In this episode of the International Web Effectiveness Podcast, Bill Hunt and Bill Scully take a deep look at global web strategies and the importance of building relationships in international markets. They cover the challenges local teams face in optimizing digital presence and highlight SEO as a collaborative effort. Bill Hunt also shares key insights from the recent PubCon, focusing on the latest trends and strategies in international SEO.
Key Takeaways
- Value of Localization: Understand that different markets require tailored approaches; avoid being overly aggressive in pushing strategies that may not fit local needs.
- Time Constraints: Acknowledge that local teams often have limited time (e.g., one hour per month) to focus on digital marketing, and prioritize helping them maximize that time effectively.
- Focus on Supportive Markets: Direct resources and efforts toward markets where your organization can provide real support, particularly post-sale.
- Collaboration in SEO: Emphasize that SEO is a team sport, requiring synergy between content, technology, and localization teams for effective outcomes.
- Data Centralization: Integrate data across platforms to uncover insights and opportunities that can enhance global strategies.
- Export Opportunities: Recognize the vast untapped export potential in the U.S. and explore avenues to assist companies in starting their export journey.
- Utilizing Tools for Market Research: Use tools like Search Console to identify international demand for products and inform marketing strategies.
Catch the Full Episode
YouTube: https://youtu.be/HFiqqVNiXDg?si=hDEUT1ngOWNZtZoE
Spotify: https://open.spotify.com/episode/71Au0NNe4GSsonNGDvSBzp?si=fgAOFU_OReSrKdvVLwiAAQ
Show Note Links
Pubcon Conference: https://www.pubcon.com/
Bill Hunt’s Article: Pubcon 2024 – Minimize the Challenges of Going Global: https://www.back-azimuth.com/pubcon-2024-minimizing-the-challenges-of-going-global/
Ryan Jones Article on why SEO’s must evolve their thinking beyond old school checklists https://www.searchengineworld.com/its-time-for-seo-to-evolve-ryanjones
Ryan Jones SerpRecon – https://serprecon.com/
Chat Assistant: https://chatgpt.com/g/g-bIrnI1355-seer-s-screaming-frog-technical-seo-companion
Pubcon Speakers:
Ryan Jones – https://www.pubcon.com/bios/ryan_jones.htm
Bruce Clay – https://www.pubcon.com/bios/bruce_clay.htm
Will Reynolds – https://www.pubcon.com/bios/wil_reynolds.htm
Transcript
Bill Scully: Welcome to episode fifteen of the International Web Effectiveness Podcast, where Bill Hunt and I, Bill Scully, focus on the strategies and techniques necessary to improve web presence and manage global websites. In our last episode, we discussed the benefits of product information management, sharing great tips on how to centralize and maximize your product information. That episode was packed with valuable insights.
In today’s episode, we’re changing things up a bit. First, we’ll dive into international SEO and Bill’s presentation at Pubcon, followed by his thoughts on the conference.
It’s great to see you, Bill! How was the trip?
Bill Hunt: It was good. Pubcon was great. Brett really nailed it when he called it “Search Chaos.” For many attendees, it felt like search is in chaos right now, especially with all the developments around AI. But for me, it was special because it marked the twenty-fifth anniversary of the first Pubcon. Only a few of us who were at the very first one were there this time.
Bill Scully: Wow, twenty-five years. So you were at the first one?
Bill Hunt: Yes, I was at the very first one—back when it was actually in a pub.
Bill Scully: Oh, really?
Bill Hunt: Yeah, that’s where the name came from, which I don’t think many people know. Over the years, Brett wrestled with this because people liked the name “Pubcon,” and some tried to spin it as “Publishers’ Conference” or whatever. But it was actually a meetup after SES London at the oldest pub in London. That’s where it all started—an epiphany moment for many people, where they learned more from the pub interaction than from the conference itself.
That gave Brett the idea: “Why not just skip the conference part and get everyone to the pub?” Of course, it evolved into a proper conference, but there was always that “Pubcon” spirit, where, for a long time, the day would include several hours at the pub, fostering real exchanges of information.
Bill Scully: That’s fascinating. I had no idea about the pub origin. So, did you share any interesting insights during your presentation?
Bill Hunt: Well, in my session, I actually mentioned how, in 2004, I gave my first Pubcon talk on enterprise search. That was the first time anyone had ever spoken about enterprise search at a conference. At that time, the audience was small, and not many people were interested in the topic. But when I explained how linking anything on our product and service page could immediately push us to the first page of Google, people got intrigued. Our product page was a PageRank 9 or 10 page, depending on the day, and I had a line of people outside the door, all eager to pay to get their link on IBM’s product page.
That led to the creation of my second agency. Within four months of that presentation, we had five of the biggest companies in the world wanting to crack international search. It was a major pivot for me.
Bill Scully: That’s incredible—amazing how one presentation can change everything. Did you notice any trends at Pubcon this year?
Bill Hunt: Yeah, I noticed something interesting: about seventy percent of the people attending were either first-time attendees or beginners, yet they wanted to dive into advanced topics. That was a unique trend. There were over sixty sessions, and twenty-eight of them had “AI” in the title. It was definitely an AI-heavy conference.
Bill Scully: I can imagine—AI is everywhere now. What were some of the key discussions?
Bill Hunt: It was a bit of a “Marsha Marsha” situation—everything, even if it wasn’t explicitly labeled as an AI session, had at least one reference to AI. People were sharing simple prompts to streamline workflows. For example, prompts to rewrite meta descriptions or titles.
I also picked up an interesting nugget from Ryan Jones, who has been analyzing AI-generated responses. He found that some AI answers were coming from meta descriptions. He suggested that we might need to rethink meta descriptions, especially since Google now generates 87 to 90 percent of them.
Bill Scully: That’s a great point. Is there anything else from the conference that stood out to you?
Bill Hunt: Yes, there were a lot of discussions on how search engines are using AI to understand content intent. Some people were talking about ways to leverage AI-generated results for SEO. I even saw my first “AI Optimization Expert” name badge in the wild. I had only seen it on LinkedIn before, but there it was at the event.
Bill Scully: It’s crazy how much AI has become a part of SEO. What were some of your takeaways?
Bill Hunt: Well, traditional SEO—whether you call it old-school or modern—is still very much alive. AI is primarily being used for workflow efficiency. There was also a noticeable international audience. I met people from all over the world—India, Japan, Korea, Thailand, Uzbekistan, and across Europe.
Bruce Clay shared an interesting shift he’s made: he’s rewritten the FAQs for all his product pages, not just with simple answers, but with more detail. This shift is paying off. Not only is he getting featured snippets, but his richer, more comprehensive answers are being picked up in AI responses too.
Bill Scully: That’s an interesting strategy. It seems like people are starting to think beyond the traditional approach to content.
Bill Hunt: Exactly. It goes against the grain of what many UX experts recommend—simplifying content and keeping it short. But when you provide more value and detail, you stand a better chance of getting picked up by AI, which is becoming more important as AI-driven search results continue to grow.
Bill Hunt: Bing recently expanded their partners around IndexNow, and they seem to be doubling down on it. In his keynote, he mentioned how expensive it is to get people to index content. It’s a costly process, and they’re really trying to get things moving. I think he referred to it as “frontier stuff”—things they’ve never seen before.
Their solution with IndexNow is definitely working. They’ve expanded their partners, including Wix, and they’re reaching out to CMS providers because they can detect when content is updated.
I haven’t promoted this yet, but I did a lengthy video recently, ranting about crawling, respecting crawling, and indexing. I’ll be posting a link to that shortly.
Another thing Bing announced is a new chatbot in Bing Webmaster Tools. It’s brilliant. Basically, you can connect to it and say, “Hey, break down my latest data from Bing Search Console. Tell me if there’s an anomaly.” The chatbot will pull out data and highlight where things have changed—whether it’s a section, keywords, or a level that’s increased or decreased. I thought that was brilliant.
If you notice a decline, you can ask it, “What are some recommendations for optimizing this content to recover?” or “How can I gain more opportunities from this content?” And it will provide suggestions.
Now, remember, it’s a chatbot using AI. The results aren’t groundbreaking, but I think it’s a valuable tool—especially for your budget clients. It’s almost like it can write some of your reports for you.
I’m curious how long it will take for Google to introduce something like this in Search Console. Bing definitely has an advantage there.
Bill Scully: So, is this pulling data directly from your console?
Bill Hunt: Yes, within Bing Webmaster Tools, the chatbot can pull data from there. It uses their AI functionality, Copilot. It’s simple: you write a prompt like, “Give me insights from this month’s data,” and it lays it out pretty clearly.
There’s a demo, and I’ll find the link. I think it’s on a blog post, and we’ll include it in the show notes. Better yet, go try it out for yourself.
Bing is really enhancing their tools for webmasters. As they continue to grow, especially in frontier web areas, they’re becoming a more powerful player.
There were two other things I want to mention. First, Will Reynolds did a great job in his keynote. He focused on customer engagement, not just traffic. He shared some LinkedIn posts about how they’re looking at engagement and what’s showing up in AI answers. It’s about how they’re truly engaging their audience.
He also mentioned that they’ve taken a lot of their knowledge base and turned it into chatbots. One of those is the Screaming Frog Technical SEO Companion.
This tool already documented some of the more technical features of Screaming Frog. I’ve used it a couple of times this week, just out of convenience, to figure out things like how to form regex or XPath for extracting specific content. I wrote the prompt, and it immediately gave me the exact code I needed. I copied it into Screaming Frog and was good to go.
It took maybe a minute to craft the prompt, and it even included exclusions that I hadn’t thought of, which saved me a lot of time. It’s an amazing adaptation of an already great tool.
He’s also written detailed articles on how to extract things, and I think he may have also written about vectors. Yeah, I’ve been using his work as a learning resource, especially with AI. It’s a game-changer in how you interact with tools like this.
Bill Scully: Turning it into a chatbot—that’s pretty wild.
Bill Hunt: Yeah, it’s cool. There’s a link to it in the Screaming Frog guide. You can click it, and if you have ChatGPT, it should work with the free version. Once you start interacting with it, it’s pretty impressive.
The last thing I want to mention is Ryan Jones. He launched a new tool called SERP Recon. He used it during his presentation on AI answers and AI results. It provides a more modern approach to understanding search engine results.
For those who aren’t familiar, Ryan works at Razorfish as their head of SEO. He’s a brilliant software developer, and I’ve had the chance to use the beta version of SERP Recon. It uses the latest methodologies, including Yandex’s code and Google’s recent hacks. It also incorporates cosine similarity and BM25 to explain why certain content ranks the way it does.
Bill Hunt:
So, it’s leaps and bounds beyond what other tools out there can do, especially when it comes to vectors and things like that. It’s really inexpensive, and it’s called SERP Recon. I’ve used it, the beta version, to do some analysis, and it was the only way I could get some pretty hard-headed people to accept what I was telling them—why their content wasn’t connecting with people.
Definitely check it out. There are probably a few blog posts out there about specific sessions, but the net takeaway is this: AI is here. You have to get used to the new methods and thinking around search. And like you mentioned, using AI to make you smarter, more efficient, and really shave time off your workflow.
Yeah, AI won’t take your job, but other people will if you don’t learn it…
I forget who said it—maybe it was Professor G—but “AI is not going to take your job, but someone who knows how to use AI is going to take your job.” I thought that was pretty compelling. Every session I attended had a reference to it. All of them, except a couple of hype sessions, talked about how to use AI to make you more efficient and effective.
I’ve been pushing the idea for a while now. I don’t think AI is a great writing tool, but I’ve been working with agents. I think I shared with you the Make.com stuff. You can break down each section in an article as its own prompt, give it its own tone and edge, and specify what it needs to include and exclude. You run it through a template, and it comes pretty close to producing really good content.
After last week when we talked about product management, I created an agent that goes to a site, gathers data from the product page, the PDF, and the spec sheet, and then generates both a short 500-word description and a long 5,000-word description. It pulls everything, analyzes it, and, like, “Oh my God!” It took me six hours to do, but it wrote 50 product descriptions in four minutes.
That’s insane. It would have taken a month to produce those—40,000 to 45,000 words of product descriptions. And now people can read and review them.
This is the thing I’ve found—we’ve talked about this before: people hate writing from scratch, but it’s easy for people to review and edit. What you just did, you could parse it. As we discussed product information management, you can store it and have someone review it.
If they reviewed a longer version, say a 500 or 300-word description, and the 10, 50, or 100-word versions are subsets of that, then they’ve already vetted it. You can be more confident in just plugging it into the database.
I think the point you made, and what a lot of people were discussing at the conference, is that they don’t put the rigor in. You need to write with your audience in mind. The more of those elements you can add, the closer it’s going to be to what you need it to be.
Now that you’ve got it and you’ve perfected it, you can apply it to any other client and get the same results, like you said, in four to five minutes. Then spend the rest of your time doing the human part, fine-tuning it.
What’s great is you can feed the fine-tuned version back in, and that will further train your model in their context. That’s what I think this was designed to do—not to replace anyone, but to augment them. We can now free up people who were spending a month rewriting all this content to focus on other tasks that humans are really good at.
Bill Scully:
Thanks for that update. That’s really good stuff. It’s funny; I always thought that PubCon was a publisher conference. I thought it was more for bloggers, so I never considered going. I knew there was a lot of SEO-related content there, but I just followed it on Twitter.
I’m heading to Brighton next month, in a couple of weeks. I’ll give you an update on that.
Bill Hunt:
Great, sounds good. A lot of people at PubCon were talking about heading to Brighton.
I think this is the second year in the States, in San Diego. How did your presentation go? Was it a keynote?
Bill Scully:
No, it was a specialist track session on international SEO. There was mine, and then one other international session where Motoko Hunt and Michael Bondfields talked about an overview of international SEO.
Bill Hunt:
Mine was really about minimizing the challenges of going global. Michael Bondfields, who might have written a post about it, actually used AI. His session was about identifying what market to target, understanding market opportunities, and how to reach the right audience. It was a series of prompts using three or four different tools to guide the process.
It was pretty cool. We filled in a matrix similar to one I worked on over thirty years ago, which had ten thousand variables for decision points to determine the best method for going global. My role was to work on the weighting logic of those variables, and Michael just breezed through it. He showed us how to fill in the criteria, like ease of doing business.
One key aspect of determining the right market is understanding what kind of content we need. The AI helped tease out persona information—like where people need more content, less content, or content with more nuance. Are there linguistic differences?
He even did a challenge where he gave people phrases and taglines and asked them to put them into an automatic translator. Interestingly, he had them do Japanese, so Motoko, a native speaker, could verify if the translations were understandable. Some worked, some didn’t.
The key takeaway, as you were talking about earlier, was when they added variables like “write it for this context” or “use a local idiom,” it improved the output. When they simply translated it, you could get the gist, but when they added specific prompts, the results were much better.
In the final test, when they made it more equivalent to a local Japanese phrase, it was almost spot-on. Motoko recommended using that approach.
Bill Scully:
One thing they had in their session that I also mentioned in mine was more of a rant. I was really trying to explain what international and global SEO really are. I see a lot of people with that title, but then they go on to discuss language- or market-specific SEO, like link building.
We can talk about semantics and nuance, but I prefer to place that under language-specific SEO (e.g., doing SEO for German or French) or market-specific SEO (e.g., targeting Korea, where we have Naver, a totally different search engine with different presentation methods).
Bill Hunt:
Exactly. It’s funny you mention that. Christopher Chin was at an event in Singapore and posted at the end of the first day, “I’m here in Asia with some of the top SEOs in the world,” but everyone only talked about Google. No one mentioned any of the international search engines. That’s the problem when people put “global” or “international” in their title but don’t respect the true differences in the market.
Bill Hunt: What I try to do is define multinational or international SEO as the effective organization and alignment of your go-to-market strategy, web infrastructure, integrated workflows, and cross-market collaboration. It’s essentially everything we’ve discussed over the course of these episodes, and we’ll continue to build on that.
I broke it down into four key impact areas. One of those is understanding the market opportunity, though I didn’t go too deep into that since it was Michael’s area of focus. However, one thing I did discuss, and we’ve touched on this before, is Google’s translated results report, which you can find in the search appearance section of Search Console. You’ll see data about words, pages, countries, and search appearances—whether it’s on mobile or desktop. A relatively new addition is the translated results section.
I shared an example from a site that had 566,000 impressions from translated results. Google had taken their English pages, in this case, the product pages, and presented them as snippets in two markets: India and Indonesia. These pages were localized for those regions, and there were 47,000 clicks on those machine-localized snippets. I showed how I drilled down into that data—examining impressions, clicks, pages, and markets. The word that drove the most traffic? “Artificial intelligence.”
I asked the audience to consider what Google had done here. In both India and Indonesia, Google took the English page on artificial intelligence, localized it into Hindi and Indonesian, and displayed it because it deemed that to be the best content for the query, even though it wasn’t originally in those languages. This is SEO 101—Google tries to surface the best content, regardless of the language or origin.
And this is textbook SEO. As we always tell people, “If Google creates a report, it’s worth looking at.” This example highlights the importance of understanding and acting on these insights. I’ve already shared this with the company since I have access to their Search Console. I suggested they drill down into the concepts and pages Google is showing in other markets and localize them further to better target the local market.
Along with other tools we’ll link to in the show notes, this is low-hanging fruit to act on. The next topic was content and localization. I showed a LinkedIn post where someone made a similar boast to what you just did, though not as smartly executed. They used AI to localize content into 21 languages and distributed it to 25 markets. A few months later, they became an Hreflang Builder client, likely due to the scale of their efforts.
They deployed these 21 localized versions across 25 markets, but in places like China, they included languages such as Hindi, Swedish, and Arabic—markets where these languages weren’t commonly used. Of course, Google didn’t favor this approach. They tried to use Hreflang to force Google to pick up these versions, but Google was less responsive to them, especially for markets with negligible traffic in those languages.
We also discussed how Bing was cracking down on this strategy, deeming it ineffective, so they didn’t index those pages. Google, similarly, was applying a canonical version—usually the primary language version, like US English—over the less relevant versions for other markets.
I also brought up the difference between popular search queries and linguistically correct ones. Popularity is a factor in this. In a session by Motoco, we discussed an example of a champagne brand caught between SEO and content teams over how to write “champagne” in Japanese. One version was what the brand manager preferred; the other was more common. The brand manager wanted to target a more affluent audience, so he favored the more exclusive term—even though it received far fewer monthly searches compared to the popular version.
This brand aimed not to drive massive traffic but to grow their membership or newsletter base. It was a case of quality over quantity—something we don’t often see in SEO debates. The external agency wanted to use the more popular term, but the brand manager stood firm on using the more prestigious term to target a specific audience. After testing, the brand manager was proven right: the more prestigious term met their objectives better, even with less traffic.
The third major area I covered was CMS and infrastructure. We’ve touched on this before, and I’m sure we’ll revisit it. One of the questions I addressed involved IP detection challenges. I’ve had some interesting conversations with people about this. For example, I’m writing a point of view on A/B testing solutions, specifically about testing IP detection methods.
I shared a couple of challenges I’ve seen. One was an issue with site maps being redirected back to the US version. This meant markets outside the US couldn’t access localized site maps because the system was set to only allow German visitors to see the German site map, for instance. Google wasn’t able to get the right maps because of this setting.
Another example involved cookie settings. Some sites forced visitors to accept cookies, but crawlers can’t accept cookies, which caused problems when I tried to visit other markets. Even with a VPN, I encountered a 503 error, indicating the site was unavailable. Worse, the site displayed a 403 error, which essentially masked the issue, telling me I couldn’t access the market due to cookie restrictions.
These types of issues are still pervasive. And I realized that there’s no tool out there to diagnose these problems effectively. For example, a site might redirect visitors from Lithuania to a Swiss-Italian version of the site—a language and currency not relevant to them. This happened due to a simple mistake where “LT” for Lithuania was confused with “IT” for Italy, a common error with Hreflang settings.
I’ve been investigating this for some time and noticed similar issues with other sites I’ve checked. It’s crucial to ensure that IP detection systems are configured correctly, and I’d like to see more testing done around this. For instance, when people encounter the banner asking if they want to visit a local market site or a foreign one, what happens when they click? Does it affect their experience or checkout process?
Bill Scully: So, I think there are some interesting tests and projects that can be done around this. Maybe it’s nothing more than coming up with prompts that trigger VPNs, testing it across all your target markets. Someone will eventually build a tool for this. But yeah, those are some of the things I discussed regarding CMS and infrastructure issues.
The last thing to close out is many of the organizational topics we’ve discussed, like we talked about a couple of episodes ago—URL uniformity, centers of excellence, sharing information. It was interesting talking to people from these other markets; they were saying, “Yes, that’s exactly what we need. We don’t get any guidance from the mothership. We don’t even get recommendations.” Just like we discussed with URLs—giving people recommended URL formats. They don’t even get that.
Bill Hunt: Right, and as we mentioned in that podcast, you can give people recommendations without making it a mandate. A lot of people will follow them. It was things like that—how to work through this. I showed a matrix I hadn’t shown in years, but I had a concept I called “protocol-compliant participation.” It was a fancy way of asking, “Do they want to play ball?” If they didn’t have their ducks in a row and didn’t attend meetings, they weren’t part of the center of excellence.
Bill Scully: If we had to decide, “Do I spend two hours with this market or another market?” the one with a framework in place, one that is cooperative and collaborative… That’s where we’d focus. I think it was like programmatic collaboration. I mean, simply put, do they want to play ball?
Bill Hunt: Exactly. Anyone who raised a “no” flag—we just didn’t engage with them unless they specifically asked us something. We’d send out mass emails but didn’t aggressively push. We focused on markets that were willing and able to participate.
Bill Scully: I got one person who thought that was harsh, but another person thought it was brilliant. The point is, unless I tell people about my matrix, they won’t know. I’m not going to go to Italy and say, “You don’t pick up the phone, we’re not doing anything with you.” We include them wherever we can, but we’re not going to force it.
Bill Hunt: Right, some regions—like where the sales manager’s secretary is responsible for the web—don’t even know the CMS. They only have copy-pasted content from the global site or major same-language sites. You can’t force them to engage. I’ve seen it firsthand.
Bill Scully: Exactly. It’s like when I did road shows—sometimes, you just have to let them be. They’re busy, they’ll never engage fully, so just redirect your efforts where they’ll make the most impact.
Bill Hunt: Yep. When I did road shows for various companies, we’d ask, “What’s your job?” and some people could allocate only about an hour a month to digital marketing. They were assistants, secretaries, or digital marketing managers handling everything else. Their focus was spread thin, so they’d say, “With that one hour, what can I do?”
Bill Scully: We’d try to help them make the most of that hour. Focus on people who want help and can make use of your time.
Bill Hunt: And that’s the takeaway: focus on markets where you can provide support. There’s nothing worse than trying to help people post-sale and not being able to deliver the support they need.
Don’t just rely on AI for content. Data centralization is key—integrating data, finding valuable insights, and aligning technology, localization, and content teams to create synergy.
One of my epiphanies from thirty years in search was that SEO is a team sport. In search, we often have KPIs with no clear responsibility.
Bill Scully: Building relationships and synergies with people is critical. You’ve mentioned this on the podcast multiple times.
Bill Hunt: Absolutely. And closing out, the U.S. government has identified over a trillion dollars in untapped opportunities for U.S. companies that haven’t exported yet. The frightening statistic is that 77% or 87% of these companies haven’t exported because they don’t even know where to start.
Bill Scully: It’s amazing. There are grants and resources available, like you mentioned from an event in Connecticut. There’s no shortage of information for companies that want to explore this.
Bill Hunt: The challenge is that many people just don’t know where to begin. I think a future episode could dive into using Search Console data and other tools to pinpoint where in the world people want your product. From there, we can explore the steps to take.
Bill Scully: That sounds like a great future episode. It’d be interesting to also look at IP detection—setting up test protocols to track success and revenue increases.
Bill Hunt: Exactly. There’s so much we can unpack with that. I think it’s a good topic to cover.
Bill Scully: That was a great wrap-up. I enjoyed this conversation.
Bill Hunt: Me too!
Bill Scully: If you enjoy our podcast, let us know! Please rate us on your podcast app and share it. Your feedback is greatly appreciated. And if you hear Bill talking about all these tools and presentations, check out the trainings on the International Web Mastery website.
Bill Hunt: I’ve put together some excellent training—it’s not basic, but I can explain complex topics in simple terms to help you understand and implement them.
Bill Scully: Any other comments?
Bill Hunt: Nope! Check them out. As I mentioned, let us know what you’d like to hear from us. We’re doing this for you. Join us on our free Facebook group—we look forward to hearing from you!