agency report before you renew its monthly package?
• Investigate pages that are crawled but not indexed

Start with the buyer who cannot find the answer

Imagine you run a software consultancy. One proposal offers twenty AI search articles a month. Another starts by fixing your migration service page. Which should you buy?

The article count cannot answer that. Your prospective customer might already have found the website and still be unable to tell whether you preserve product URLs during a WooCommerce migration. That missing answer belongs on the service page, where the customer is deciding whether to contact you.

Already have a clear service page, but never appear when buyers compare suppliers? Ask the agency to investigate those comparisons. Find out which questions buyers ask, which sources the answers cite and what those sources say about your service.

This is a buying example, not a client result. Its point is simple: the first proposal should explain the problem it will solve before it promises a volume of content.

When does ordinary SEO work need to come first?

Start with access, clear service information and a working enquiry path. Adding another content programme to pages that cannot be fetched or understood gives you more work to diagnose.

Google’s AI search guidance says supporting links in AI Overviews and AI Mode require indexing and snippet eligibility. It does not prescribe special AI markup. That requirement applies to Google; ask for a separate explanation of any other platform in the proposal.

• Can an anonymous visitor read the main service explanation without a login or button click?
• Does the page describe who the service suits, what is included and how to ask for a quote?
• Do enquiries reach the right person, with a record of their source where available?
• Can someone approve and implement technical or content changes?

If these checks fail, commission the specific repair first. You can still plan AI research alongside it, but keep the responsibilities and budget visible.

How do you compare an AI SEO proposal with an SEO proposal?

Compare the work and the evidence, rather than the acronym in the heading.

A supplier can legitimately offer research without implementation. The problem is paying for implementation while receiving a spreadsheet that your team has no time to act on.

What should the first paid project include?

The first project should tell your team what to change and why. A long report that leaves every task waiting for someone else is a poor starting point for a small team.

Agree which buyer questions will be tested and which pages will be reviewed. For each proposed change, the agency should explain the finding, the person who will fix it and the check that confirms it reached the website.

Then settle the practical details. Will the agency publish the copy, or send a document? Does the quote include a developer? Who pays for research tools? Your answers can change the cost and usefulness of the same recommendation.

This is our suggested starting scope, not a universal package. A larger team with its own developers may need research alone. A founder without development support may need the agency to carry the change through to publication.

Which evidence should you ask to see?

Ask for evidence that resembles your situation. A screenshot of a brand mention does not establish that a provider can improve your lead quality.

Look at an audit finding beside the change it led to. Then see how the agency checked that change. Private client information may need redaction. A supplier can explain the method without exposing confidential records.

Keep measured outcomes separate from examples. If an agency shows an illustrative report, it should be labelled. If it cites a traffic gain, ask about the date range, starting point, affected pages and other changes that happened during the period.

Use our AI SEO reporting buying guide to define what a mention, citation, visit and enquiry mean before they appear in a contract.

What can make a sensible proposal a poor fit?

A good plan can still be too much for your team to carry out.

A founder-led business with no developer may need implementation support more than another audit. A larger B2B team may have writers already and need help deciding which buyer questions to investigate. Buy the help your team is missing.

Pause if the proposal promises permanent placement, offers no account access, or cannot identify the pages it will change. Also pause if the agency asks you to publish claims your team cannot support. Your website carries those assertions after the engagement ends.

In the buying meeting, ask: “What would you advise us to postpone?” A useful answer shows that the supplier can prioritise within a real budget.

Buyer questions

Sources and editorial method

Technical claims link to primary documentation in the relevant section. Buying worksheets are Digital Magma’s editorial recommendations. Examples are illustrative, not client results. This guide was prepared with AI assistance; no measured search volume or ranking outcome is asserted. Send corrections through our contact page.

Related buying guides

• What should an AI SEO agency report before you renew its monthly package?
• WordPress SEO audit or redesign: what should you buy when pages are not indexed?
• Investigate pages that are crawled but not indexed

First, decide whether you want this URL in search. Your main service page probably belongs there. A shopping basket or a filter URL that repeats another product listing may not. Fixing the latter just to increase your indexed-page count can send you in the wrong direction.

Google’s Page indexing report says the status can change and there is no need to resubmit an unchanged URL.

The label does not tell you what caused the exclusion. If someone recommends a rewrite immediately, put the page and its inspection result in front of them first. What, exactly, needs fixing?

Which indexing status are you actually seeing?

Read the full status label first. A URL Google has discovered needs a different investigation from one it has already fetched.

Use the label shown for your exact URL. Similar-looking labels can point to very different next steps.

Why does the live test pass while the page stays unindexed?

The live test checks what Google can fetch now. The indexed result describes the version recorded in Google’s systems. Start with that recorded result.

Copy the exact URL into Google Search Console. Record the indexed result before running a live test. Then note the last crawl date and the Google-selected canonical where it is available. Save both screenshots.

A passing live test is useful: it rules out some current access problems. But it does not run all the checks Google uses to select pages for its index, including duplicate and canonical evaluation. That is why the green result and the indexing exclusion can both be correct.

Check the dates. If Google last crawled the URL before your rewrite, the inspection may still describe the older page.

Google explains these limitations in its URL Inspection documentation. For a single URL, read its inspection details alongside the broader report. Their dates and scope differ.

Is Google choosing another URL as the canonical?

When several URLs contain similar material, Google may choose one to represent them in search. Confirm which one you intended to keep.

Look at the HTML canonical, redirects and internal links together. If a product exists at both a clean address and a tracking-parameter address, link to the version you want indexed. Keep that choice consistent in your sitemap.

Compare the Google-selected canonical with the address you want visitors to find. If they differ, inspect both pages before changing the annotation.

For example, if a tracking URL repeats the clean product page, make your canonical, internal links and sitemap agree on the clean address. If the two pages serve different needs, compare their content before treating them as duplicates.

Do not replace every canonical with the homepage URL. That can confuse the relationship between unrelated pages. Google’s canonical documentation covers redirects, HTML annotations, sitemaps and inconsistent signals.

What should you check on a WordPress page?

Check the page while logged out. Your administrator view can hide a problem that visitors and crawlers still encounter.

Open the published URL in a private browser window. Does it show the full article? Then check the document response, canonical and robots directives. A correct setting in your SEO plugin helps only if the public page delivers it.

• Confirm the production site is not carrying a staging noindex directive.
• Check the SEO plugin’s setting for this post and its content type.
• Inspect the visible content after your page cache and CDN refresh.
• Check that a security plugin is not returning a challenge or empty response.
• Follow the published navigation and content links to the page.
• Verify the primary copy is available in rendered HTML.

A saved change does not always reach the public page immediately. If your editor shows the new version while an anonymous visitor sees the old one, inspect the page cache and CDN response. Another rewrite will leave that delivery problem unresolved.

For JavaScript-heavy pages, inspect the rendered version as well as the initial response. A working interface can conceal missing text from a crawler. Google’s JavaScript SEO guide explains how crawling, rendering and indexing fit together.

Does this page answer anything your other pages miss?

Put this page beside the other article on your site that answers the same question. Can you explain why a visitor needs both?

A city-name swap rarely answers that question. Neither does another 500 words of background. Look for the missing detail: a local constraint, a worked example or the step someone keeps getting wrong.

Add something the reader can use: an observed error, a worked example, a dated source or a decision the existing explanation leaves unresolved. Extra length alone supplies none of those.

Suppose two guides explain exactly how to submit a sitemap. Keep the stronger guide, merge any useful missing details and plan an appropriate redirect if you retire the redundant URL. If they solve separate problems, make the distinction visible in the title, opening answer and internal links.

Keep useful old material. Before consolidating anything, export the content and check its existing traffic, links and business role.

What evidence should you record before choosing a fix?

Write down the observation before deciding what it means. “Google last crawled the old version” is evidence. “The new page is too short” is a guess until you investigate.

Start with the URLs that matter to your business. The hypothetical worksheet below shows how one observation can lead to a specific check. It is not a client recovery report.

Keep one row per URL. Record the crawl date, canonical, response status and robots directives, then add what you changed and when. When Google crawls again, you can compare the new result with that record instead of trying to remember what was live last week.

How do you save a useful record of the problem?

Save the document response before changing a setting. A screenshot of the status alone leaves your developer with very little to investigate.

In your browser’s developer tools, open Network and reload the page. Select the main document request rather than an image or stylesheet. Save its final URL, status and response time, along with the date and timezone of your check. Remove cookies and authorization headers before sharing an export.

If present, save the Cache-Control, Age, Content-Type and X-Robots-Tag headers. They can help your developer investigate caching or an indexing directive. To check whether an update reached the public page, compare its canonical and opening paragraph with the copy saved in WordPress.

If your developers use command-line checks, ask them to preserve the redirect chain, HTTP status and response body as separate artifacts. A header-only request cannot tell you whether the article renders correctly. Keep credentials out of screenshots, shared HAR exports and public issue reports.

Link the saved response to the URL’s worksheet row. If a firewall challenge disappears before your developer opens the page, that capture may be the only record of what you saw.

What did our own website check establish?

Our breadcrumb update exposed a useful limit of public website checks: a working page tells you little about whether Google has selected it for search.

On 4 October 2026, Digital Magma checked 145 sitemap-linked URLs, including the homepage, during its breadcrumb update. All returned HTTP 200, and every non-home URL contained BreadcrumbList markup. This was a public HTTP audit, not a fresh Search Console indexing export.

So we could confirm the responses and markup. We could not use that result to claim 145 indexed pages. That would require current Search Console evidence.

When should you request another crawl?

Request a crawl after publishing a new page or making a meaningful correction. The request tells Google there is something to fetch, without deciding whether it belongs in the index.

Once the page is accessible and the change is worth recrawling, use URL Inspection for an individual address. Use an updated sitemap to communicate a collection of new or revised URLs. Keep modification dates truthful.

Keep a dated record of the change. Google’s recrawl guidance says crawling can take days to weeks and repeated requests do not speed it up. There is no acceptance deadline.

When the next crawl appears, compare its date with your update. If it happened earlier, give Google time to fetch the revised page before judging the result.

Can an unindexed page appear as a Google AI search link?

Indexing is required for a page to qualify as a supporting link in Google’s AI Overviews or AI Mode.

Google’s AI search guidance says the page must also be eligible for a search snippet. No extra technical requirement applies solely to those answers. Eligibility does not promise a citation.

For your article, explain the answer clearly and link to the evidence. Use BlogPosting, Person and Organization markup to describe the actual article, author and publisher. Adding those types cannot make an excluded URL eligible by itself.

Other engines have independent retrieval systems. Google’s indexing status cannot certify visibility in ChatGPT, Perplexity or Bing Copilot. Check crawler access and actual cited answers separately when measuring those platforms.

Crawled currently not indexed: common questions

These answers address the decisions that usually come up after the first inspection.

Sources, author and how this guide was prepared

This guide uses Google’s documentation and the dated public-site check described above.

Research also reviewed public Reddit questions about conflicting reports and ecommerce parameters. Community explanations were treated as questions to verify, not technical authority. English, Russian and Chinese versions of Google’s recrawl guidance were cross-checked. No third-party article was translated into this post.

AI tools assisted the research and drafting. The technical references are linked above, the worksheet is hypothetical, and our site observation covers only the public checks described in that section. Found an error? Send it through our contact page.