Siege Media

CRO is GEO. With CRO, we're thinking about bottom funnel messaging. We're trying to convince users to buy our product. So at the same time, we're effectively trying to convince LLMs to buy (and therefore recommend) our product, too. To do that, we should hit LLMs in the face with the things that would make users buy and not rely on off-site signals. By doing so, seemingly, we may also make LLMs recommend us. Some common examples I've been seeing more recently: Say you apply for awards, or maybe you're top in category on G2. It's common to use these across award things on your site as trust symbols. But don't make it hard for LLMs to find that context on your site, and ideally put it on your homepage with crawlable text. They may find it themselves, sure, but make it easy on your own site by having it front and center. Have the same review stars. You could say your client retention, your NPS scores, your case studies, the quotes that impress prospectives. This is common across a "credibility" block I'll often see ~3-4ths of the way down. All of those features are CRO components traditionally, but they're also making it very easy for LLMs to recommend you as well. Don't just focus on the off-site review signals, bring those forward so LLMs never have to work to find them in order to win.

Excited to be hitting the "road" for a book audio/IRL tour, commencing now! A few of the exciting pods I've either already taken part in or will be publishing soon or are already live, with topics for each: Already live: - Edward Sturm's SEO pod, topic: "The page type that drives the most AI traffic" - Unscripted SEO pod w/ Jeremy Rivera: "GEO is CRO and the Versus-Page Signal" Live soon: - The GEO Blueprint Podcast w/ George Vlasyev: "YouTube GEO Best Practices" - Startups for the Rest of Us w/ Rob Walling: "SaaS GEO Best Practices" - The Best SEO Pod w/ Matthew Bertram: "Listacles, Schema and Conversion Tracking" Recording soon: - BuzzStream Podcast w/ Vince Nero: "Digital PR's Impact on GEO" - Near Memo w/ David Mihm, Greg Sterling, Mike Blumenthal: "GEO for Local" - The Majestic SEO Podcast w/ David Bain: "Actionable Off-Site GEO Tactics" - Real Talk With Sean w/ Sean Barber: TBD - The SEO Video Show w/ Paul Andre de Vera: TBD On the conference front, I'm excited to be speaking at Ahrefs Evolve on "PR For Machines: Earning Coverage That LLMs Repeat" on October 13th. And at Tech SEO Connect over November 5th and 6th, attendees will get access to the book/I'll be available for signing (or questions) as well. Definitely will get sick of my own voice/thoughts, but gotta get out there. If you have a relevant podcast you think I might be a fit for, I'd love to hear from you. Cheers!

"If the [TOFU traffic] quality is great, no matter how bad you are at everything else, you win. If the [TOFU traffic] quality is bad, no matter how good you are at everything else, you lose." “Not all of your TOFU traffic is high quality, right?” “It'll hit a landing page, and then what we do is we A/B test on things like landing pages and the homepage. And through the A/B test, we'll just figure out what works best for this traffic. And that's how we'll improve the conversion rate anyway.” “Because when we go get more traffic in SEO. Let's say all that new traffic is garbage. Then what is going through your A/B test?” “So when we improve the conversion rate, we are converting garbage. We're convincing more garbage to continue.” “The customers that are not going to buy are what we're optimizing for.” Jason Cohen founded two unicorns, most recently WP Engine. In our new podcast, Jason diagnosed why SEO programs generating TOFU content that's off-topic can be truly challenging for the organization beyond just its ability to convert. One of the biggest ones -- that it leads to downstream optimizations that lead us to optimize for the wrong customers.

The LLM few are considering, but many should be: Reddit’s. Reddit just announced their default UI will start testing an “Ask AI” inclusion as a nav option. In addition, searches already include the equivalent of AI overviews in their results. Unlike Gemini and ChatGPT, there are no external websites linked. In my opinion, this is *the* reason it’s flying under the radar so far. But brands are still mentioned in droves, curating from existing Reddit content to make the recommendations. Quietly, Reddit is building towards a variant of ChatGPT and Google. It already is the fifth most popular website online. This continues to build the story: yes, citations on ChatGPT might have dropped. But if your audience is on Reddit, who cares? Their influence will continue on, and if you’re not considering and aiming to impact that influence, you may eventually be left behind.

One of the most underrated GEO factors of any action: halo effect. When you create a blog post on “best X” or a product page that manages to rank that your own brand shows up for, the power of ranking is of course helpful for generating additional visibility. But for a newer brand, that’s only the start. The halo effect of other creators finding you is arguably larger than the original action in isolation. As an example of this, we’ve gone in and out of ranking for “GEO agency” terms for our homepage in the last few weeks. Not surprisingly, though, the mentions from other sites and the snowball effect of having appeared even briefly led to a halo effect of several other great mentions – because we belong on that list.  If you search the term today, you’ll see us on almost every list from a third party site – and we did nothing direct to be placed on any of them. Our own pages are just one example, but there are several others. Bidding to be placed on an affiliate listing driving the same influence. Actively participating on Reddit. PPC. As an example of indirect Reddit influence, consider the search “best CRM”. The first result? Four different links from Reddit on the subject. ChatGPT Reddit citations went down recently, and it’s not directly impactful on Claude. But do you really think these results aren’t influencing the creation of all the listacles that still influence ChatGPT and Claude directly? According to Ahrefs, the aggregated value of the first result is 50,200 searches per month, or $62,400 in traffic value. The people who create these articles almost always start on Google. And if they don’t, they probably go there second. This is why “advanced” GEO is taking strategy a step further than source URLs. It’s instead aiming to impact the areas of customer influence with *actual traffic*, and as a byproduct, increasing the probability source URLs mention you positively without action on a long enough timeline. That process of influencing the influencers is infinitely more scalable than carving out a single source URL to an audience of zero, hoping it drives a sustainable model change. That's why most are overreacting to the Reddit decline. It still has outsized influence for many audiences, and until it doesn't, that's where many brands should be participating in order to drive LLM mentions over a long enough timeline.