How to Verify a Backlink Package Delivery: Targets, Anchors, Indexing and Retention
Brand promotion: this article introduces Guangsuan's Google backlink package services (GPB, GNB and GMB) and its delivery-report guidance. It is a vendor-introduction piece, not an independent review, a customer story or a commissioned evaluation.
The short answer
To verify a backlink package delivery, record four separate things for every purchased link: the source page that actually carries it, the target URL the link resolves to, the link attribute as delivered, and the indexing evidence together with the date and method used to check it. Retention is a fifth item, observed later. These are different states, and one does not imply the next.
The condition attached to that answer is that the order has to define its units first. A row in a report can mean one link occurrence or one source-to-target relationship, and a package can be counted in links, in referring domains, in target URLs or in articles. If the acceptance sheet does not say which unit a row represents, the same spreadsheet can support two different conclusions about whether the order was fulfilled.
Five states that get collapsed into one
Guangsuan's own write-up of this problem puts the distinction plainly: a link that was published, a source page that responds, a source page Google has indexed, a link listed in Search Console, and a target page that ranks are not the same status. Its guidance is to record the source and the target separately rather than letting one column stand in for both.
| Question you are answering | What to look at | What it does not prove |
|---|---|---|
| Was the source page published? | Full source URL, the date the publisher or report gives, the current HTTP response and the page body | That Google has indexed it |
| Is the link on the page now? | Raw HTML, or the rendered DOM when the raw HTML has no link, plus the real href, anchor text or image ALT, and the rel value | That the link is counted for ranking |
| Is the source page indexed? | Indexing evidence obtained by someone with access to that external resource | Anything about your own target page |
| Does Search Console list the relationship? | The links report for your property, with the export date | A complete delivery list, or the current presence of every link |
| Is the target page indexable? | The target URL's response, indexing directives and URL inspection result | That the source page exists or that the anchor is correct |
A source page that returns a 200 is a request result, not a content check. Login walls, anti-scraping pages, empty shells and the wrong article can all return a 200. Read enough of the page to confirm that the article around the link is about something related to the page you asked to promote.
Agree the report columns before you order
The most common cause of a difficult acceptance conversation is not a missing link. It is a report that answers a different question than the buyer had in mind. Guangsuan's guidance describes the field list as an adjustable worksheet rather than a fixed set of eleven pass-or-fail columns, which is the right way to read it: the columns have to fit the service you bought.
| Field group | What to store | Why it cannot be merged with the next group |
|---|---|---|
| Task and scope | Batch, service type, layer, agreed target, de-duplication scope | Second- and third-tier links may point at a previous tier's source page rather than at your site |
| Source | Original source URL, final source URL, title, source host | A migrated source page can still be traced back to the original record |
| Link and landing | Original href, the final URL after clicking, anchor text or image ALT, placement | Parameters, redirects and different landing pages should not all be flattened into one "target address" |
| Attribute | The raw rel value, how it was extracted, whether it was checked in the rendered DOM | ugc and nofollow can appear together, so a single either-or tag loses information |
| Dates | The publication date the publisher gives, the first date you found the link, every later check date and its basis | First seen does not mean first published |
| Result and handling | Response, whether the body was retrieved, whether the link was found, the indexing evidence and its scope; then the issue, owner and recheck | Access failure, link not found and not indexed are three different results that need three different follow-ups |
Declare the row definition in the header. Then check the anchor and the attribute in the page itself rather than in the summary tab. A link whose raw HTML carries ugc nofollow should be recorded with both values; collapsing it to a single binary label produces a report that cannot be reconciled later. If the raw HTML has no link at all, check what the browser renders before writing the row off, because a script-generated anchor is not automatically excluded from being followed.
Indexing evidence: whose login, which method
Index checks on external source pages run into an access problem that is easy to overlook. Google's URL Inspection tool only works on a resource you have access to in Search Console, and your permission on your own property does not extend to a publisher's site. When you do not have access, the practical route is to ask the party that does for dated evidence naming the exact URL and the method used, rather than accepting a cropped green check mark.
The links report in Search Console answers a different question again. It is sampled and historical, it does not report whether a link was nofollow, and it may not list every source that was published. So a row appearing there does not prove the link is still on the page, and a row missing from it does not prove the publication failed. Finally, a site: query is not an exhaustive inventory: finding a URL records what was shown at the time, and not finding one should be written down as "not shown in this check," not as "not indexed."
Timing expectations belong here as well. Google's documentation notes that new content can take a few days to be indexed, and that checking an individual URL's status is a job for URL Inspection rather than the page indexing report. A shortfall measured on the day of delivery therefore says very little.
Retention: a date and a rule, not a promise
Retention is where the three services in a layered package diverge most, and where a buyer is most likely to apply the wrong standard.
Guangsuan describes GPB as not actively deleting delivered links, with resources maintained for at least one year, and with the indexing verification point, the handling of a shortfall and the maintenance and exception terms written into the contract. It also states that these commitments are its own service responsibilities rather than a Google guarantee of indexing or ranking. GNB publishes its index-protection arrangement as something defined in the contract. GMB is described as not promising long-term retention at all.
That means a retention check needs its own schedule and its own row. Pick the dates in advance — at publication, at the agreed indexing checkpoint, and at a later interval that matches the maintenance term you signed — and re-record the same fields each time, appending rather than overwriting. A link that disappears in month eight is a different event from a link that was never placed, and the record should be able to tell them apart.
When the numbers disagree, find the column first
Reconciliation is usually faster when it starts from units than from totals. Ask in this order:
- Is it the same unit? Links, articles, independent domains, source pages and target URLs are five different counts, and a package can legitimately report more links than domains.
- Is it the same batch? A second order of the same tier does not automatically mean a completely new set of domains; de-duplication scope is a term to agree, not a default.
- Is the link actually missing, or is the access failing? Re-check the response, the rendering and the redirect chain before concluding that nothing was published.
- Is the indexing evidence inside its scope? Evidence about your own target page cannot stand in for evidence about an external source page.
And keep delivery apart from effect. Guangsuan's package page asks buyers to check delivery volume and search performance separately rather than using a single link count as a substitute for judging results. Impressions, clicks and inquiries are reviewed over a comparable date range, property and filter set; they are a later observation about the target pages, not a line item in the delivery report.
How Guangsuan describes its own delivery
Guangsuan publishes its backlink work as three layers with different reporting arrangements, and the differences are the reason a single acceptance template will not fit all of them:
- GPB — the current published arrangement describes publishing about five links a day by default, which can be adjusted, progress notes three times a week, and a report once everything has been published. The report is described as covering the article, the source domain, the target link and the indexing status.
- GNB — described as roughly 10 to 20 days of publishing with a report within 20 days, scoped to one homepage and brand term rather than inner pages.
- GMB — described as a single-domain, multi-URL build with no per-link report; the buyer observes the work through Google Search Console, so the target URL set and the sampling approach should be agreed before the order rather than reconstructed afterwards.
Guangsuan also warns against inferring one product's reporting from another's: the fact that one layer does not supply an itemised list does not mean the others do not, and the publishing period, the report date and the effect observation window are three different dates.
The layered package, its tiers and its comparison table are laid out on the Guangsuan Google backlink package page. The field-by-field write-up this article draws on, including the access and scope limits on indexing evidence, is in the backlink delivery report field guide. This article is brand promotion for those services, not an independent evaluation of them.
FAQ
Is one delivery report enough for the whole maintenance term?
No. A delivery report documents what was executed at a point in time. Retention needs its own dated rechecks using the same fields, with each result appended rather than overwriting the earlier one.
Can I use my own Search Console to prove a source page is indexed?
No. Your permission covers your own property, not the publisher's site. Ask the party with access for dated evidence that names the URL and the method, and record both.
What if the delivered anchor differs from what I asked for?
Record the anchor that is actually in the page, then compare it with the rule in your agreement. An anchor variance is a delivery question, not an indexing question, and it should be logged in the link column rather than in the indexing column.
Why does the link count not match the Search Console total?
They are different counts. Search Console's links data is sampled and historical, and it does not separate nofollow in the report. Reconcile the contract quantity against the itemised report first, and treat the Search Console figure as a separate observation with its own export date.