You paid for a link, or spent three weeks earning one, and it is live. You can see it on the page. Six weeks later nothing has moved, and you are left with the question nobody in the outreach thread wants to answer: did that link ever count for anything?
The takeaway up front: a link being visible in your browser proves almost nothing. Before a backlink can pass any value, it has to clear a chain of separate technical conditions — it must exist in the HTML a crawler receives, sit on a page that can be crawled and gets indexed, point directly at your real URL, carry no restricting attribute, and sit in a placement search engines have reason to trust. A link can be perfectly visible to you and fail at any one of those links in the chain. Most "the backlink didn't work" cases are not mysteries about algorithms; they are one broken step you can check in about ten minutes.
What "passing value" actually means
Strip away the plumbing metaphors and a backlink does two mechanical things. It gives a crawler a path to your URL, and it associates your page with the anchor text and surrounding context on a page the engine has already formed some opinion about. Neither happens automatically; both depend on the engine actually processing that specific link on that specific page.
So the useful question is never "how much authority does this domain have." It is a sequence of yes/no questions: Did the crawler see the link? Was the page it sits on worth storing? Does the href resolve to my URL? Is the publisher vouching for it? Any single "no" ends the chain, and the domain's authority score is irrelevant from that point on.
The chain a link has to clear
Run these in order. Each is cheap, and each one knocks out a category of dead links that look fine on screen.
1. Is the link in the HTML a crawler receives?
Your browser runs JavaScript. A crawler's first pass often does not, and rendering is a second, queued, resource-limited stage. If the link is injected client-side — by a widget, a "related partners" carousel, a comment system, a cookie-gated content block — it may not be present in the raw HTML at all.
Check the source the server actually sends, not the DOM in your inspector:
curl -sL "https://example.com/host-page" | grep -i "yoursite.com"
No match while the link is plainly visible in the browser means the link is render-dependent. It is not necessarily worthless, but it is now waiting on a slower, less reliable pipeline and you should treat it as unproven.
2. Can the host page be crawled, and is it indexed?
A link on a page the engine will not store is a link in a locked room. Three things commonly close the door:
- robots.txt disallows the path. The crawler never fetches the page, so it never sees the link.
- The page carries
noindex. It can be crawled, and links on it can still be followed, but the page itself has no standing — and pages set to noindex long-term tend to get treated as effectively nofollow over time. - The page is an orphan. Nothing internal links to it, it is in no sitemap, and nothing external points at it. Freshly published "guest post" pages on neglected sites often sit like this for months.
The quickest test is a site: query on the host page's exact URL, or a distinctive verbatim sentence from it in quotes. If the page itself is not in the index, your link on it is not doing anything yet. The mechanics of crawl paths, sitemaps, and why links sit undiscovered are covered in the backlink indexing guide.
3. Does the href point at your URL — directly?
This is where paid placements quietly fail. Read the actual href, not the visible anchor text. Common breakages:
- A redirect script. The link goes to
example.com/out?url=yoursite.com, and that/outpath is disallowed in the site's robots.txt. The crawler stops at a door it is not allowed to open. This is standard on many bookmarking, forum, and aggregator platforms. - A chain of hops. Every extra 301 is another step, and any 302 in the chain signals "temporary," which is a weaker signal to consolidate.
- The wrong destination. A trailing slash difference, an
httpversion that redirects, awwwmismatch, a stale URL that now 404s, or a page that canonicalises to something else entirely. If your target page canonicalises elsewhere, the signal is consolidated to the canonical, not to the URL in the link. - Tracking parameters that produce a distinct URL. Usually resolved by canonical tags, but worth confirming rather than assuming.
Trace it properly instead of clicking:
curl -sIL "https://example.com/host-page-link-target" | grep -iE "^(HTTP|location)"
You want to see your canonical URL returning a 200 at the end of a short, clean chain.
4. What does the rel attribute say?
Read the anchor tag itself. rel="nofollow", rel="sponsored", and rel="ugc" all tell search engines the publisher is not offering an editorial endorsement. Google has described these as hints rather than absolute directives, so the engine decides how to use them — but you should plan on the assumption that a restricting attribute means no ranking signal.
Two traps are easy to miss. A site can apply <meta name="robots" content="nofollow"> sitewide, so every outbound link is restricted with no visible per-link attribute. And on some platforms, the link is fine but the outbound redirector described above is blocked — functionally nofollow, with clean-looking markup.
A restricting attribute does not make the placement pointless: referral traffic, brand association, and URL exposure are all real. It does mean you should stop expecting it to move rankings, and stop paying for it as though it will.
5. Is the placement one an engine has reason to trust?
The last check is judgement rather than a command-line test. The same domain can give you a link that counts and one that does not, depending entirely on where it sits. An editorial link inside body copy, surrounded by topically related text, is the shape search engines have every reason to weigh. A link in a sitewide footer, a blogroll, a repeated author bio, a comment, or a block of forty outbound links under the heading "Our Partners" is the shape they have spent two decades learning to discount.
Ask the honest question: would this link exist on this page if you had not asked for it? The further the answer is from "yes," the less you should expect from it.
Reading the evidence after the link is live
Suppose the link clears all five checks. You still cannot isolate its ranking effect with any precision, and anyone who tells you they can is selling something. What you can do is gather converging evidence:
- Confirm the engine has seen it. Search Console's Links report shows external links Google has associated with your site. It is incomplete and slow to update, but an appearance there is positive confirmation. Absence after a couple of months on an indexed host page is a real signal that something in the chain is broken.
- Watch referral traffic. Even one or two real referral sessions in analytics prove the page has a live audience — which is the thing you were paying for anyway.
- Track the target page's own metrics, not sitewide ones. Impressions and average position for the specific page's query set, in Search Console, is where a genuine link effect would surface first. Sitewide traffic is too noisy to attribute anything to.
- Give it a realistic window and expect noise. Recrawling, reindexing, and signal consolidation take time, and other pages, other links, and algorithm updates all move at once. Treat a single link as a contributor to a trend, never as a measurable line item.
When the link is live but the answer is still "no"
Some links pass every technical check and still cannot do much. A link on a page nobody visits, on a site whose own profile is a wall of exact-match anchors, in a paragraph written solely to house it, is technically flawless and practically inert. Technical validity is the floor, not the ceiling: it tells you the link can count, not that it is worth counting.
FAQ
How do I know if a backlink is passing link juice?
You cannot measure it directly — no tool reports the value transferred by a single link. What you can verify is whether the link is eligible to pass anything: present in the raw HTML, on a crawlable and indexed page, pointing directly at your live canonical URL, and carrying no nofollow, sponsored, or ugc attribute. If all four hold and the placement is editorial, the link is doing whatever it can. If any one fails, that is your answer.
Does a backlink count if the page it is on is not indexed?
Essentially no, at least not yet. If the host page has not been crawled and stored, the engine has not processed the link on it. Links on crawled-but-noindexed pages are a partial exception — they can still be followed — but a page kept out of the index long-term tends to lose that ability. Check whether the host page is in the index before concluding anything about your link.
Do nofollow backlinks pass any value?
Not ranking value in the way a followed link does. Treat nofollow, sponsored, and ugc as "no authority expected." They still deliver referral visitors, expose your URL to crawlers, and carry brand mentions, all of which are worth having. The mistake is paying followed-link prices for restricted-link placements, or refusing a strong, relevant mention because of the attribute.
How long before a backlink shows any effect?
There is no reliable number, and any specific figure you see quoted is invented. The sequence is crawl, index, then consolidate the signal, and each stage depends on how often the host page is crawled — which for a neglected site can be rarely. Practically: confirm the technical chain immediately, then look for the link in Search Console and for movement on the specific target page over a period of weeks, not days.
Why is my backlink not showing in Search Console?
Most often because one of the chain steps failed: the host page is not indexed, the link is injected by JavaScript, or the href runs through a blocked redirector. It can also be simple reporting lag or sampling — the Links report is not exhaustive and does not show everything. Verify the chain first; if every step passes and the host page is indexed, wait, then treat continued absence as evidence the link is not being counted.
Next step
Before you commission another placement, audit the ones you already own. Run each live link through the chain — present in raw HTML, crawlable and indexed host page, direct hop to your canonical URL, no restricting attribute, editorial placement — and you will usually find that a share of what you paid for never had a route to counting. That audit costs less than any new link, and it tells you which vendors and tactics to stop funding. To systematise the checks, pick your auditing stack from the curated list of SEO tools for marketers and agencies.
