Why Brand Mentions Are Starting to Beat Backlinks

A Houston client added six new links in March and still dropped four spots. Here is what actually moved the needle instead.

In March 2026, a garage door client of mine added six new backlinks in a single month. Solid ones too, from local trade directories and one guest post on a mid tier home services blog. The site still dropped four positions on its main term. Nothing technical had changed. Page speed was fine. The content was still fresh from a February update. So I pulled up Search Console and started looking at what else was different that month.

What I found was a stack of mentions. The client’s business name had shown up in a Reddit thread about garage door brands in Houston, in a local news roundup of small businesses to watch, and in two YouTube reviews. None of those linked back. Not one. But every single one used the brand name in a real sentence, next to real context about garage doors and Houston.

I checked the timing against Google’s own status updates and this lined up with a broader core update rolling out that same week. Other clients in the same account portfolio, running similar link building pace, held steady or gained. The garage door client was the only one that lost ground, and the only real difference I could find was that stack of mentions sitting in places I had never thought to track before.

That was the moment I stopped treating backlinks as the whole game.

The habit that is hard to break

For twenty years the industry ran on one instruction: get more links. Referring domains, Domain Rating, link velocity, these were the numbers every strategy deck lived on. I built entire client reports around them. A backlink was proof that someone else vouched for you, and Google counted the vouches.

That model still works. It just is not the only model anymore.

Google’s own patent filings from over a decade ago described something called an implied link, a reference to a brand or resource that carries weight even without a clickable link attached. Search Engine Journal has pointed out that the patent language actually leans toward branded search behavior rather than plain on page mentions, so the exact mechanism is still debated among practitioners. What is not debated is the outcome I keep seeing in client accounts: brands that get talked about, linked or not, are showing up more in both organic results and AI answers.

Why AI search changed the math

Backlinks were built for a web made of pages linking to pages. AI answer engines do not work that way. When someone asks ChatGPT or Google’s AI Overview for a garage door repair company in Houston, the system is not crawling a chain of links to decide who ranks first. It is pulling from a wide pool of text where a brand name shows up, checking whether that brand keeps appearing in consistent, credible contexts, and picking two or three names to answer with.

I tested this directly for three home service clients over an eleven week stretch this spring. I ran the same ten prompts across ChatGPT, Perplexity, and Google’s AI Overview each week and logged which brands got named. The client with the most Reddit and local news mentions, despite having the weakest backlink profile of the three, showed up in AI answers almost twice as often as the other two combined.

Every result went into a plain spreadsheet, nothing fancy, just the prompt, the date, and which brands got named in the answer. By week six the pattern was obvious enough that I stopped needing the spreadsheet to notice it. The client with the strongest backlink profile barely showed up in AI Overview at all, even though it still held a solid position three ranking on Google’s regular results. The two systems were clearly reading different signals.

Search Engine Land’s own reporting on this backs it up with a wider lens. A mention without a link still tells search systems the brand exists and gets talked about, even though it does not carry the same technical authority transfer a backlink does. That distinction matters more now that so many answers skip the click entirely.

What a mention actually does that a link does not

A backlink passes a technical signal. It tells Google’s crawler that page A trusts page B enough to send readers there. That is useful, but it is also gameable. Anyone can buy a link on a low quality blog network. It takes real work to get five independent sources talking about the same brand in the same week.

A mention works differently. It shows up in a Reddit comment because someone genuinely used the product. It shows up in a local news roundup because a reporter thought the business was worth naming. It shows up in a YouTube review because a creator picked it to compare against three others. None of that is easy to fake at scale, which is exactly why search engines and AI systems have started leaning on it.

Search Engine Land has been writing about this same shift lately, describing visibility as something that now depends on how often a brand gets talked about and trusted across sources, not just on how many links it collects. I have watched that play out in three separate niches this year, from Ayurvedic wellness to HVAC, and the pattern holds every time.

A small experiment worth running

Here is something I now do with every new client in the first month. I set up branded search alerts for the business name, the owner’s name if they are public facing, and any product names. Then I spend twenty minutes a week checking three things: where the brand got mentioned, whether the context was positive, and whether the mention sat next to language that matches what the business actually wants to be known for.

For one Kerala based Ayurvedic hospital client, this turned up eleven unlinked mentions across health forums and one regional news site in a single quarter. None were backlinks. All eleven used consistent language around Panchakarma treatment, which happened to be the exact phrase the client wanted associated with their name. Three months later, that phrase started pulling an AI Overview citation that had never appeared before.

That is not proof of a direct cause. I want to be honest about that. But it lines up with what practitioners across the industry keep reporting, and it costs almost nothing to track.

I ran the same setup for a UK based wellness client in April. Fewer mentions this time, just four across the quarter, but two came from a well known health forum and one from a podcast transcript that got indexed. The client’s branded search volume in Search Console rose by roughly a third over the following eight weeks. Branded search volume is not the same thing as rankings, but it is the clearest early signal I have found that a mention actually landed with real people, not just with a crawler.

Where backlinks still win

I am not telling anyone to stop building links. A single link from a genuinely relevant, high authority site still does something a mention cannot. It passes actual equity to a specific page and can move that exact page for a specific term. If you need a product page to rank for a competitive keyword next quarter, a backlink campaign is still the faster lever.

Backlinks still matter, and reviewing what makes one strong versus weak is worth ten minutes if it has been a while since you checked.

The shift is not backlinks versus mentions. It is that mentions used to be treated as a nice side effect of good PR, and now they carry weight on their own. Treating them as a real line item, tracked and reported the same way you track referring domains, changes how a strategy gets built.

What this means for your next report

If your monthly client report still leads with new referring domains and nothing else, it is missing half the picture in 2026. Add a mentions section. Pull Google Alerts or a mention tracking tool, log where the brand showed up without a link, and note whether AI Overview or ChatGPT started citing the brand for any of the same terms. It takes maybe thirty minutes a month once the alerts are set up.

Getting a brand discoverable inside AI answers takes a slightly different shape than classic link building, and it is worth understanding the mechanics before you build a plan around it.

I still open Ahrefs every week and check referring domains for every client. That habit is not going anywhere. But the first thing I check now is whether the brand got talked about somewhere new, link or no link. Six months ago that would have been an afterthought. Now it is the first line of the report.

Try this on one account this week. Set three brand alerts. Check them for a month. You will probably find mentions you did not know existed, and at least one of them will be worth a direct message asking for a link. The rest still count for something, even without one.