In this episode, Bill Scully and Bill Hunt discuss their experiences at the Google Search Central event in New York City. They discuss AI’s growing role in search, challenges with image indexing, and managing Google Business Profiles.
The episode also covers data inconsistencies across Google systems, the impact of hourly data on marketing, and the importance of content control in an AI-driven world. They highlight the fight against spam, the need for industry standards, and how marketers must adapt to evolving search trends.
Key Takeaways
- The Google Search Central event provided valuable insights into current SEO trends.
- AI is becoming increasingly important in search strategies.
- Image indexing, schema usage, and Google Business Profiles are key for effective SEO.
- Data discrepancies across Google systems complicate analysis, highlighting the need for clearer reporting.
- Networking at industry events provides valuable insights and connections.
- Continuous learning is essential in the fast-changing digital marketing landscape.
- Hourly data can reveal immediate marketing impacts.
- AI is changing how users search and interact with content.
- Content control is crucial for managing AI-generated snippets.
- SEO standards are necessary for future-proofing strategies.
- High-quality content helps stand out in a spam-filled web.
- Understanding Google’s serving team can improve search outcomes.
- Marketers must adapt to evolving user needs.
- Feedback loops are essential for refining marketing campaigns.
- AI tools require careful oversight to ensure accuracy.
- Tracking financial trends can offer insights into user behavior.
Catch the Full Episode
Youtube: https://youtu.be/3O1PL6E8btI
Spotify: https://open.spotify.com/episode/1HDBTwYcECNbwQu5Sw0JVv?si=OKXq4CD0TG6rf8OnHhc84A
Apple Podcast: https://podcasts.apple.com/us/podcast/episode-30-google-search-central-live-nyc-2025-recap/id1755789311?i=1000702070792
Show Note Links
Google Search Central Meetup NYC 2025 by Bill Hunt
https://internationalwebmastery.com/blog/google-search-central-meetup-nyc-2025/
Transcript
Bill Scully (00:01)
Welcome to episode 30 of the International Web Effectiveness Podcast, where Bill Hunt and I, Bill Scully, focus on the strategies and techniques necessary to improve your web presence and manage your global websites. In episode 29, we discussed the need for a VP of Answers and our Answers Management Program and Strategy.
In this episode, we want to talk about our recent trip to New York City for the Google Search Central event. Hey, Bill, how’s it going?
Bill Hunt (00:31)
Doing well. And you?
Bill Scully (00:32)
Good. Yeah, so we both got to go, as well as Motoko. That was nice. I didn’t get a lot of shocking information from it, but I thought it was good. It seemed like they were tailoring a lot of the content to the vast majority of people in digital marketing.
SEO. What did you think about it?
Bill Hunt (01:07)
I mean, in the first 15 minutes, I thought, Well, I just wasted a day. But it’s never really a waste because you get to see people you haven’t seen in a while, and that’s always good. The best part is getting direct access to people at Google and hearing what they have to say.
Bill Hunt (02:53)
Yeah, I thought it was a good event. It’s always valuable when we can get people in front of Google to ask questions. That’s what’s nice—there was really nothing off-limits. Of course, there were some questions they weren’t necessarily going to answer directly, and some they skirted around.
We had search relations people, search outreach or support teams, and partner representatives. There was also a lot of discussion about AI. At first, I was a little skeptical. I think I mentioned this to you—I thought, Oh wow. John Mueller started with What is search and how does it work? and I thought, You’ve got a room full of some of the top SEOs in the world—why are we starting here? But as you said, it was designed to cater to a broad audience.
If I had to pick two major takeaways, the first would be AI. AI answers, AI optimization—AI was the dominant topic. The second, which we’ll discuss in a few minutes, was evolution. I must have heard that word a hundred times. If we took the transcript and did a search, it would show up at least a hundred times.
Bill Scully (04:08)
Yeah.
Bill Hunt (04:21)
So those were the two big themes. There was nothing earth-shattering, but in the article we’re putting together, we do call out some interesting points. A few sound bites stood out, which we’ll discuss as we go deeper. But overall, it was great. They took care of us well, and managing such a large group must have been a challenge.
It was also my first time in that building, and I thought it was really nice. Definitely worth the time investment to be there.
Bill Scully (05:02)
Yeah, and like you said, getting access to people is key. At the end, I told you I was going to say hi to someone—I ended up going back in and speaking to Daniel Waisberg, the Search Console guy. I was trying to figure out some issues with image indexing for schema.
Some interesting information came out of that conversation. There’s really no good way to confirm if images are indexed. Even if you use SRC: and the image URL, it will show you images, but clicking through only tells you if it’s visible—not necessarily if it’s indexed.
Then he introduced me to Ryan Levering—you remember him? The guy wearing the schema shirt at the end? That was another valuable conversation. We’ve been battling thumbnail issues in organic results, trying to handle them with schema, but they don’t actually use schema for those images.
Bill Hunt (06:09)
Interesting.
Bill Scully (06:33)
So let me check my notes here.
I had a couple of questions for him. First, I asked if schema images that were not on the page would get crawled. In schema, you put a URL, but it’s not a clickable link—it’s just text inside the schema block. He said that eventually, Google would get to it, but it could take time. He wasn’t sure about the exact go/no-go decisions for indexing.
What he did say is that to get thumbnails indexed, it’s best to use a schema image URL that is also visible on the page. The next best thing would be to have an image map to help speed up crawling and indexation, or an XML sitemap. Schema image URLs are only used for specific features, like mobile carousels, which are now in a 1:1 format. They used to use schema images for organic thumbnails, but they stopped doing that.
Bill Hunt (07:44)
So, in an image XML sitemap?
Bill Scully (08:09)
Exactly. And he thinks Google probably never documented that change, so maybe we’ll see an update as a result of our conversation.
I left my house at 5:30 AM and got home at 10 PM, but getting those answers made the trip worthwhile.
Bill Hunt (08:27)
That’s good. Yeah, exactly. Because if it’s not documented anywhere, who do you ask, right? There’s no better person. And you’re not asking anything that’s necessarily privileged. In the first 10 minutes when we were let in, there were so many people asking John Mueller specific questions about their sites. He kept saying, “Look, I can’t tell you about your site because I’m not looking at it, but I can give you generalized feedback.” I think that’s a perfect question, and we should probably include that in our meetup notes. Those are the golden insights. Even if they can’t give you a direct answer, they help point you in the right direction. And when they say, “I don’t think we can do that,” at least you know you’re not going crazy after spending so much time trying to figure it out.
There were two things I thought were important regarding images. They mentioned that AI-generated images, especially from image-building tools, often get flagged. But you should still ensure… what was interesting is, just two weeks ago, I was judging the AI Search Awards. I better not say which one, but two people discussed how they were modifying images—very creatively—for shopping search and the paid side.
Bill Scully (09:34)
Mm-hmm.
Bill Hunt (09:56)
Immediately, I thought, “Wow,” because they said if you don’t have the correct tags—I’ll share the screenshot in the notes—like “synthetic,” “modified,” or “conjoined”—if you don’t have one of those three tags, it’s an issue. They didn’t seem to care about it then, but it popped into my mind when you mentioned it. So, that’s something to keep in mind, especially if you’re working in paid e-commerce search. If you’re creating dynamic images or working with images in a way that’s relevant to your question, like the auto-sized images, I’m curious if you asked about that.
Bill Scully (10:04)
Yeah.
Bill Hunt (10:25)
With Cloudflare and Akamai, for example, you can load a full-size, high-res image and use a code string to specify how it’s resized. I’ve used this setup before, where, instead of having three different actual images in a recipe carousel, we just use the dynamic modification functionality in Cloudflare. The image is adjusted on demand. Did you happen to ask if that approach would meet their requirements?
Bill Scully (11:47)
Mm-hmm.
Bill Hunt (11:52)
Yeah, because in some cases, those modifications are part of the filename—like adding “dot JPEG”—but in other cases, they’re not.
Bill Scully (12:05)
Right.
Bill Hunt (12:18)
Exactly. We’ve used dynamic modification, where images are sized on demand, rather than having multiple versions. It’s the same idea, and it’s something we’ve done with carousel images, where you dynamically size images based on how they’re displayed—standard, vertical, or horizontal. This is great for performance and SEO because you only serve the appropriate version of the image for the user’s screen size.
But in terms of schema, we always try to submit an image XML sitemap that includes all the images we’re using in our schema. That’s an interesting point to consider. Even if you’re not getting a direct answer, sometimes you can still apply generalized advice to your situation.
Bill Scully (13:34)
Yeah, you have to think like an engineer and ask your questions as if you were solving a technical problem.
Bill Hunt (13:39)
Exactly.
Bill Scully (13:42)
Yeah, I guess they have office hours or something where they answer specific site questions, but you’re more likely to get an answer if you ask a generalized question. John mentioned during the meeting that they can’t give preference to any specific site or provide answers that would benefit one person over another.
Bill Hunt (13:47)
Yeah.
Bill Scully (13:50)
Right, they give generalized advice that applies to multiple cases.
Bill Hunt (14:10)
For me, the value was still there. We had dinner with our daughter… and it was a great reason to go to New York. But even beyond that, I walked away with four or five key takeaways. It’s like when you go to a conference and get a nugget of information that you can use. It was interesting to see how I interpreted things differently when comparing my write-up to others from the event.
Bill Scully (14:30)
Yeah.
Bill Hunt (14:42)
Exactly.
Bill Hunt (15:01):
And we’ll talk about a few of those, but yeah, so, you know, I mentioned the two big things for me. Well, there were three, but the two were AI, AI, AI, all day. Anything you took away from that?
Bill Scully (15:22):
Yeah, the same. They’re definitely leaning into AI. It’s changing so fast. They’re not backfilling their systems to report on anything yet. You can’t get any specific information, like overviews, out of Search Console yet.
It was interesting to see. I mean, I think everyone might have guessed it, but it’s interesting to see how they use RAG to update the overviews. We know that some feedback is going into the LLM to help get the right answer. We both know that…
You know, if you’re looking for it to write a generic paragraph, it’s great. But if you’re looking for a specific number, or time, or address, it falls short. I just saw a presentation today where they asked questions I wouldn’t have thought of—coming out of our VP of Answers podcast, but they asked questions like, “Who is the manager of X company?” or “Who is the purchasing manager of X company?” And it was 60% wrong.
Bill Hunt (17:04):
Right. And to be fair, where would they get that? I mean, that’s the whole premise of our Answers Management program. Our VP of Answers podcast talked about that: what are they asking? And I think you gave a great example of that, asking the sales team, “What are people asking you?” The procurement manager makes sense because we want to get in front of that person. These are the kinds of things you’d typically go to LinkedIn for.
And so it makes sense. Sort of jumping ahead, one of the big takeaways was during the Q&A, where one of the women asked, venting in a way. She said, “Hey, we posted—she and a few others posted some horrific examples from their clients—where the address and phone number coming up in AI answers were wrong.”
And they said, “Look, you’ve got this amazing thing with Google…” I still think of it as Google My Business, but I’m trying to remember what it’s called—Google…
Bill Scully (18:18):
Yeah.
Bill Hunt (18:21):
Google… I’m going to call it Google Business Profiles. I’m old school, so it’s hard to get rid of the old vocabulary. But, and this is something I want to talk about a little bit later. My other takeaway is… and I’ll just throw it out there: I got more takeaways than I thought I had.
For one second, I think too many people think of Google as this highly functional, all-encompassing, highly efficient thing. They don’t understand that it’s a series of systems. And as any of us know, I mean, just go to your web team now and say, “Hey, can you make a change?” You know, you’ve got to fill out a ticket, justify it—all that nonsense. All the highly valuable, actionable tasks you need to make an efficient organization work.
So one thing they said was, “Can’t Google connect to the Google Business Profile?” I’ve got a business, and someone wants the phone number. It’s in the Google Business Profile—why can’t you pull it from there? And I think this is what you were just talking about, Bill, with how we have an answer. Where does it come from? How are we vetting it? And I think they had a slide showing someone asking a question.
Google can look both at their search engine and their LLM, then sort of massage the two and figure it out. What they’re calling “grounding” is an answer with links. Apparently, those links tend to be—or should be—I didn’t quite get the legally correct wording they used about where they got that answer from.
That was the interesting thing. Over time, Danny said they’d try to figure out if they can pull such answers from the business profile. That makes sense, right? Many businesses—though not all—do edit those. In a Google environment, that makes sense. And, you know, I’ll give my little rant about Google’s systems a little later, but that was an interesting discussion about the quality of the information and where those legitimate answers are coming from.
Bill Scully (20:52):
Yep.
Bill Hunt (21:12):
This was later on, but it almost acted as the catalyst for a number of other questions. They were essentially preemptively addressing what they anticipated would be the key questions, especially around AI and their direction. The biggest thing I heard from people collectively was, “When can we get the data?” People wanted to know when Search Console or any other tool would provide things like impressions, clicks, and other key metrics. That’s really what people were asking for.
I think the community as a whole just wanted to know: Is it impacting us or not? I thought Mike King actually asked a really good question, though I didn’t catch the exact wording. He said something like, “We’ve got people we have to answer to, our clients. If we could point to something from Google, something official—not just Sally, Jim, Bob, or Steve speculating—but actual numbers or data from Google, then that’s helpful.” Their answer was that they’d take the feedback and consider it.
On the one hand, it seemed like people just wanted basic data: how many queries, how many times something appears, what the quality and frequency are, and what the average click rate is. Generalized data. They seemed hopeful that Google would consider it.
The second part of the conversation was about why they can’t give us AI answers yet. I totally get it. They first pointed out that it’s complicated, and that AI is still slightly out of beta. Some markets don’t have it yet. They also mentioned prompt size—one thing we discussed in a previous episode was that the average prompt is just 22 words. When you think about the data side of that, they have it, but can they process and parse it?
They also mentioned that 25% of all queries, especially with AI functionality, had never been seen before. So, will the data be usable if 90% of the queries only show one impression because it was asked only once due to the nuance of the prompt? I get that concern.
One of the more condescending moments of the conversation came when they were asked, “What would you do with this data if we gave it to you?” That question really pushed the need for clear direction: we just want to know definitively where things are heading.
Bill Scully (24:31):
Yeah.
Bill Hunt (24:42):
That was interesting. Another clever moment came from Paul Shapiro, I think from Uber, when he asked if they use UTM parameters, because ChatGPT does. These are the little tracking parameters added to the referring URL.
Bill Scully (25:06):
Right, as a referring parameter.
Bill Hunt (25:10):
Yeah. Now, we don’t get query data anymore, so at least UTM parameters would be useful. What was interesting was that they said it’s not likely. If they did that, it would give preference to Google’s analytics products, which leverage UTM parameters. It was sort of a policy-driven answer. I thought, Okay, we’re not going to promote something from Google, but this makes sense.
It seemed like they took the feedback seriously because the demand was overwhelming. They said, We’ll see what we can do. They also emphasized that it’s still a work in progress.
Bill Scully (26:18):
Yeah, I liked the session where they talked about why the data doesn’t match, comparing Search Console to Google Trends to GA4.
Bill Hunt (26:25):
Yeah.
Bill Scully (26:34):
It never really crossed my mind, but you can customize Google Analytics’ time zone, while Search Console is fixed to California time. So, if you’re in Europe, you’re getting data based on California’s time zone. That’s good to know for anyone listening from overseas. Then they went into the differences between Search Console and Google Trends. Trends looks at topics, not individual keywords. So when you see keywords in Trends, they’re aggregated around topics, using various methodologies. That was interesting.
Bill Hunt (27:25):
They want to align things better, though. I thought it was a good point. Daniel, who gave the presentation, used to be part of Google Analytics, and this kind of highlights the disparity between different systems. Each system has its own objectives, and this shows that disparity.
In a way, this illustrates why it’s difficult to connect data across platforms. Some people ask, Why can’t we tie everything together exactly? And Daniel referenced time zone differences and other variables, which make it complicated. They did announce that you can now pull 24-hour data via the API for people using BigQuery, which is a good step forward.
Yeah, that was a really good session.
Bill Scully (29:02):
And they announced that they’re going to come out with eight days of hourly data. So, he thought it might be that day, but he wasn’t sure. I haven’t heard anything about it. I’m sure it hasn’t come out yet—people would be all over that. Let me ask you a question, a practical one: how would you use that data?
Bill Hunt (29:15):
Yeah.
Bill Scully (29:24):
No, I mean, how would you use eight days of hourly data?
Bill Hunt (29:29):
I mean, I would look at…
Bill Scully (29:30):
Sorry, eight days of hourly data.
Bill Hunt (29:33):
Yeah, I would just look at it over the day. Well, so, one way that immediately comes to mind is, going back to our session on combining, you know, making search play nice with everybody else. And I think I did nudge you when he was talking about this and say, “OK, we just launched a campaign this morning, you know, on TV or the radio or whatever archaic media.” Even if we launched it now, we could see, if we launched it this morning, we could start to see it. Let’s say, at 9 a.m. on drive-time radio or whatever, we could look at the potential time on that first day. Was there an uplift in queries?
Bill Scully (30:21):
That makes total sense. I hadn’t thought about that. So you create an event that could trigger something instant. And so you’d see something happen at 9 o’clock, then again at 12 o’clock.
Bill Hunt (30:37):
Or same day. I mean, let’s keep it realistic. I’m not listening to somebody immediately turn and start querying it, but I may. What we can see is, again, if we were launching a brand new product—this is where I’d see it. If I was launching, let’s say, the next generation or fifth-generation Prius, what’s the word, phrase, or concept? They were rebranding it as the fifth-generation Prius. So, we were able to find, in search, the immediate uptick and see on which day we could notice an uplift in people using that query when we introduced that new word or phrase into their vocabulary. So, I’d use it the same way. If I was running any net-new campaign, especially anything where I’m introducing new concepts, new products, I’d want to see, you know, not literally how quickly they started Googling for it. This also ties back to one of the points you made: if we’re launching something, you know, don’t come to us a week later and say, “Search team, how are we ranking on this?” And we’re like, “We had no idea you even launched this.” So, I think with data like this, we could see, is there a time wave, if you will? Are they doing it more in the morning?
Bill Scully (31:49):
Right.
Bill Hunt (32:02):
Is it a lunchtime thing? I’d be curious to know. Like right now, I’m not a big basketball fan, but with March Madness, is there any correlation in times when the game’s on? Because I remember, I’ve had people take me to dinner because of my client’s ads on TV during the game. They didn’t remember who it was, but it drove a ton of traffic to them because we didn’t do a good job of matching content to the commercial. But yeah, that would be an interesting way to use that time window.
Bill Scully (32:48):
Yeah, so I can see how you could use it. Now, I’m thinking the real benefit here is feedback for external actions. Because, I mean, depending on the time of day you search, it’s going to go up and down. Up and down. OK.
Bill Hunt (33:11):
Right, exactly. It’s just another level of granularity. And, you know, it is what it is. Anything else on AI?
Bill Scully (33:16):
Mm-hmm. Yep. OK.
Bill Scully (33:19):
Anything else on AI? You know, they shared anecdotally lots of information about how people’s search behaviors have changed. You know, longer search queries, people using different modes, different assets. Did they launch the AI search mode at the event?
Bill Hunt (33:57):
AI mode. Yeah, it’s still not public. You have to opt into it, but I’ve been playing with it. And… yeah, I don’t…
Bill Scully (34:06):
Yeah, I’ve… yeah. How is that? I signed up for it. I got a note that I’m in, but I haven’t had a chance to test it yet in labs.
Bill Hunt (34:16):
I mean, it’s OK. It’s nothing earth-shattering. I did a variety of queries on it, and it just… yeah, it didn’t impress me. But, you know, I think it’s a notch above. I think it really depends on what you’re searching for. Like, for example, a search I did yesterday—it gave me a more detailed response when I wanted to know how to plan a route from Malaga, Spain, up through Alhambra into Sevilla. It broke it down: step one, step two, step three, and then A versus B type of thing. And again, like everything, it comes down to what you ask it to do. I thought it gave me a much better answer than what I was getting from regular Google, which just gave me a bunch of affiliate sites. AI answers, though, didn’t necessarily answer the question directly, but they did provide options. In that regard, I think it’s good if you ask a complex question with multiple “if-then” scenarios. I haven’t compared it to other tools yet, but the one thing I like is that it can fall back into Google, which has recency data. And so, again, that’s my bias toward continuing to use Google—it can tap into Google proper and pull things from, say, yesterday. There’s a slight benefit there.
Bill Scully (36:10)
Yeah. I was going to mention a couple of things that also came up. Most people are aware that publishers have control over their content. They can use the robots.txt file to prevent AI from using their content for training. Another interesting point Daniel brought up was snippet controls. You can actually prevent content from appearing in AI-generated snippets. His example was with coupon codes—you want your page to show up, but not the actual code in the search results. You can use snippet controls to manage that.
Bill Hunt (36:56)
Yep, exactly. These are all ways to make things more efficient. You want people to find your site if they’re searching for your coupon code, but you don’t necessarily want the code to appear in the search results. The idea is to have them visit your site to access the code. Personally, I’m torn on this. From a user perspective, just give me the code. But as a marketer, the goal is to drive traffic to your site, especially when you want visitors to get it from a trusted source rather than one of the many coupon scraper sites out there.
Bill Scully (37:39)
He also mentioned that there are industry-wide discussions about standardization and future considerations, and he encouraged people to get involved in those discussions.
Bill Hunt (37:55)
Yep. Gary and I, I forget who else from Google, but I missed the window to write up my thoughts. The standards body, the IEEE or something, asked for input from the community on robots.txt adaptations, modifications, and improvements. I know Gary and another Googler submitted their input, and they recently went to Ireland to defend or argue their points. I think there’s another round happening soon, possibly in Singapore. Gary, or maybe John, mentioned they’d be attending that event. Google is very active in this, and I suspect other AI companies are as well, as they’re exploring ways to control this.
Just before this, I was dealing with an issue where something semantically simple went wrong—a market folder wasn’t included when it was supposed to be a wildcard. Google didn’t respect the market folder, which led to a huge number of requests against the content. Even though it was a minor oversight on Google’s part, it led to a lot of issues.
I think we need more work on this. We need logic tools that allow people to test variations and ensure things are being blocked as expected. There was a tester I saw once where you could input a URL to check if it would be blocked or not. These are some interesting ideas that get my brain working.
Another point I want to mention is using AI for tasks like schema writing. I was doing this recently and, being a bit lazy, I had the AI extrapolate some schema variations for me. It did a great job, but it added unnecessary spacing and mixed case, which is very specific in schema. I asked it why it was doing this, and while Google might overlook some of this, I’m a perfectionist when it comes to code. So, I always go back and re-check. Now, when I ask AI to make a correction or change, I remind it to run everything through a validator to ensure the code is correct. You have to prompt it, just like you would a human.
Bill Scully (41:16)
Yes.
Bill Hunt (41:42):
Hey, did you proofread this? Did you spell check it?
Bill Scully (42:14):
I did previously.
Bill Hunt (42:16):
Well, you just edited 14 paragraphs, so you probably need to do it again. That’s just a side note on some things you need to keep in mind when you’re relying on AI tools. Anything else about the day that stood out to you?
Bill Scully (42:29):
Nothing really jumps out at me.
Bill Hunt (42:31):
No? They went over the typical things, like the new mentions in the quality guidelines for reducing filler content and addressing AI-generated content. You brought that up—mark your images as AI-generated. They also mentioned avoiding large-scale thin content strategies.
Bill Scully (42:49):
Yeah, for people who think AI will be their savior for building websites, they’re just going to end up pumping out generic content that’s already online. It’s not going to be seen as quality, and they’ll have a tough time ranking.
Bill Hunt (43:08):
And they talked about spam. They shared a data point that 50% of all documents they encounter on the web are considered spam. Then they gave a couple of criteria for identifying malicious content. I think Danny made a good point, which we’ve discussed before—what you’re producing has to be materially different and offer real value beyond what Google already has.
Bill Scully (43:18):
Yeah.
Bill Hunt (43:38):
This is something I’ve struggled with my whole career: trying to get people to understand that their “baby” might actually be ugly.
Bill Scully (44:03):
Yeah.
Bill Hunt (44:04):
Like, what did I see yesterday? It was something I thought, “My God, do we need another one of these?” I think it was a delivery service. It’s like, really? Do we need another one of those? And someone probably put millions of dollars behind it.
Bill Scully (44:29):
Yeah.
Bill Hunt (44:29):
The point is, Danny was saying that it’s about creating something different. For example, if you have third-party content that’s there legitimately, that’s fine, but not for the purpose of fooling Google.
Bill Scully (44:53):
Right.
Bill Hunt (44:53):
It’s like the analogy they gave—imagine Babies “R” Us suddenly deciding to publish finance content, like 401k advice. Sure, you could probably spin it with some tips on saving for your child’s education, but if you’re talking about the 14 best CDs for your child’s education, that’s not materially different.
Bill Scully (44:53):
Exactly.
Bill Hunt (44:53):
And it’s like the major newspapers with their coupon sections—it’s nothing but coupon codes for every product under the sun. That kind of content has no business being on that site, available to Google. You can have it if it’s part of your business model, but if it’s spam, Google won’t reward you for it.
Bill Scully (45:04):
Yeah.
Bill Hunt (45:08):
Spam crawling and those types of things are huge problems. And if you think about it, 50% of what Google is encountering is spam and garbage. They have to figure that out. So, if your site isn’t significantly more relevant or better than others, you might not get indexed.
Bill Scully (45:20):
Right.
Bill Hunt (45:20):
Just keep that in mind. It ties into another point we’ll discuss more in our next episode: AI and inaccurate information. Google is trying to improve here, and it’s good that they’ve acknowledged this.
Bill Scully (45:38):
Exactly.
Bill Hunt (45:40):
The other big takeaway for me was how much they mentioned evolution. Liz Reed even said, “The search problem is never solved.” Google wants to think differently. They shared five points, and one thing that stood out was that Google is a for-profit company. It’s not all altruistic, and we need to adapt to changes. Evolution is key in their strategy—dealing with spam, evolving user needs, and how the SERP will evolve.
Bill Scully (45:56):
Yeah, that’s true.
Bill Hunt (46:00):
The idea of how Google is constantly evolving ties into how the search engine results pages (SERP) are changing. John discussed how all the pieces fit together—pages, crawlers, indexers, ranking, and serving. It’s not just a simple process; each team handles separate systems. And when we, as international search professionals, talk about HREFLang, it’s frustrating when Google serves the wrong version of our pages.
Bill Scully (46:20):
Right.
Bill Hunt (46:22):
Google is just like any other company—it’s not one giant orb of efficiency. Things can go wrong, and it’s not always seamless. But understanding how these separate systems interact helps us navigate the complexities of search better.
Bill Scully (49:22):
Yeah.
Yeah, I mean, I think it would be really good to understand that it’s almost exactly like your company, right? There’s a mission, and people are in different roles. The output of your company would be a product, and the output of Google would be the search results, in the case of search. There’s the person who has to buy the raw materials, then somebody has to ship them to the manufacturer, and then someone else has to receive them. After they receive them, they have to put them in queue for production. Then someone cuts them, and someone else assembles them. It’s no different in a search engine.
Bill Hunt (50:21):
Exactly. They all have different goals, and then somebody — I love this analogy — the person sourcing the products, right? They’re probably trying to get them as inexpensively as possible. And does that open up tolerance faults, which someone downstream has to deal with? That’s kind of what these guys have alluded to. But coming back to the serving team, right? I think…
Bill Scully (50:52):
Yeah, they have to take all the elements that everybody puts together for them and figure out how to get the best user experience.
Bill Hunt (51:02):
Exactly. And that’s exactly how John phrased it. They have the job of taking the query, interpreting it, and then reaching down into the depths of Google to pull out all the bits that answer it. Is it best with images? Is it best with maps? Is it best with text? Is it best with an AI answer? And so on. The hope is that all of those bits — like, back to your question about what image is being served in image search — all fit together well. They talked about the evolution of things like Lens and Circle Search, and how all these different elements are now operating. If they’re all using faulty materials… I love this analogy — procurement screwing up things downstream. I remember my wife coming home from her first job in the United States with a Brother typewriter. People would order parts, and they’d have a whole team dedicated to fixing issues caused by misfit parts, just to make the whole work because the tolerance wasn’t right. Yeah, it’s like knowing what ingredients went into the sauce. But the other thing these guys have to deal with is Google making money.
Bill Scully (52:47):
I can’t even imagine a team with the dual mission of serving the best results in the right context, in the right format, while also managing KPIs around how to generate revenue and keep it sustainable.
Bill Hunt (52:49):
It’s no different than any other company. Yeah. You can make a million-dollar pen, but you’ll never sell it. You won’t make money.
Bill Scully (53:14):
Exactly. How do we merchandise this stuff without… I mean, it’s…
Like, I love this. See, this is why it’s good to talk like this and then go deeper, because it’s like when you go to your supermarket, and they happen to put their branded products on the end cap instead of Barilla pasta. How does that look to people, right? It’s the same thing. It’s like, okay, now we’ve got the people paying to be there, and then we’ve got the other side. So it’s an interesting thing. I think we should drill down into this.
Bill Hunt (53:29):
But anyway, I just wanted to say that they’re no different than the chaos in your company. The more we can figure this out, the better. And the last thing I want to leave people with is to follow the money. Think about where things are heading. Google referenced something, I think it was passive-aggressively or maybe cryptically, about how rapid changes are impacting behavior patterns. They said that users are increasingly turning to closed-garden platforms. I love how they phrased that. It’s very much like AOL back in the day.
I didn’t grab the slide number, but it’s the one that has one, two, and three. They mention new platforms, new user preferences, and new technologies. Specifically, they say users are increasingly turning to closed-garden platforms for information needs that were previously met by search engines. This is resulting in less traffic to the open web. Essentially, they’re saying that if you use ChatGPT, you’re stuck.
Google, on the other hand, still sends people out to the public web to some degree, so we’re not a closed garden. When I start hearing words like that integrated into discussions of rapid change, it’s pretty interesting. These were the same words people used against AOL when it was the dominant platform. Now, those words are being used in a different context. They’re not saying “don’t go there” because it only shows you what they want you to see. Instead, they’re saying, “Stay with Google because we’re open and free—we let you pull in anyone’s stuff.”
I just thought that was interesting. They’re using this as a reason for why search is evolving.
Bill Scully (54:07):
Yeah.
Bill Hunt (54:26):
And the third item they mentioned was new technologies and behaviors. I think this ties directly back to the serving team and their mission. I just want to leave people with this thought: if you haven’t done this recently for your business, it’s worth revisiting. I’m working on an episode about three to five efficient things you can do to make your website more effective. One of the first things is to take a look at the SERP.
And like Bill, how you started with images—your client wants their images to appear because for certain queries, images are what show up. The question is, are your images showing up? Are they the right ones? So, for each of these modules—each iteration of the search results page—ask yourself: do we have representation? Whether we call it omnichannel or something else, do we have the right context? Are we answering the question in the way Google or whatever system thinks it needs to be answered?
If you haven’t done this recently, go look at the SERPs. Take time to really examine what’s being presented for different queries. Are we there? Can we be there? Do we have the resources to be there? As Bill mentioned earlier, use schema and XML sitemaps, especially if it’s an image-heavy play. Make sure the context is understood. Learn how image search works.
Bill Scully (56:20):
Yeah.
Bill Hunt (56:33):
I think another article we should talk about in an upcoming episode involves Liz and some former Googlers discussing Google’s shift to a more closed-garden approach. The interview with 21 former Googlers talked about this—how Google is trying to keep people within its ecosystem, presenting only Google content. But the question is: how do we play in that environment? At the end of the day, Google is just a collector of information. I won’t use the “P” word, but you get the idea—they leverage other people’s content to present things to you.
As we consider this, we need to be mindful of where things are headed. How can we benefit, and how can we ensure our content is part of that ecosystem?
It was a good event, lots of valuable sound bites. As I’m talking through this, I realize there are four or five other things I want to dig into. Hopefully, they’ll release the video at some point so you can see it for yourself.
Bill Scully (58:42):
They’ll probably carve it up into the different agenda topics.
Bill Hunt (58:47):
Yeah, that’s true. Looking back, I’d say it was really worth attending. I’d definitely go again. And if anyone ever has the chance to go, they should definitely do it.
Bill Scully (58:58):
Okay, I think this is a good place to wrap up. What do you think?
Bill Hunt (59:02):
Sounds good.
Bill Scully (59:15):
If you’re enjoying the show, we’d really appreciate it if you’d take a moment to rate and review the podcast on your favorite platform. We also want to hear from you. Your feedback helps motivate us and plan future episodes.
Bill Hunt (59:39):
Perfect. Yeah, that sounds good.
Bill Scully (59:47):
You can find us by searching for International Webmastery on LinkedIn, Facebook, and X. And please—don’t just give us a star. We’d love to get some feedback and know that you’re actually listening and getting something out of this. Tell us what you’d like to hear next.
Bill Hunt (1:00:06):
Sounds good. See you next time.
Bill Scully (1:00:09):
Yep. Bye.