The Whitespark Local Update

🔥 Join the webinar

Darren Shaw and Claire Carlile are back with another roundup of the biggest updates, insights, and conversations shaping local SEO and AI search, including:

📝 If You Don't Understand Skills, You Don't Understand Claude (Eliot Prince)
📝 AI Overviews getting demoted in most local search results (Michael B. Snow)
📝 “Major updates” coming for SABs (Payal Badhran)
📝 Third-Party Citation Optimization Prioritization in an AI Search Era: A Framework (Aleyda Solis)
📝 Why Google Deletes 2% of Reviews, Siri Search, Google’s New In-Stock Check & Its Ad Revenue Miss (Near Media)
📝 Gone in 7 days: Why Google reviews disappear (Reviewflowz)
📝 The risks of optimizing SEO for AI over users (Andrea Schultz)
📝 Great webpage pruning and content focusing example (Jan-Willem Bobbink)
📝 Google’s Ranking Signal Distribution System (Koray Tugberk GUBUR)

---

Follow us ❤️

📑 LinkedIn:
Darren Shaw: https://www.linkedin.com/in/darrenshawwhitespark/
Claire Carlile: https://www.linkedin.com/in/clairecarlile/

---

Listen on 🎧
Spotify: https://open.spotify.com/show/4lB61Txo1qXheQe4nrzOKO
Apple Podcasts: https://podcasts.apple.com/us/podcast/the-whitespark-local-update/id1837196873

#googlebusinessprofile #AI #localseo

What is The Whitespark Local Update?

The Whitespark Local Update is the go-to podcast for Local SEOs and Marketers who want to stay ahead of the curve in local search and the local visibility space.

Join industry experts Claire Carlile and Darren Shaw for a lively, insightful roundup of their carefully curated selection of top “must-read” and “must-watch” links, including news, trends, and can't-miss resources.

[Darren Shaw]
Welcome to another episode of the Whitespark Local Update podcast with me, Darren Shaw.

[Claire Carlile]
And with me, Claire of the Carlile.

[Darren Shaw]
I mean, we did not even coordinate our sweaters, but look at this. Let me see your sweater. What is happening?

Like it's like we just showed up both wearing our like Nordic sweaters. This is amazing.

[Claire Carlile]
Yeah. For anyone listening and not watching, side note, we are both wearing Nordic sweaters.

[Darren Shaw]
Or jumpers. If you're in the UK, you can call it a jumper. This is a jumper, right?

[Claire Carlile]
Yeah. That's a lovely jumper.

[Darren Shaw]
Your jumper is quite lovely as well.

[Claire Carlile]
Thank you.

[Darren Shaw]
Claire, tell me what is happening in the world of local search. What have you discovered this past week? Anything interesting?

[Claire Carlile]
Well, it's funny you should say that Darren. I have actually been away. I have been in Barcelona on the Product Expert Summit, which was very, very lovely and absolutely wonderful and so much fun times.

So I've actually only just got back. So then suddenly I went, I need to find my linky links. So suddenly I found some linky links.

So I don't know what I'm talking about. I'm just going to, uh, so these are some things that I have found very recently in the last hour and a half. Um, I'll go with my first one.

It is a video on the YouTube from a guy called Elliott Prince that I have been following for a little bit of time and I would actually like to do his course. So he writes a lot about Claude and how to use Claude. Have you, have you, have you seen any of Elliott Prince's stuff Darren?

[Darren Shaw]
I have not, no, but you know, that's all I do all day long is use Claude.

[Claire Carlile]
So he's a British guy. Um, and with me and Claude, I mean, I have a very close relationship with Claude. Claude knows most things about me.

Um, but I think with like use of AI generally, I think a lot of people are sort of just doing things without necessarily knowing very much. So I would like to be a little bit more structured in how I use, um, AI. But anyway, so in this video, Elliott is talking about skills.

So Claude skills, what they are, how you use them. Um, I have set them up, it appears some, um, but I didn't know that I had, so it's worth watching this video. He just, I like the way that he sets it out.

So a Claude skill is a checklist that runs itself. So, um, I work with some like quite long prompts, which I end up copying from project to project and pasting from window to window. So his point is you set this up, you don't copy and paste your prompt in, it sits there ready to run when you need it.

And he likens it to, um, like a standard operating procedure, something that you would run through with a new hire. Um, but also when you set up a skill, it can actually fire itself. So you might be giving it like a verbal instruction and it will pick up on what it needs to do, uh, based on your verbal instruction.

So he talks about it like this. He talks about the anatomy of a skill being, um, the description, which is the label on the folder, which is what Claude uses it for and people don't necessarily pay attention to that bit. And I'll get to why in a moment, um, the skill.md folder itself, um, which is like the text file, which is like basically your long prompt as far as I can see. So that's the steps in order. The main set of instructions, it's a super detailed prompt. Think of it as the decisions, what looks good and what it must never do.

And then he talks about the reference files so that they're the supporting details. They're not necessarily read every time, but they can be pulled in when the job needs it. So if you've got your brand guidelines, your style guide or whatever it is, if you're using that for the skill, it would sit there.

Um, and I paused on the screen where he ran through really quickly his skills and saw some of the examples that he has set up in his blog. So he has a content strategy boss with a funnel, keyword analyzer, connects SEO data, content brief builder. Obviously he has worked with SEO clients, but you would build your own skills based around what you need to do.

Um, one of the ones that I have built is one that builds out my invoices when I need invoices built. And he says, you can schedule them to run and Claude does them automatically. So if you have like a job you need doing at the end of every month, which is like, okay, take the sales data and then update the spreadsheet with the sales data and then push that to my data studio or whatever it is that you're reporting in.

Anyway, really good video. Enjoyed it. The three mistakes that he sees people making is, um, skipping the run through.

So basically they build the skill, but they don't train it like you would, which is like, try this, try that, identify what works, but really don't just put it in because you know what the, the quality of what you put in is, is going to influence the quality of what comes out. He says, don't make one skill trying to do five jobs. Um, there's more room for errors within that and for the output not to be great.

So break things into separate skills, just one particular task or job with a realistic outcome. And finally, he said, make the label very specific so it knows when to invoke the skill. So anyway, I really liked that video and if I had time, I would do his course and pay the money for it, but I don't have time.

[Darren Shaw]
You love courses.

[Claire Carlile]
I love a course. Well, Darren, I love a course.

[Darren Shaw]
You really love a course.

[Claire Carlile]
Yeah. I love a course.

[Darren Shaw]
I like to learn from like the shortest LinkedIn posts. It should give me the shortest little bit of information I need in order to execute. But yeah, skills are great.

Use them at, uh, in my own work quite a bit. I've made a few skills and, uh, yeah, the run through is critical. You gotta like, I've, I've, I've had to really browbeat Claude into getting my skill right.

It takes time to go back and forth back and forth. Um, my first thing I want to say is that I'm excited about our upcoming webinar, Claire. Uh, so this will be on Tuesday, November 10th.

If you're listening to this, we just published the podcast. Like it's probably happening right now. You, if you're listening to this, we're probably doing the webinar right now because we published the podcast on Tuesday morning and then the webinars on Tuesday morning ish.

So, uh, we're basically going to publish the podcast and then be doing the webinar within about an hour of publishing podcast. So if you're listening right now, pause this listening, go to the white spark channel, go to the live section and join us for the podcast. I mean for the, uh, for the webinar.

So webinar about local ranking grid is going to show you all the awesome things it can do to help you in your agency and your local SEO life. And, uh, if you are listening to this podcast, well, guess what? The recording is there on the YouTube channel.

You can still go and watch it. What? Okay.

So, uh, my first link, I got two of them. I got to combine them. So they're just, they're just so quick.

I just want to say what they are. I think they're very interesting. So Mike B.

Snow said he's noticing, uh, that local packs seem to be more commonly placed above the AIO now. So, um, I like that very much because I find that the AI overview version of local results are, well, first they're completely randomized. They don't seem to be based on which business is actually the best business.

So the trashier results, um, and they're AI ick. They just something off about them. I love the local pack.

Maybe I'm a stubborn old local search curmudgeon, but I prefer the local pack. I think it's a better experience. Uh, Ben Fisher commented on this post.

Someone, uh, was like, Hey, my map pin is in the middle of the ocean. And then Amy and Ben Fisher said that this is a common bug. People know about it.

I said, Oh, well, if you set your whole service area to be United States, that happens because Hawaii pulls your map into, uh, into the ocean. Um, but then Ben Fisher, ding, ding, ding, this guy's a diamond, uh, GBP product expert. So he loves to like drop little hints about stuff that's coming.

Claire's face. If you're listening to the fire, Claire is like, Oh, geez. Uh, okay.

But he says there's some major updates coming. We hope to SAB soon. Hopefully, it's like a major hint that some of the BS happening with Google business profiles and service area businesses.

This to me indicates, he knows that Google is trying to improve the system. So that's all there is to really say about it is that sounds like people at Google might be working on fixing some of the difficulties that SABs have. So I just wanted to highlight that.

And that's all I got to say about it. And Claire, do you have anything to add?

[Claire Carlile]
No.

[Darren Shaw]
Excellent. What's your next link then?

[Claire Carlile]
My second link and is a biggie. So I won't really bore you with it too much, but it's not boring. It's just like whenever a lady writes anything is going to be a big post, but it is going to be worth reading.

Also, interestingly, it comes with a worksheet. Obviously I love a worksheet the same way. I love a course, but also with a Claude skill.

Whoa, you can use other people's skills. So this one is all about citations. It's about third party citations.

And, you know, we're in this position where we're going, well, you know, where do I need to appear? How do I need to appear? I want to influence citations and recommendations in the AIs.

So, right. What do I do? Okay.

So short one in terms of, let's just think about some of the things that she talks about. So she talks about formats to maximize. So she gives 13 high value types of citations.

And so I highlighted, well, a few that I think that are going to be relevant for local businesses, obviously independent firsthand reviews and tests, evidence led comparisons and recommendation shortlists. We're seeing more and more and more and more of these than ever, especially in the world of SAS, but this is relevant for anyone where there is a decision making process that needs to compare competitors, practitioner demonstrations and task led tutorials, authentic community Q and A and troubleshooting discussions, editorial coverage, an original reporting specialist, expert commentary, data led articles, detailed customer reviews, and maintained profiles, et cetera, et cetera. Industry directories, independent databases is worth going in and finding the full list from her. Even if you're not going to use the framework, you should be looking at this in terms of this is my complete list to all of the marketing things that we actually need to be thinking about.

She also goes through some of the low value, risky types of AI crappiness to avoid. Again, another one to look at. Another thing to use this for is if you are working with someone who is offering you, I don't know, A I E O E O E O, optimization services, then make sure that they're leaning into the activities, which are going to be high value, not the crappy ones that are going to be crappy value.

[Darren Shaw]
Yeah. Citations. I'm actually, we just discussed it on our department lead call this morning.

I'm just like, we need to improve the local citation finder. And so a lot of that functionality is going to get pulled into that software where we will help you identify your AI citations. Also related.

No, yes. I read two interesting articles, which I did not link in this, but I'll just quickly mention them because they're related to this, is that I read one post where the guy was a case study and he, what he did is he like cleaned up his website, got a bunch of pages out and then he took a lot of that content and started, he published 570 posts on like social media, YouTube comments, Reddit, like he just sort of called them external. He didn't define his sources.

Right. But that's what we're talking about here. Right.

It's like, and his LLM mentions went through the roof, right? He had a huge boost in being recommended by LLMs. And then I read another post where the guy said, we did none of that crap. All we did was just clean up our website and our LLM mentions went through the roof.

So which is it?

[Claire Carlile]
But also how are they measuring? Blah, blah, blah. What are the mentions?

You know, it's all a nonsense. So let's just lean into good marketing and not silly measurement.

[Darren Shaw]
But how do you know if your work is working?

[Claire Carlile]
Because you, you sell more stuff, you get more phone calls, people come in more to your branch.

[Darren Shaw]
Yes. You're getting recommended by large language models.

[Claire Carlile]
Yeah. You, you would see, you would hope to see an uplift in your actual difference making, um, actions that people can take on whatever you do.

[Darren Shaw]
That's all that matters, right? Oh, are we getting more leads through our contact form? Yeah.

[Claire Carlile]
Yeah.

[Darren Shaw]
All right. Well, let me talk about my actual next link, which is near media guys. They're smart.

Uh, Mike Lumenthal and Greg Sterling did an episode with just the two of them, no guest. And they talked about two things that I think are very interesting and worth sharing with our audience. So go and, go and watch near media's podcast episode that I'm going to link in the show notes.

Um, but also just like subscribe there. They're the fricking best. So make sure you subscribe to near media.

So the first thing that they covered on their latest episode was a study of reviews from a company called review flows. They tracked about 5 million reviews over two weeks. I was about 9,000 Google business profiles they looked at and they were tracking this whole thing about Google's deleting reviews, right?

And so they're trying to understand why are these reviews getting deleted? So out of the 5,000 reviews, 2% were deleted within one week of posting. So now they're looking at those ones, home care services, like home, home services, like plumbing, furnace repair, that kind of stuff.

Those and healthcare had the highest removal rates. Uh, hospitality had one of the lowest hotels and maybe they include restaurants in that one star reviews were removed at about three times the rate of five star reviews. I find that interesting.

So they're removing more negative reviews than positive reviews. Kind of surprising. Really.

I thought this was a whole like big review thing, but maybe that was because of review sabotage, right? That could be that, um, profiles were 85% or more of reviews were five stars. So roughly double the removal rate.

So that's interesting. Um, and similar to how Yelp works it. If you're a first time reviewer or you have like low history of reviews, those are much more likely to be removed than, uh, reviews from people who have an established track record.

That of course is a reasonable thing for them to do with the review filtering review length, images, mentioning employees or stock phrases like highly recommended. These are things that people think have an impact on whether your reveal get removed or not. They did not find any correlation with that.

So that's interesting to me. So if you look at a bunch of reviews that got removed, um, these things like, you know, putting the staff name did not actually seem to raise removal rates despite the new policy. Google said they would do that.

Um, the biggest factor that they found was review velocity. This is fascinating. So it's very useful to know that if you're getting reviews at a rate of like one per month and then all of a sudden you get like 40 Google's like, yeah, we see, we know what bullshit you're up to.

And then they delete those reviews. So you people that are just sending out an email blast twice a year to your mailing list, say leave us a review. That's just a really bad idea.

The review filter is designed to, to flag that. So don't do that because you know what it looks like. It looks like you bought a big bunch of fake reviews.

So just don't do it. Um, and a five star review that got an owner response within 48 hours was about half as likely to be removed. So respond to reviews, don't send them in bursts.

And that's mostly what you need to know. Uh, also, uh, they talk about Apple maps and Siri. So Mike, he's a big Apple fan.

He says Siri sucks a lot less since this 27.2 update. And so, Oh look, Siri is asking me if I want a thing. No, I don't.

Uh, I, I didn't, I wasn't actually talking to you Siri, leave me alone. Uh, so it now handles web and local search pretty well. What does this mean?

It's a hint that I think in the not too distant future, Siri could become one of the more significant players in LLM usage. Because like right now it's like the, the, the battle between chat, GBT, Gemini and Claude, that feels like to be the battle and like no one can speak to Siri. But frigging Apple iPhones are the most prolific device on the planet.

And so if everyone just finds it easy to ask Siri for stuff and Siri does a good job, I could see some shifting happening there and Siri could be an important platform. What do you need to do? Well, a hundred percent of your local business you need to keep your Apple business profile.

If you haven't claimed it, go and claim it. Fully optimize it. Fill out every field and keep it fresh with fresh content and photos.

Try to earn Apple ratings. Apple ratings are little thumbs up. Get some thumbs up from some of those people.

See if you can get some of those you're going to drive reviews. Reviews on Yelp also help with that. So anyways, great podcast episode.

You should go and listen to it. Those are some of the highlights. And that's all I got to say.

[Claire Carlile]
Oh, Apple. Apple Maps is one of those things where I'm just like, need to do that, need to do that. And then you do it.

And then you realize why have I suddenly got eight duplicates suddenly of the set? It just seems like such a, the infrastructure seems quite broken in terms of the backend, which is why I keep thinking it's important. And I come back to it and I'm like, this is worse than Bing.

[Darren Shaw]
It sounds like you're talking about Bing. Bing is absolute trash, but I have not experienced the same thing that you're saying on Apple.

[Claire Carlile]
Yeah. Yeah. All right.

I'll go back in and have a look. Right. My final linky link is another one that I picked up from Women in Tech SEO, the newsletter.

They always highlight corkers. So I really like this. It is the risks of optimizing for AI over users by Andrea Schultz.

And all the way through AI Mageddon, it has made me just go, do we did this already? Come on, everyone. We did this already.

Can you not remember all the spammy awful stuff that we used to do for SEO? Can you not remember those terrible times? This is just another iteration.

So it's a little bit of a warning. Well, not warning because we should know this anyway from her. So she talks about, as marketers chase citations and visibility in AI search, some are repeating an old SEO mistake, optimizing for search systems rather than the people who use them.

And I know it is a bit fluffy to say, you know, human centric, customer centric first. Obviously we need our websites and information to be findable, passable, indexable, or whatever, whether it's by Google or whether it's by AI agents, but we do need to be appealing to human beings. So it's another call to arms for reminding us about this.

So, you know, we've already had horrible list of calls that were getting us, uh, to the, to the right, uh, with loads of mentions in AI, um, in the LLMs. So she talks about AI search visibility obviously matters, but it shouldn't come at the expense of useful original content. So let's just remind ourselves what that can look like. So she talks again about, um, how the winning pages in search tend to include elements that AI can't independently produce.

So original experiences, unique industry insights, opinions, and viewpoints, proprietary data, expert analysis, and real use cases, which is what we have been referring to as non commodity content. Um, so she talks about human first search optimization, which she gives four pillars. Now everyone appears to love a pillar at the moment.

Um, so here, uh, here are the pillars experience. So providing firsthand knowledge, lessons, learns, implementation, implementation, challenges, real world observations, blah, blah, blah opinion. So taking a click, this is what everything just looks exactly the same because all of these websites built with AI and everything, which is just built to like one be a quick hack and made with AI, or I want to build something that will give me visibility.

Everything looks the same. So there is no opinion. There's no clear positioning, um, or interpretation.

So we need to work out what our voice is, what our positioning is, um, you know, what we actually stand for. We talk about this every week, but you know, people still aren't doing it proof. So supporting claims with original data and evidence such as customer reviews, internal research, direct experiments, and identity.

So establishing trust through real people, the number of people that I've seen recently that don't have an about us page. I mean, you talked about ours being bad, but at least we have one. I'm not going to buy even from a SaaS, but I've worked with a SaaS that said, well, we don't need an about us page.

No one cares about that. And I was like, I would not buy from you anyway. Really nice piece.

Just worth dipping back into if you are interested in humanity and um, maybe the human race staying, staying alive.

[Darren Shaw]
Hey, you like humans? You should read that post.

[Claire Carlile]
Anyway, that's my third link. It was good.

[Darren Shaw]
Yeah, that does sound good. It sounds like we're beating the drum again, but it's, it's, that's what we all need to hear, right? Frickin tactics, short term tactics.

No. Hey, you know what? I already mentioned this piece.

I didn't realize that I was one of my links, but I already talked about this guy who pruned his website, cleaned up a bunch of content. Uh, they moved. Uh, so this is Jan Willem Bobbink.

I'm sorry if I butchered your name on LinkedIn. Uh, he, he put this post and this is the one where he did that I mentioned earlier where he did the 570 pieces of external content, which is around 27 a day, including social media posts, articles on other platforms and community answers all connected to the same smaller set of pages. So they cleaned up the website, merged a bunch of stuff.

So this content pruning is a pretty big deal. I actually also am going to link to, and I won't go into detail about it, but Corey Huber, Huber also put your name, uh, Cori, you probably seen this guy around. Uh, he's been SEO for a long, long time, but he talked about Google's ranking signal distribution system.

And he's talking about when you have a bunch of pages, you're diluting, and there's an actual thing in the patents about this, where you just, Google will not index all of these pages. So content pruning can be very valuable. You get rid of all these extra pages.

You focus in on the, what the core pages are, um, by getting rid of the cruft and all your SEO goes up and your AI goes up as well because of all that. So within three weeks, this guy says his chat GPT clicks rose from 12 to about 1100 per day clicks from chat GPT. Is that what he's saying?

I'm so doubtful. This is a bullshit post. Oh, wait a minute.

You're getting 1100 clicks per day from chat GPT.

[Claire Carlile]
My SEO went up and my AI went up.

[Darren Shaw]
Yeah. Anyways, it's still good work you should do. I'm curious about these external pieces.

Um, someone in the comments asked for more info about that and he was pretty cagey about it. He was like, man, I'm not telling you. He's like, it's all, it's all dependent on your industry.

Like you got to find the ones that are important for your industry. But reading between the lines, it's Reddit, other forums, YouTube comments, LinkedIn, other social sites that get mentioned in chat GPT for your specific query space. So you got to do a little digging to like find out like what are the places that are often getting mentioned and then be there like sometimes Pinterest, Instagram.

Anyways, uh, do read that. If you want to have your SEO and AI go up, read that, do the same thing he did.

[Claire Carlile]
I want it to go up and right, up and right.

[Darren Shaw]
Up and to the right. If you want that, just read this post, do what he did and be a big winner. Yeah.

You want to be a winner in life? There you go. Don't be a loser.

Be a winner. Do this stuff. Yeah.

I think that's the podcast. I think we did it all right.

[Claire Carlile]
Finished. I mean, complete.

[Darren Shaw]
I will remind you all we have the Local Ranking Grits webinar. If you're still listening to this and you're not on the webinar, go and check cause we might, we might still be live. If you're getting this two to three hours after we published the podcast, then you can still go and watch it on our YouTube channel and you can see all the cool things you can do with that software to have to be a winner in life.

If you want to be a winner in life, go and watch our webinars. See ya.

[Claire Carlile]
All right. Thank you. Bye.

Bye. Bye. Bye.