How To Create a Mention Building Program

Michael 00:00 – 00:16

whatever you do build, it has to lean on your company's expertise. I can't say this enough. Generic content that doesn't say anything new and doesn't establish who you are won't do nearly as much for getting results as original thinking will.

Michael 00:16 – 00:34

Search used to be treated mostly as performance marketing. And AI search is really turning this into brand building.

Michael 00:34 – 01:04

Hey, welcome back to The Search Signal. I'm Michael Transon. I'm the founder and CEO of a search marketing agency called Victorious. And here at The Search Signal, our goal is to help marketers understand and also take action on what is happening in the quickly evolving world of search marketing. And we have been running campaigns ourselves for nearly 13 years. We also work with some of the fastest and the biggest growing brands out there. And we bring what we learn from those experiences, plus a lot of our own first-party research, to The Search Signal here to help you with your own work. So

Michael 01:04 – 01:32

AI assistants mostly decide whether to name your company based off of what other websites are going to say about you. And once marketing leaders accept that, they ask how to get more of those websites talking about them. Today's episode is going to answer that in detail. And we're going to talk about three programs that we recommend running. And then also I'll walk through how to set up each of those. But before we get into that, let's take some time this week talking about what happened in the world of search marketing news. So,

Michael 01:32 – 01:57

Google this week updated its guidance on AI generated content, and they did that on October 1st. So the page that they use for that recommendation is now saying that it's critical, that's their words, to manually fact-check and review all AI generated content for accuracy before you publish it. And that the review that you do should cover your title tags, your meta descriptions, your structured data.

Michael 01:57 – 02:48

And your image alt text on top of your body content. So the reason why I bring that up is a lot of teams right now are generating those other areas with AI, and they don't tend to look at them after that. But I would also say the next day, Google then added another section to that page, and that section was about helpful content. Google defines a page's main content in that and then lists four things that Google's quality raters are trained to look for, which are effort, originality, talent or skill, and accuracy. That same page now describes made-up author profiles as deception. And if you had not seen at Google Search Conference or Google Search Central Live event, it was in Barcelona, they had said that scaled content.

Michael 02:48 – 03:31

Is now actually becoming a bigger problem for Google than what has historically been its bigger problem, which is link spam. So plenty of people are reading this as a sign that there is a big update in the future that will be aimed at AI content. That could be true. However, Google's own FAQ on AI content, which that new page still links to, says that. the question is whether a content is helpful and original, regardless of how it's being produced. So my point is I don't have a problem with any team using AI to write a first draft. Somebody who knows the subject though needs to read every piece and check it before it goes live. I would give that job to a specific person for any type of specific content that you are publishing.

Michael 03:31 – 03:45

And I would also make sure you include the metadata and the schema that Google had specifically called out in this update. And I would make that person's review a required non-negotiable step before anything gets published on your website.

Michael 03:45 – 04:56

We talked about this in a prior episode for a news update, but coming back to it with a little more information. Google is paying some publishers for content that it's using in its AI answers. A website called The Information reported last week on a pilot with about a hundred publishers, where the payments are now working out to roughly a tenth of a percent of their ad revenue. There was one early participant I saw that is reportedly making over a million bucks a year doing this, which is great for larger companies. And some small sites have made, you know, less than a thousand bucks over several months. So Google seems to be deciding who gets paid right now based off of whether a page contributes significantly to generating the answer. A page that gets linked afterwards or that gets used to confirm a fact once the answer is already written does not qualify. So what that means is Google is separating the content that shapes what AI says from the content attached to the answer after the fact. And the content that shapes the answer is the content that has something original in it, information or like a POV that the model could not get from anywhere else.

Michael 04:56 – 05:22

Google has also started showing AI Overviews on most branded searches, meaning searches that are for a company's name. So I saw that DemandSphere, which is a company that tracks branded keywords, saw that AI Overviews existed on about a quarter of them at the start of September. But by the end of the month, it was at over 80% and they saw a peak of almost 90%.

Michael 05:22 – 06:15

Ahrefs looked at US desktop results and found that AI Overviews existed on about 83% of branded searches as of September 29th. And in about three quarters of those, the AI Overview sat at the very, very top of the page. Now Google has not announced anything about this, but both studies measured whether an AI Overview appeared. I don't think either one of them looked at what it said about the company itself. So, you know, for most of the history of search, someone who Googled your company typically and usually saw your own website first, but now they are seeing a summary of what Google wrote. And Ahrefs found that outside sources were cited most often in those summaries. And some of those sources included Wikipedia and also LinkedIn. Point being, other websites are now writing the first thing that people are seeing when they're searching our name or your name.

Michael 06:15 – 06:32

I would say if you haven't, go search your company's name this week, read the AI Overview, look at which sites it's pulling from because those sites' descriptions and descriptions of you are now sitting above your own homepage. Now that's the news for this week. Let's dive in.

Michael 06:32 – 08:24

So we've talked in a lot of detail on this podcast about how getting your company mentioned on other websites is the biggest driver of how you perform an AI search. Now that also needs a foundation underneath it, which is all the technical and the entity work that we've talked about before. But assuming that you have all of that in place, let's talk about the three ways that you can build a brand mentioned program. And this is based off of all the research that we've done and where we're actually seeing results. So the three programs are mention building, link building, and citation outreach. And each one of those sequentially is more targeted than the one before it. So let's talk briefly at a high level before we jump into the details. So mention building is the broadest. This is PR and brand building work. This is like press releases and speaking at events, podcast appearances, etc, etc and things like directory profiles. And its job is to raise the total number of pages that mention your company and to put your name next to your category on as many of those pages as possible. A lot of them sit on sites outside of your industry too, like local news. Link building is a narrower version of mention building that goes after specific websites that cover the kind of work that your company does. And each placement includes an actual backlink back to your website, and that carries a little bit more weight, from our research than just a mention alone. And then citation outreach is the most targeted. It goes after the exact pages right now that AI systems are already citing when your buyers are asking about your category. And it works to change what those pages say about you directly. So before we go into how to run each of these, I do want to spend a couple of minutes on why they work because that also will shape how you run them.

Michael 08:24 – 09:18

The reason why these programs work is when somebody asks ChatGPT or Google's AI Mode to recommend a company in your category, the answer is going to come from two different places. Part of it, we talked about this before, comes from what the model learned in training, which is an enormous amount of text from across the entire internet. The rest comes from a live search. And this is where the assistant reads a handful of pages and then pulls from them. Either way, your website is going to be just one voice among thousands. And the model is going to build its picture of your category mostly from everybody else's websites and pages. Ahrefs, for example, looked at about 75,000 brands to see what tracked most closely with getting mentioned in Google AI Overviews. And the strongest factor they found was how often the brand got mentioned across the web. And that came out actually ahead of backlinks, and it also came out ahead of domain rating.

Michael 09:18 – 09:29

So all three of these programs depend on something you have to build before any of them. And the first thing is expertise inside of your own company.

Michael 09:29 – 09:59

Getting mentioned more depends on your company actually having something to say that other companies do not have. So a real durable presence in AI search has to come from your own information and your own point of view. And these are things that help the people that are asking the AI questions. And spread as widely as you can on pages about your category and consistently and for a very long time. You need to think about this as your share of voice across the whole internet.

Michael 09:59 – 10:52

The three programs that we're going to talk about today handle the spreading of that content. But the information and the point of view have to come from inside of your own company. And the reason is because a company that is doing the same work as its competitors and saying all of the same things that the competitors say gives a mention program very little meaningful to send out. And more outreach and stuff will not fix that. So I want to give two practices on building out an expertise system in your business and to do it on purpose. The first one is one that we actually do at Victorious, and you've heard me talk about a lot, which is fund original research. And if you can, fund it on a schedule. This could be like a recurring study every quarter or every month or even twice a year on questions and information that nobody in your industry has good data on and gives your company something new to say for months.

Michael 10:52 – 11:18

And you can have the articles and the interviews that you do that follow all of it have that source as an opportunity for interviews and podcast episodes to point back to. It does not need to be like crazy large or expensive. You could do a survey of your customers. That would count. You could try to measure a question in your industry that everybody seems to argue about. And you can use data that you might even already have at your company.

Michael 11:18 – 13:07

And if you have a lot of questions that don't make it into the final episode or the final report, you can also use that as material for blog posts later. Then once we've got that, we can use the people, and I would recommend you use the people inside of your company who actually know things. Make a list of everybody in your business, that can speak with authority. And next to each of their names, you can have them write topics they can speak on, and you can use them to turn around quick quotes. Because if, for example, you're using outreach to journalists, they're gonna be on deadlines, they're gonna tend to use whoever answers first. I would recommend if you're building this expertise program, interview these people. Maybe interview them for 20 minutes, maybe once a month or something like that. Something that they have an opinion on. What they know that customers get wrong most often, what questions that they are answering, you know, every single week that neither us nor anybody else in the industry have published a good answer to. You could also ask them what most people in the industry believe that they think is wrong. We can record these conversations. You should record these conversations, and you can have more points of view, honestly, than you could probably publish. The point is, like when we're building these expert programs for these three mention building programs, most companies have a lot more to say than they think they have to say. So once your experts can get on record and they can have opinions, almost any news in your industry can become something that you can actually comment on. And every comment that you make is another chance to put your name next to your category. The other thing after building a expert program for your mention building program before running any of these programs, is to understand about co-occurrence and where your name needs to sit on the page for the program that you're building.

Michael 13:07 – 13:34

Okay, so let's talk about co-occurrence because no matter which program you choose to deploy, whether you are going with a broad reach mention building program, you're going link building, or you're doing citation outreach, co-occurrence is very important because the useful mentions out of all your mentions are the ones that put your company's name close to the words people use for your category, and ideally in the same sentence.

Michael 13:34 – 14:54

The industry term, like I just said for that, is something called co-occurrence. There's a company called Brand Mentions I saw. They do mention tracking software. They had looked at about 200 or 250 brands and found that mentions within a few words of the thing the brand wants to be known for tracked AI visibility much more closely than just the raw count of mentions did. About half of the mentions for the more visible brands also didn't even include a link at all. That study compares brands that already show up with brands that don't. So it shows a pattern, doesn't show the cause. But point being is when we are building these programs and we're preparing content for outreach and for inclusion in PR releases or for link building or reaching out to these citations that currently exist, our goal is to be able to write clear, plain sentences that say what your company is, uses the words a buyer would use for your category and put that exact sentence everywhere your company gets described. These would be things like speaker bios, directory profiles, author bios, the press releases themselves, or if you do a podcast like your show notes. Over time, every one of these pages that puts your name next to your category in the same words becomes an opportunity for the AI systems to understand what you actually do.

Michael 14:54 – 15:44

So once we have that foundation in place, taking a step back, those are two very important things before we build these programs. We need to focus on building a expert program where we have people in our business who can provide POV, can provide quotes, can provide perspective that is unique to your business. And then the second thing, once we recognize the value and importance of co-occurrence, that we can't just go out and get our site or our brand randomly mentioned in a bunch of publications, but it needs to be done in context of the types of entities, which are products, services, and value we bring to our buyers, as those two are the most important things that we have as foundation. We then can start to talk about our three programs. And we can start with mention building.

Michael 15:44 – 16:00

Okay, so mention building is wide. There's a lot of ways that you can do this. For today's episode, I'm just going to talk about a few examples. And these are ways in which you can build your mention presence across the web in very practical ways.

Michael 16:00 – 16:07

Most of mention building is brand building work that also leaves your name in writing on other people's websites.

Michael 16:07 – 17:41

So for one example, speaking at industry events. This puts your company and also puts your speaker on the event's website and in all of the coverage that happens afterwards. Another one is podcast guest spots. These really work well because for most podcasts, the show notes and the transcript get published and they put your name and your experts name in writing and usually it's going to be on a site about your industry. Another one is journalists oftentime post requests for like experts and expert sources on sites every day on services built exactly for that and answering those types of questions quickly with a very useful quote from somebody on your staff or your roster is going to help get your company named in articles about your category. Another one is contributor articles. So contributing in articles on publications in your industry that your buyers are reading, those come with an author bio, which is where we talked about earlier that one-sentence description of your company can go. There are also directory sites and listing sites in your industry that let you create and maintain a profile for free. And AI reads those profiles. So if we fill them out, we can fill out every field and we can keep them current. Also, I would also say LinkedIn totally counts too, right? We've seen this time and time again, and we talked about it in prior episodes. LinkedIn showed up among the outside sources that are cited most often in branded AI Overviews, and we talked about that in the news. So, posts there from your experts are pages that AI reads.

Michael 17:41 – 18:22

Another one that's been going back and forth on its value is Reddit, where if you're building a brand mention program, you can participate as a known employee who is there to help. There's a company called Octolens, which sells some social listening software, and they did an analysis of hundreds of millions of Reddit mentions of software companies and comments from people who disclosed where they worked from and they worked at the company, got more than twice the upvotes of anonymous promotional comments and helpful answers that actually solved someone's problem did way better than any like promotional posts that were disguised as questions.

Michael 18:22 – 19:16

So I just hit you with a couple just real fast ways in which you can deploy a brand mentioned program very practically. There's one that I want to spend a little bit more time talking about inside of these programs that I think has a lot of potential. I am skeptically optimistic. And that is press releases. So I think press releases are a great part and has a lot of potential part of a mention building program too, but they work a little differently than everything else that I've talked about on this list. So historically, a press release used to mean you had some big news to announce, like you had gotten a funding round or maybe you had made a big hire. Expert opinions and having that program can change that. So any development in your industry that one of your experts has a view on can actually become a press release.

Michael 19:16 – 19:36

A good press release runs probably a couple hundred words. It usually needs to be tied to something current. And then I would say we need to include links to at least one data point to its source and then also states our company's POV. Now a press release costs very, very little and it's quick. It can go out the same week you write it.

Michael 19:36 – 20:50

When a wire service ultimately syndicates the press release, which means it is distributing it to other sites that republish it, copies can actually appear on a lot of websites, including local news sites that also do not maybe have actually anything to do with your industry, but it increases the surface of brand mentions, and every copy is going to put your name next to whatever phrase that you had written it next to. So I'm optimistic about press releases. Like I said, I am a little skeptical at the same time. There was a company out there that I think is called Muck Rack, it sells PR software. They had studied a lot of links in AI answers, and they found that press releases show up almost entirely in answers to questions about industry trends and about three and a half times the rate of the best of answers. So a release can help you when somebody's asking AI about where your industry is headed and what's happening in the industry, it might help a little bit less than when someone is actually asking like which company that they need to hire. I don't know how long the effect is gonna last with press releases, and also press releases only spread as far as the other sites pick it up. And we do not control which websites pick up our press releases.

Michael 20:50 – 21:17

Okay, and then like the last part of mention building that I want to talk about is paying for placements, which is probably the most debated part of a mention building program right now. My position on it, just to state it, is that a sponsored mention on a relevant page, as far as we can tell, as of right now, counts the same as an earned one. Now, Google has said some things that do push against that.

Michael 21:17 – 21:43

And you should definitely read them. Google's guide to its AI search features says that seeking out inauthentic mentions across the web isn't as helpful as it might seem. And its spam policies do cover attempts to manipulate generative AI responses similarly to the traditional ones for SEO. And they've said for the record they compare buying mentions to buying links.

Michael 21:43 – 22:00

Now that's true. However, I will say, honestly speaking, we've seen a lot of studies online about how purchasing links can be an effective way to build an SEO campaign, and they're still efficacious even to this day.

Michael 22:00 – 22:26

So my position that I'm holding is because of actually just one word in Google's guidance, which is inauthentic. A sponsored article on an established publication that covers your industry about something that your company knows, with sponsorship disclosed or not, is an authentic mention. The model reads the page the same way it reads every other page on the site.

Michael 22:26 – 23:23

Fake activity fits Google's word. Like if you make up an account or you invent an author profile, or if you decided to plant reviews or sites that like literally only exist to sell mentions. And you should stay away from that. And then also one old rule that they've always had still applies too. When you pay for a placement, you might want to consider using the sponsored tag in that page's code and treating it so these spots see that it's an endorsement. So point being is sponsored content can be an absolutely viable way to build a mention building program. Now going to be more expensive because you are literally paying for the placement versus earning the placement. And they probably will be more expensive than other methods like press releases, which are probably more affordable than sponsored mentions. But sponsored content does have the potential if it's on the right website with the right co-occurring content to be very, very efficacious.

Michael 23:23 – 23:47

Okay, so that's mention building. Now the second program we'll talk about is link building. And this is a little bit narrower than mention building. This has been around forever. Companies that start investing in mentions often ask whether they still need link building, and I would say that they do. Okay, so a backlink, if you don't know, is a link from another website to yours. And in practice, it's like a mention with a link attached.

Michael 23:47 – 24:21

The AI reads your name on that page in the same way that it reads any other mention. And then Google also counts the link as a ranking signal for your site. So a mention that comes with a link from a relevant site does do a little bit more work for outcomes than just a mention alone. Google has used links as a core ranking signal for decades. And we do know that Google's AI features draw from Google's own search index. So the thesis is the links that help your page rank in Google should also shape what Google's AI has available to cite.

Michael 24:21 – 24:58

Now, links on their own aren't the strongest signal for AI. I talked about the Ahrefs study earlier, but that same study found that mentions tracked AI Overview visibility more closely than backlinks did. So link building runs alongside mention building. Mention building is going to give us the volume. And then link building adds weight on the sites that are closest to your industry. So link building is going to target specific websites that cover your kind of work. There are a lot of ways to earn those links. You can do contributions to articles to industry sites.

Michael 24:58 – 25:19

You can get listed on resource pages, like we talked about with the expert program publishing data on that other sites want to reference. You can also offer your page as a replacement when a site links to something or a URL that no longer exists. When you do this though, every link needs to go through, in the way that we do at Victorious, a four-level review before we actually count it.

Michael 25:19 – 25:43

We need to check the page that the link sits on, the website as a whole, and we need to check whether the site is relevant to the topic and whether it's safe for the brand. And there are two really important rules. Number one is relevance definitely in link building beats raw authority. So a smaller site around your exact topic usually will help more than a big site about something completely irrelevant.

Michael 25:43 – 25:53

And a link from a site that has never linked to you is going to count far more than a link from a site that already has linked you.

Michael 25:53 – 26:49

So that's link building as part of your brand mention program. That's the second program that we recommend. And that has been pretty consistent. If you've been running any meaningful SEO program at your business, I would say continue to do that. Now let's talk about the third program. And this is the most targeted one. And we call this citation outreach. So when we talked about this earlier, when an AI assistant is answering a question, it often is going to show the pages that it pulled from as sources. And those sources are called citations. And they tell you exactly which pages that the AI is actually reading for its response. Citation outreach goes after those pages directly and works to get your company added to them. So it starts with the questions your buyers are asking AI when they are choosing a company like yours. We've talked about this before, but we want to track maybe let's start with 20 or 50 of them.

Michael 26:49 – 27:12

In the words that your buyers would use, like who are the best providers in your space, or which option fits a company of their certain size, what pricing is, all those things. If you use a tracking software for AI, you can use this. But if you don't, you can just manually run those questions through those AI platforms that your buyers use. Like ChatGPT, AI Mode. And then we look at each answer along with the pages that it cites.

Michael 27:12 – 27:39

Again, software that tracks AI answers can run the questions on a schedule, which is the best case scenario, and log the citations for you, but a spreadsheet also works too. It's just gonna take a lot longer. Once we've got that, then we sort the cited pages and pull out the ones that mention your competitors and then leave you out. You're mostly gonna find four kinds of citations on review and directory websites. The fix for this is usually going to be just creating or completing your own profile that might not be there yet.

Michael 27:39 – 28:27

You also might find best of lists and some comparison articles. And we can reach out to the website and ask the author to add to you. And then you also might see forum threads where someone from your company can jump in, say where they work, that they're on staff, and then can add context or a helpful answer. And then the fourth thing would be news articles. And it's pretty much impossible to get a news article edited or updated with your POV. So the more realistic goal on this is getting into the next story that the reporter writes on that topic since reporters are again, they're rarely going to go back and edit pages. So establishing those relationships, reaching out to those authors, providing your experts as a POV.

Michael 28:27 – 29:44

Now, depending how many prompts you're tracking, you might find a lot of citations and it could feel overwhelming. You might be tracking hundreds of prompts and seeing thousands of citations. And these citations can change week over week. So we want to make it digestible. So my recommendation if you're just starting, every month take just the 50 pages that get cited most often in answers for prompts that we are tracking where a competitor is going to appear and you don't. We then find the right contact for each one of those websites, usually the author or the site's editor. We then can just send an email, really a short note explaining who you are, why your company belongs on the page. And then if possible, include like an entry already written and it matches the length and the format of some of the other entries, if it's like a best of list. So the editor could just pick it up and paste it, not have to do any work. And again, we've talked about this before, but oftentimes they're gonna be asking for some sort of economic consideration. And so you need to recognize if you're going to build a citation outreach strategy, we have to have budget associated for this. And we can just follow up week after week. And then if you don't hear responses, just move on. Most of these emails will probably go unanswered.

Michael 29:44 – 29:48

But that is also why the program has to run every single month.

Michael 29:48 – 30:42

So taking a step back, these are the three programs that need to run at the same time. And right now, based off of everything that we know, they are the best ways we know how to influence how your company is actually showing up in AI search. And here's the truth: like all of them are gonna take budget, and all of them are gonna take a lot of time. So you either need a team inside of your company or an outside team that is going to be running these for you. And there are a lot of tools that can help with parts of this. And you're gonna find a lot of them online if you do your research, but none of them are going to effectively replace human judgment about what your company should say and also where it should say it. And whatever you do build, it has to lean on your company's expertise. I can't say this enough. Generic content that doesn't say anything new and doesn't establish who you are won't do nearly as much for getting results as original thinking will.

Michael 30:42 – 31:05

Search used to be treated mostly as performance marketing. And AI search is really turning this into brand building. And there are also a lot of other ways to get your company onto websites that we did not get to today because there's honestly a lot that can be done here. It is a really exciting time in AI search because businesses have a chance to build authority right now, but doing it well is going to take time and it's going to take resources.

Michael 31:05 – 31:16

So thank you for listening to this week's episode of The Search Signal, where we talked about how to build a brand mentioned program. If this was useful for you, please go ahead and subscribe so the next episode finds you. And I will see you next week.