Sitewide, footer and widget links: are they a problem?
Sitewide, footer and widget links are not a problem in themselves. They become one when they are distributed widely, carry keyword anchors or are paid for. Google's spam policies name both footer and widget links, and most such links are discounted even when they break no rule.
In short
- Google's spam policies, last updated in August 2026, list widely distributed links in the footers or templates of various sites as link spam.
- The same policies list keyword-rich, hidden or low-quality links embedded in widgets that are distributed across various sites.
- A sitewide link adds one referring domain however many pages it appears on, so thousands of backlinks from it count for little.
- A Google patent known as the reasonable surfer model is reported to describe footer and sidebar links passing less value than links in the main text.
- The compliant form of a credit, badge or widget link uses a branded anchor, is visible to readers and carries a nofollow attribute.
Sitewide, footer and widget links are not a problem in themselves. Sister brands link to each other in footers, agencies sign their work, and tools carry a maker’s credit. The trouble starts when a link like this is spread across many sites, uses a keyword as its anchor or is paid for. Google’s spam policies name both patterns. Even when no rule is broken, these links tend to be discounted, so they are rarely worth chasing.
What the terms mean
The four terms overlap, so it helps to fix them before going further. A sitewide link is a link that appears on every page of a site, or nearly every page, because it sits in the template. A footer link is a link in the page footer, and it is usually sitewide for that reason. A blogroll or sidebar link belongs to a list of links in a sidebar, repeated on every page. A widget, badge or embed link is a link built into something other sites install or paste in, such as a calculator, a plugin, a theme, an award badge or the embed code of an infographic. The first three describe where a link sits on one site. The fourth describes how a link travels to many sites at once, which is why it gets a policy line of its own.
This lesson is about links between different sites. The navigation and footer of your own site, linking to your own pages, are a separate subject. What ties the four kinds together is that nobody placed them page by page. Someone added the link to a template or a piece of code once, and it then appeared wherever that template or code was used, so the number of pages carrying it says nothing about how many separate decisions were made to link.
What Google’s policy says
Google’s spam policies, last updated in August 2026, define link spam as creating links “primarily for the purpose of manipulating search rankings”. Two of the listed examples apply here. One is “Widely distributed links in the footers or templates of various sites”. The other is “Keyword-rich, hidden, or low-quality links embedded in widgets that are distributed across various sites”. Read the wording closely. The first is about links spread across “various sites”, not one footer link between two related businesses. The second is about widget links that are keyword-rich, hidden or low quality, not every widget credit. In both lines the qualifying words carry the meaning, and a reading that drops them would condemn ordinary practice that the policy does not mention.
The same page says links bought for advertising are not a violation when they carry rel="nofollow" or rel="sponsored". A rented footer link with neither attribute is a paid link like any other. So there are two separate tests, and a link can fail either. One asks about distribution and anchors: is this a keyword link pushed out across many sites through a template or a widget? The other asks about money: was the link paid for and left unqualified? A rented sitewide link with a keyword anchor fails both, which is why it appears at the risky end of the table further down.
Why these links count for less anyway
Two things work against a template link even when it is innocent. The first is position. A Google patent known as the reasonable surfer model is reported to describe links on a page passing different amounts of value according to position, prominence and how likely a reader is to click. A link in the main text is treated as stronger than one in the footer or sidebar. The direction is well supported. The size of the discount is not known. The logic is easy to follow, though: a link inside a paragraph is there because the writer wanted the reader to follow it at that point, while a footer link sits where a reader is less likely to click, so a system that weights links by the chance of a click will give it less.
The second is repetition. A link on 5,000 pages of one site is 5,000 backlinks and one referring domain. Ahrefs’ January 2025 study of one million keywords found referring domains correlate with rankings slightly better than backlinks, at 0.255 against 0.248. The 2024 leak includes a field for repeated links from one source to one target, which analysts read as support for diminishing returns. Referring domains vs backlinks explains the counting. The gap between those two correlations is small, and neither is strong, so the study does not prove that the extra pages are worthless. On our reading, the result is consistent with the idea that the 5,000th copy of a link from one site tells Google little that the first did not, though the study itself does not test that.
Put position and repetition together and the commercial case weakens before any policy question arises. A sitewide footer link looks large in a backlink report and is, on this evidence, one weakly placed vote from one site. That is the reason for the advice in the opening paragraph: where such a link arises naturally there is no need to refuse it, but there is little reason to pay or campaign for one.
Common cases, sorted
| Case | Typical form | Problem? |
|---|---|---|
| Sister brands in one group | Footer links between sites with the same owner, branded anchors | No. A legitimate use |
| Genuine partner or sponsor | Logo or name in a footer or partners strip | Low risk with a branded anchor. Sponsor links that were paid for should be labelled |
| “Website by” credit | Agency link in the footer of every client site | It is a widely distributed template link. Use a branded anchor and nofollow |
| Theme or plugin credit | Link in a free theme footer or plugin output | Same as above. Keyword anchors here are the classic violation |
| Award badge or “as seen on” badge | Image with a link back, given to winners or members | Compliant when branded, visible and nofollowed |
| Infographic embed code | Pasted code containing a link with a chosen anchor | Keyword anchors in embed code are widget-style spam |
| Rented footer or sidebar link | Monthly fee for a sitewide link, often with a keyword anchor | Yes. A paid link and a template link at once |
| Blogroll swap | Two sites add each other to a sidebar | Falls under link exchanges when done at scale |
Sponsor links have their own rules, covered under sponsorships and events. Swaps are covered under link exchanges. Reading down the table, the same link in the same place moves from acceptable to risky as three things change: the anchor goes from brand to keyword, the number of unrelated sites carrying it grows, and money enters. The “website by” credit can come as a surprise, because it feels like a signature. Repeated across every client site it matches the first policy line word for word, which is why the cautious form is a branded anchor with nofollow.
The patterns are also easy to detect. Sitewide links leave an obvious footprint: thousands of links from one domain, all with the same anchor. When that anchor is a commercial keyword, your anchor text profile shifts sharply towards one phrase, and when the link goes live the anchor count jumps overnight. The 2024 leak documents fields that count spam-phrase anchors and measure spikes of them. Those fields are documented. Their weights are not known. Widget schemes leave a different footprint: the same keyword-rich link on hundreds of unrelated sites that all installed the same code. Practitioners report that widget link schemes have drawn manual actions in the past.
What happens next follows the usual pattern. Google mostly ignores links it judges to be spam. Where manipulative links dominate a profile, a manual action for unnatural links is possible. If you sell or host such links, the site carrying them can receive an action for unnatural links from the site. So the risk runs in both directions, and a site owner who rents out a footer position is exposed as well as the buyer. Google link penalties and recovery explains both.
Where sitewide links are still sold
Our research found that homepage and sitewide links are still actively sold in some non-English markets and in iGaming, and are reported to work in the short term. That evidence is low confidence: practitioner reports with no published dataset. We also found no verifiable 2026 source for rental prices, so we do not quote any. Both gaps matter if someone offers you such a link with a confident claim about results, because there is no public figure against which to test either the claim or the price.
Operators who rent such links tend to use them on sites they can afford to lose. For a licensed brand or any domain that must last, a rented sitewide link with a keyword anchor is one of the easier paid patterns to spot. That difference in position explains how the practice can be reported to work and still be a poor choice for a business whose domain must last. Someone running a site with a short planned life weighs a quick gain against a loss they have already accepted. A business with one domain it intends to keep is weighing the same gain against the footprint described above, sitting on its main asset for as long as the rental runs. Our lesson on white, grey and black hat tactics sets out how to weigh that risk.
What to do with the ones you have
Start with links pointing to you. Find them in a backlink tool by sorting referring domains by number of links. Any tool in our backlink tools comparison will show this, and the sitewide links stand out at once because a single domain accounts for hundreds or thousands of rows. Then sort them into three groups.
- Links from sister sites, partners and real credits with a branded anchor: leave them.
- Links you placed or paid for with a keyword anchor: ask for the anchor to be changed to your brand, for a nofollow to be added, or for the link to be moved to a single relevant page.
- Links you did not create and cannot explain: usually leave them. Google’s guidance limits the disavow tool to sites with a manual action or a considerable number of links they built or paid for. Backlink audits and toxic links covers when a disavow is justified.
Then turn to the links you give out. If you make themes, plugins, widgets or badges, or you build sites for clients, four habits keep the credit on the right side of the policy. Use your brand name as the anchor, not a keyword. Keep the link visible. Add rel="nofollow". And let the other party remove the credit, because Google’s policy also lists requiring a link as part of a contract without letting the other party qualify it. A credit handled this way still does what a signature is for: people who see the work can find out who made it and visit. What it gives up is the ranking credit, which the two reasons above suggest was worth little in a footer anyway.
Keep or accept a sitewide link when it would exist if search engines did not: a group of brands showing what else it owns, a sponsor being thanked, a maker signing its work. Use the brand as the anchor and label anything paid. That test settles most of the rows in the table without reference to the policy text, since a link that exists to tell readers something has a reason to be there whatever Google makes of it. Definitions of blogroll, widget link and related terms are in the glossary.
Common questions
Are footer links bad for SEO?
Not in themselves. A footer link between sister brands or to a genuine partner is normal. Footer links placed across many sites to pass ranking credit, especially with keyword anchors, are named in Google's spam policies.
Do sitewide links count as many backlinks?
Tools count one backlink per page, so a sitewide link can show as thousands. It is still one referring domain, and repeated links from one site add little after the first.
Should a "website by" credit link be nofollow?
That is the cautious choice. A design or development credit repeated across many client sites is a widely distributed template link. A branded anchor with nofollow keeps the credit and removes the risk.
Are widget links against Google's guidelines?
Only some. The policy names keyword-rich, hidden or low-quality links embedded in widgets. A visible, branded, nofollowed credit on a widget is the compliant version.
Should I disavow sitewide links pointing to my site?
Usually not. Google's guidance limits the disavow tool to sites with a manual action or many links they built or paid for. If you placed the links, change or remove them at the source first.

