Client, Partner and Brand Logos on Your Site: Should You Link Them?

Not by default. A logo that links to a client's, a partner's or a brand's site neither helps nor hurts your ranking: Google has said so since 2016. The only real SEO risk is a network that links back and forth with itself. The sales cost of a logo bar has never actually been measured.

On this page

Includes a Falia study of 203 Canadian companies

What Google writes, and what its spokespeople have repeated

The fear of "losing link juice" by pointing outward is a shortcut inherited from the 2000s. It still circulates because it sounds technical, and it costs design decisions made to protect something that doesn't exist. It rests on no Google document.

The starting point is a public Q&A session from January 2016. John Mueller, Search Advocate at Google, answered there that links from your site to other sites are not specifically a ranking factor, that they can add value for your content, and that whether you mark them nofollow or not "doesn't really matter." In December 2019, asked about studies correlating outbound links with better rankings, he answered with one word: "No." In July 2023, on Reddit, he put it another way: nothing happens, and naming a dictionary doesn't fix your spelling mistakes. His rule fits in one question: does this link provide real value to the reader? If yes, link normally. If not, don't.

The corollary is confirmed in another session, in July 2019: there is no ranking advantage to marking all your outbound links nofollow, and a site that uses normal links doesn't rank worse than a site that blocks them all. Linking doesn't hurt. Linking doesn't help. It's neutral, and that's the only defensible answer on the ranking side. So the decision gets made elsewhere: on sales, on network risk, and on accessibility.

The least-known point concerns nofollow, and it's worth reading twice because it undoes the protection many people think it gives them. In June 2009, Matt Cutts explained that a page with ten links, five of them nofollow, no longer concentrates its value on the five remaining links: each link gets a tenth, and the nofollow links' share simply isn't passed on. Putting nofollow everywhere doesn't keep anything for you, then. It only takes something away from the destination site, often that of a client you wanted to help. Since March 2020, Google also treats the attribute as a hint it can follow or ignore. Nofollow isn't a wall anymore. It's a label.

The two directions of the link, almost always confused

Most discussions of this topic mix up two links that don't share the same effect, the same risk, or the same owner.

One special case deserves a note: the link you place yourself on someone else's site. The certification badge you provide to your dealer network, the "Powered by" badge for your tool displayed on customer sites, the credit an affiliate program requires. In March 2015, John Mueller said he'd "definitely" put a nofollow on these links, not because they're bad, but because they weren't placed by the site owner editorially. Google's spam policies have since named "links widely distributed in the footers or templates of various sites" and links embedded in widgets among examples of link spam, and note that they're acceptable when they carry rel="nofollow" or rel="sponsored". If you provide a badge or embed code to your partners, that's where the attribute belongs, not on your own logo bar.

The one real risk: a network that links back and forth with itself

A reciprocal link isn't a violation. What becomes a signal is the same exchange reproduced mechanically across dozens of sites that belong to the same commercial network.

Google's spam policies prohibit "excessive link exchanges (link to me and I'll link to you) or partner pages created exclusively for cross-linking." The word that matters is "excessive," and the example that matters is "partner pages." In January 2020, John Mueller wrote that reciprocal links aren't necessarily bad, and that they're fairly obvious to Google's systems. In January 2022, he added that it's "perfectly fine" and even natural, with the example of a local business that links its neighbors, and one caveat: as long as you're not making strange arrangements behind the scenes.

The data confirms that reciprocity is the norm on the web. In 2019, Ahrefs analyzed 140,592 domains getting at least 10,000 monthly visits from Google. 73.6% of them had reciprocal links with at least one other site, and 43.7% of the sample's 112,440 top-ranking pages did too. A site with zero reciprocal links would be the anomaly. It's a vendor study, run on its own index, but the order of magnitude is enough to rule out the idea that a single mirror link is a problem.

A second test, this one done by hand, nuances the first. We checked 30 specific pairs: a company displays the logo of a client, a partner or a vendor, does the target link back? The link back existed in 5 of 30 cases, or 17%, and one of those five isn't even a genuine reciprocal partnership: Mercer displays Formula 1's logo pointing to Marsh, its own parent company, not to an independent partner. Real reciprocity, logo for logo, runs closer to 13% than to the 73.6% cited above. What Ahrefs measures is reciprocal links anywhere on a domain; what a logo wall creates is a narrower case, and in an isolated instance, a fairly rare one.

The scenario that carries real risk is narrower, and it shows up in three types of networks you'll find everywhere.

In all three cases, the risk doesn't come from the reciprocal link itself. It comes from the volume, the systematic pattern, and the lack of content around the link, three things you control.

Does an outbound link really reduce conversion rate?

This is the argument you hear most often against clickable logos: every exit costs sales. It deserves more than a bare claim, because it's used to justify design decisions on pages that generate revenue. We went looking for the published tests. The result is more nuanced than what conversion guides tend to say, and it changes the price you should be willing to pay for this rule.

What's documented concerns single-goal landing pages. In 2014, HubSpot tested five of its high-traffic pages, with and without navigation or outbound links. On top-of-funnel pages, the gain was 0% to 4%. On mid-funnel pages, 16% and 28%. So the same test contains both the gain everyone cites and the zero nobody cites. In 2011, Yuppiechef doubled the conversion rate of a wedding registry page, from 3% to 6%, by removing the navigation menu. In 2014, MarketingExperiments removed navigation from a checkout process and measured 10% more completed orders and 19.95% more revenue per visit. And in 2018, Unbounce analyzed 18,639 landing pages: those with a single link converted at 13.5% on average, versus 10.5% for those with several.

These numbers carry three limits that need naming before you decide anything. None of these tests publishes its sample size or confidence interval, and all of them are published by a tool vendor or an agency. Flat tests never get published, which inflates the averages. And above all, they all deal with internal navigation links removed from a page that has one job to do and gets paid traffic. That's the case for a landing page, not a homepage, which by definition serves several intents at once: the buyer ready to purchase, the one comparing options, the one checking a price, the one confirming you exist.

The only published test dealing with an element linked outward is Taloon's, a Finnish hardware retailer. Removing the Facebook, Google+ and Pinterest sharing buttons on its product pages raised "Add to cart" clicks by 11.92%, at 95% confidence. But VWO, which published the case, offers two possible explanations: distraction, and the fact that almost every product showed zero shares, which is negative social proof. The outbound link is only one of two possible causes. No published test, anywhere, compares a clickable logo bar to a static one on conversion rate or sales. And no analytics study measures the share of visitors who click outward, then come back and convert. The number you're looking for, sales lost per link, doesn't exist.

There's even data pointing the other way. The Baymard Institute recommends always linking to external sources of reviews, ratings and awards, because its usability tests show that a non-clickable third-party rating pushes some participants to leave the site to go verify it themselves, with one participant calling the unlinked rating a "dead link." The outbound link, in that case, reduces abandonment and protects the sale. Stanford's web credibility guidelines, built from research with over 4,500 people, put easy verifiability of information first, with this note: even if people don't follow the links, you've shown confidence in your material. And Jakob Nielsen wrote as early as 1999 that not being afraid to link to other sites is a sign of confidence, and that third-party sites are far more credible than anything you can say about yourself. As for the "choice overload" argument used to justify a one-link-per-page rule, a meta-analysis of 50 experiments and 5,036 participants published in 2010 in the Journal of Consumer Research finds an average effect of zero.

What the reading data adds is that the logo bar gets seen. In 2018, the Nielsen Norman Group measured, across 120 participants and over 130,000 eye fixations, that 57% of a page's reading time happens above the fold and 74% within the first two screens. Contentsquare's 2023 Digital Experience Benchmark, drawn from over 35 billion sessions, puts the average scroll rate at 50.2%. A bar placed right under the hero sits in the most-watched zone of your site, the one where the prospect decides whether to keep going. That's not an argument for or against the link. It's the reason the decision deserves to be made deliberately rather than left to the template.

The technical details almost everyone gets wrong

Four attributes show up on logo links. Two change nothing about ranking, one is now implicit, and the last one cancels out the gift you think you're giving.

Accessibility, where a clickable logo is really judged

A clickable logo is a link whose only content is an image, and the customer using it has the right to know where it leads. Accessibility rules are precise on this case, and they're rarely applied to logo bars, even though part of your customer base depends on a screen reader or a keyboard.

WCAG technique H30 establishes that when an image is a link's only content, its alternative text describes the link's function, not the image. The W3C's tutorial on functional images gives the example of a logo that leads to the homepage: the alt text is "Home," not "logo." For a clickable logo, the accessible name should therefore say where the link goes. For a non-clickable logo, the company name is enough, and the word "logo" has no place in the alt attribute.

If the link opens a new tab, technique G200 asks that you only do this when necessary, and technique G201 asks that you warn the user before the context change. In practice, the accessible name becomes "Client name, opens in a new tab," through visually hidden text or the alt attribute itself. The screen reader then announces the destination and the window change before the click. Going back becomes impossible in a new tab, and that's exactly what these techniques try to compensate for the customer who can't see the screen.

Target size matters too. WCAG 2.2 sets a minimum of 24 by 24 CSS pixels at level AA (criterion 2.5.8) and 44 by 44 at level AAA (criterion 2.5.5). Android's guidelines call for 48 by 48 dp, and Apple's default control size in its guidelines is 44 by 44 points. A logo bar squeezed on mobile, where each logo is 18 pixels tall, produces targets that are impossible to tap precisely, and customers who give up without you ever seeing it. Finally, an auto-scrolling animation that runs for more than five seconds and shows alongside other content must offer a way to pause, stop or hide it (criterion 2.2.2, level A), and should respect the system's reduced-motion preference. The Nielsen Norman Group's 2018 study on banner blindness adds another reason to skip the motion: animation is one of the visual treatments users associate with advertising, and skip past. The obligations that apply in Quebec are covered in our article on web accessibility obligations for private businesses.

What 203 B2B companies in Canada do, a study conducted by Falia in September 2026

The data usually cited on this topic, ours included until now, comes from a handful of American software giants. Useful to illustrate a principle, but not very representative of a B2B company here, whose homepage was often built once by a designer rather than maintained by a marketing team. So we looked wider, closer to home, and logo by logo.

The behavior splits into four groups, but not at random. Martinrea, an auto-parts manufacturer based in Ontario, links its 45 client logos straight to Audi, BMW, Aston Martin, Caterpillar and 41 other automakers, no exceptions. On the other end, six well-known ad and marketing agencies here (Bleublancrouge, Cossette, Kabane, Republik, Substance, Tink) display 97 client logos combined without a single one clickable. Same restraint at several recognized B2B software vendors: Coveo, Local Logic, Lightspeed Commerce, Kinaxis, Dialogue, Achievers and Workleap link none of theirs either.

A wall that includes partner logos is nearly twice as likely to be entirely clickable as a client-only wall, and close to four times less likely to have no links at all. That confirms what this article already recommends: a partner relationship almost always carries a contractual obligation or a badge to display, which forces the decision. A client logo, on the other hand, often just gets copied into a template once, with no one coming back to decide anything.

The factor that matters most, though, is neither the sector nor the logo category. It's where the logo lives. Across 1,342 logos whose page carried an identifiable type, the click-through rate climbs from 32.6% on a homepage to 74.2% on a dedicated "our clients" page, 88.4% on a "partners" page, and 100% on a case-study page.

The simplest explanation is also the most likely one. A dedicated page is a decision made once, by someone building a list on purpose. The homepage bar, by contrast, is often copied from a template or added by a designer with no clear mandate on what the click should actually do. That's exactly the gap the hierarchy of solutions below is meant to close.

CGI, headquartered in Montreal, illustrates the first two solutions in that hierarchy on a single page, by itself. Its 14 most strategic technology partners (AWS, Microsoft, SAP, Salesforce, ServiceNow, Snowflake, Databricks, Google Cloud, among others) each get their own dedicated CGI page for the alliance, never a link to the partner's site. Its other 128 partners, less central, are simply listed with a direct link to their external site. BDC goes further: its link to Futurpreneur Canada leads to a page Futurpreneur built itself for the relationship, addressed "bdc-web-referral."

B2B software is the most cautious sector: 38 of the 87 companies with a wall come from it, and 17 of those link no logo at all. Manufacturers, financial institutions and economic development organizations link more readily, often because the relationship they display is itself contractual: a distribution program, a guaranteed loan, a membership.

This study has limits worth naming before you use it. It's a one-time snapshot, not a tracked trend. Status stayed undetermined for 195 logos, most of them behind dynamic loading that a page read can't follow all the way through. And a logo's presence says nothing about the agreement behind it: there's no way to know from the outside whether every clickable logo carries written authorization or a publicity clause, the question raised earlier in this article.

See the 203 companies reviewed, by sector

Aerospace (3) — CAE, Héroux-Devtek, Mecachrome.

Artificial intelligence (1) — IVADO Labs.

B2B distribution (15) — CDW Canada, Deschênes Group, Emco Corporation, Finning, Fortier Auto, Girardin, Guillevin International, Hamster, Lumen, Novexco, Russel Metals, Toromont, Wajax, Wolseley Canada, Wurth Canada.

B2B financing (5) — Banque Nationale, BDC, Desjardins Entreprises, Exportation et developpement Canada, Fondaction.

B2B fintech (2) — Flinks, Nuvei.

B2B services (1) — Solotech.

B2B software (38) — Absorb LMS, ACCEO Solutions, Achievers, AlayaCare, Benevity, Certn, Clio, Coconut Software, Copperleaf, Coveo, Dayforce, Dialogue, Docebo, Dye and Durham, Hootsuite, Jane App, Kinaxis, Klue, Later, Lightspeed Commerce, Local Logic, Logibec, Nethris, Novisto, OpenText, Petal, Plusgrade, Poka, Purkinje, Rewind, Shopify, Thinkific, Trulioo, Unbounce, Vena Solutions, Vidyard, Visier, Workleap.

Construction (10) — Aecon, Bird Construction, Broccolini, EBC, Graham, Groupe Montoni, Ledcor, Magil Construction, PCL Construction, Pomerleau.

Consulting services (2) — Levio, Talsom.

Cybersecurity (4) — Difenda, Herjavec Group, Mondata, Okiok.

Economic development organization (12) — Aéro Montréal, Centech, Chambre de commerce du Montréal métropolitain, Groupe CTT, Investissement Québec, Manufacturiers et Exportateurs du Québec, Nergica, Prompt, Propulsion Québec, Scale AI, TechnoCompétences, Zu.

Energy (5) — Boralex, Energir, Hydro-Québec, Innergex, Kruger Energy.

Engineering consulting (10) — AtkinsRealis, BBA, CIMA+, Englobe, FNX-INNOV, GBi, Hatch, Norda Stelo, Stantec, WSP.

Industrial technology (3) — FLO, LeddarTech, Optel Group.

Insurance (5) — Beneva, Beneva La Capitale, BFL Canada, Croix Bleue du Québec, Lussier Dale Parizeau.

IT and consulting (1) — CGI.

IT services (18) — Alithya, Compugen, Createch, ESI Technologies, Groupe Access, Groupe SL, ITCloud.ca, Micro Logic, Momentum Technologies, MS Solutions, Nventive, Osedea, Present, Sherweb, Softchoice, Spiria, Victrix, Vooban.

Manufacturing (24) — Artopex, ATS Corporation, Bouty, Béton Provincial, Cascades, Celestica, Chantiers Chibougamau, Groupe Canam, Groupe Lacasse, Groupe Lebel, Groupe Soucy, Kruger, Linamar, Lion Electric, Magna International, Manac, Martinrea, Premier Tech, Prevost, Prolam, Rotobec, Soprema Canada, Uniboard, Velan.

Marketing agency (15) — Absolunet, Adviso, Akufen, Bleublancrouge, Cossette, Deux Huit Huit, Globalia, Kabane, lg2, Locomotive, Republik, Sid Lee, Substance, Tink, Uzik.

Professional services (21) — Adecco Canada, BCF avocats d'affaires, BDO Canada, Bédard Ressources, Demers Beaulne, Fasken, FBL, Fed Group, Groupe Deloitte Canada, Langlois avocats, Lavery, Mallette, Mercer Canada, MNP, Normandin Beaudry, PSB Boisjoli, Randstad Canada, Raymond Chabot Grant Thornton, Richter, ROBIC, Therrien Couture Joli-Coeur.

Research services (1) — Potloc.

Telecommunications (1) — Telus Business.

Transportation and logistics (6) — Bison Transport, Day and Ross, Groupe Guilbault, Groupe Robert, Purolator, TFI International.

The legal question: the logo and the link aren't the same problem

Two separate questions hide behind the logo bar: do you have the right to display the logo, and do you have the right to link. In Canada, the second one is settled. The first depends on your agreement.

On the link, the Supreme Court of Canada ruled in 2011 in Crookes v. Newton: a hyperlink, by itself, never constitutes publication of the content it points to. The majority wrote that hyperlinks are essentially references, neutral as to content, and called them the web's "synapses." Quebec courts applied this reasoning as early as 2012. Linking to a client's, partner's or brand's public site therefore requires no permission. The caution applies to what you write around the link, never to the link itself.

On the logo, the Trademarks Act gives the owner of a registered trademark the exclusive right to use it, and section 22 prohibits anyone from using another's trademark in a way that's likely to depreciate the value of the goodwill attached to it. Displaying the logo of a real client, or of a brand you actually distribute, in a clearly labeled section, doesn't use their mark to distinguish your own products or services, and it's hard to see depreciation in that. The real risk isn't a lawsuit. It's an email from the other company's marketing department asking for removal, because the agreement didn't cover it, because the client is gone or the relationship ended, or because the company doesn't want to be publicly associated with you. Distributed brands often have precise usage guidelines for their retailers, and partner programs impose their own badges. For clients, common practice is to include a publicity clause in the commercial agreement: the right to name the client and display their logo, which channels it covers, respect for their visual identity, removal on simple written notice, and a statement that the display implies no endorsement. This article describes the framework and doesn't replace legal advice.

Quebec's Law 25 doesn't come into play. A click on a logo passes your domain name to the linked site as a referrer, nothing more, and no personal information is involved.

The hierarchy of solutions, from best to simplest

The question doesn't call for a yes or no. It calls for ranking three options, based on what you actually have to show behind each logo.

The bar doesn't only hold customer logos. Partners, distributed brands, sponsors, press mentions and certifications don't follow the same rules or carry the same risk, and mixing them into one unlabeled bar muddies the proof.

The checklist before you publish

The role links play within your own site is covered in our article on internal linking, and where proof fits into Google's quality criteria in what E-E-A-T actually covers. The mechanics of social proof in the buying decision, the logic of conversion rate optimization, and the homepage that sells in designing for three kinds of readers round out this reading list.

The general framework for search is laid out in our complete guide for SMBs and large companies.

Frequently asked questions about clickable logos and outbound links