Analytics
Dale McGeorge
Head of Product
This interview-style discussion centers on Google Business Profile (GBP) data and its effective use for local SEO. The speakers delve into the importance of focusing on meaningful metrics tied to business goals, such as calls, website clicks, and form submissions, while cautioning against relying on potentially misleading metrics like impressions.
They address a recent bug inflating driving direction requests and the inherent limitations of GBP data accuracy, emphasizing the need to understand data collection methods and potential delays. The conversation highlights the value of the GBP API for in-depth analysis and reporting, especially for multi-location businesses.
The speakers introduce a free tool that leverages AI to guide users in selecting appropriate categories and services for their GBP profiles. They also explore challenges arising from Google's evolving local pack display and the impact of Google Ads data on GBP performance, ultimately advocating for prioritizing first-party data over third-party sources for strategic SEO decisions. The discussion concludes with a rapid-fire Q&A session and an upcoming webinar focused on website strategies for local businesses.
In the realm of digital marketing, particularly search engine optimization (SEO), accessing and interpreting the right data is fundamental to crafting successful strategies. Google Business Performance data and Google Search Console provide a wealth of information that, when utilized correctly, can significantly enhance SEO efforts.
Google Business Performance data offers insights into how businesses appear in search results and the actions potential customers take upon seeing these listings. Metrics such as website clicks, call clicks, and direction requests are crucial for understanding user behavior and the effectiveness of your Google Business Profile.
1. Website Clicks: This metric denotes the number of times users clicked to visit a website from a Google Business Listing. It's a direct indicator of how well your listing drives traffic to your site.
2. Call Clicks: Unlike phone call counts, call clicks measure how often people click the call button on your listing. It's vital to note that this action doesn't guarantee a call was made; it simply shows interest.
3. Direction Requests: Especially relevant for physical businesses, the number of direction requests can reveal the local interest in visiting your store or facility.
Google Search Console (GSC) complements Business Performance data by providing a deeper dive into how websites perform in search results overall, not just from the business profile. GSC offers data on impressions, clicks, and keyword performance, allowing SEOs to optimize their site's visibility and content.
1. Query Data: GSC shows which search queries bring users to your site. This insight is invaluable for understanding what your audience is looking for and how well your content meets those needs.
2. Page Performance: Beyond queries, GSC illustrates how individual pages perform, offering opportunities to tweak or enhance underperforming content.
3. Click-Through Rates (CTR): By examining the CTR, you can gauge the effectiveness of your title tags and meta descriptions in enticing clicks from search results.
The real power lies in integrating Google Business Performance data with Google Search Console insights. This combined approach allows businesses and marketers to have a complete view of their online performance, from how users interact with their Google Business Profile to how they engage with the website after clicking through.
1. Combining Metrics: Look at website clicks from your business profile alongside site visits and behavior from GSC data. This can help identify whether users' expectations align with what they find on your site.
2. Analyzing Search Queries: Use GSC's query data to inform the keywords and categories you emphasize in your Google Business Profile, ensuring consistency in SEO efforts across your online presence.
3. Evaluating User Behavior: Assessing direction requests and call clicks in conjunction with page performance metrics from GSC can reveal insights into the customer journey, from discovery to action.
While Google's tools offer robust data, marketers face challenges in interpreting this information. Data delays and anomalies, like bugs affecting direction requests or discrepancies in performance metrics over time, require vigilance and a critical eye to ensure accurate analysis and reporting.
1. Data Accuracy: Stay aware of any reported issues or anomalies in Google's data collection and reporting. Regularly verify data consistency across reporting periods.
2. Integration with Other Tools: Consider how these insights integrate with data from other SEO and marketing tools. A holistic view that includes social media, email marketing, and other digital channels will provide the most comprehensive insights.
3. Actionable Insights: Always strive to translate data into actionable SEO strategies. Identifying trends and anomalies is only beneficial if it leads to informed decisions that enhance your online presence and performance.
Effectively leveraging Google Business Performance and Search Console data is essential for any comprehensive SEO strategy. By understanding and integrating these insights, businesses can optimize their online presence, improve user engagement, and ultimately drive better SEO results. Remember, the goal is not just to collect data but to use it to inform decisions that lead to measurable improvements in visibility, traffic, and conversions.
Joy: Good afternoon everyone, or I should probably say morning for some of you. Dale, what time is it currently where you are? You're in Australia, right?
Dale: 7 AM, yes.
Joy: Oh my goodness. Okay. So, if you're tuning in from this side of the world, good morning. For those of us over in North America, good afternoon. I'm Joy Hawkins, owner of Sterling Sky and I would prefer to let our guests introduce themselves. So Dale, from Jepto, why don't we start with you?
Dale: Sure. so I'm the founder and head of product at Jepto. My role is to interface with customers, make sure we're solving the right problems, and then translate that into engineering talk. been in the industry for 25 years now and really loving getting into the data science part of digital marketing.
Joy: All right, and Noah?
Noah: Hey! I'm the Director of Innovation at Sterling Sky and this is a total hoot for me. I got to know Dale in 2019 when I was running a show called Agency Automators. He was our second or third guest and completely blew me away with how he thought about marketing and how to leverage like data and APIs and automations to make all that stuff happen. And that was like at the beginning of his Jepto journey really, and we've collaborated a ton on feature sets inside their tool in terms of like things to think about next, whether it's like search console data or all kinds of functionality around Google business profile stuff.
Joy: Well, I have promised our audience, something that we frequently run into working with Noah, his brain exceeds the average person's brain and sometimes when he gets talking about things, he's rolling his eyes, we're all like, "What are you talking about? Like you're just basically way too smart." And I would put Dale in that category as well. So, if these guys say something and you're like, "I don't have a clue what you're talking about," do you let us know in the comments because I'll make sure we db it down a little bit. That's going to be the challenge for you two today, for sure. Noah's learning, he he tells me all the time when to to let him know if he's talking over my head but that's what happens when you just have people that are really smart. So, we're hopefully going to db down some of the issues that come into play with tracking when it comes to Google business performance data, search console data, possibly GA4. these are all different data sources, it can be really confusing trying to make sense of them. so let's start with a simple question. I'd like to hear from both of you and I guess Noah we'll start with you. What metrics are the things that actually matter? let's start with that question first.
Noah: So in Google Business Profile, there's tons of different things that you can measure or metrics. And the ones that we really like are the ones that align with people's business goals. So for most of the businesses that we're dealing with, it happens to be website clicks or call clicks. And there are other metrics that matter whether it's bookings or whether it's direction requests. But the value of these different metrics really depends on your business and they also depend on whether or not that particular metric is suffering from a bug in the moment. generally website clicks and call clicks are the most valuable. But a lot of these come with like nuance and interesting things about them that that make them either more or less valuable. For example, like the call clicks, that's not a measurement of phone calls. That's a measurement of how many times people hit the button that then opens up the interface on their phone that reads to the user, "Are you sure you want to call this this phone number?" which Dale gets me to the power of UTM parameters, right? Like, do you have any thoughts on that?
Dale: Yeah, I think that understanding the way that different metrics are actually captured really matters, so that when you are reporting on that, you understand and can communicate the way that they're captured versus the way that they might be perceived. So the the particular example there with call clicks is great, and then the same with website clicks. So if you've got different types of clicks coming from say a business profile, then you want to be able to to tag that for appropriate attribution. and then one of the the things that you might want to consider is where else is that particular link focused. And then if you do have, let's say, a a syndicate of a a particular profile, being aware that you haven't just literally copied and pasted all those details across that can then cause issues with your data as well.
Joy: So let's talk about one in particular that has caused a lot of buzz recently, which is driving direction requests. So, there's a tweet, which I'll have put up on the screen here, that I tweeted out talking about this massive spike that we saw in January across all kinds of businesses when it comes to the number of driving direction requests. This number was inflated so high that there's just absolutely no way it could be accurate. what are you guys' thoughts on on what's going on there and what should people do about it? Let's start with Noah.
Noah: Is that me?
Joy: Yes. Yes, that's you.
Noah: So, I actually don't know. And just like you when I have crazy questions with the API, I turned to Crystal Tang from Uberall. And, when I reached out to her, her answer on this was, "It's a bug, they're aware of it. They're working on resolving it." That's all I know. And I've reached out to other people and that's all I can find out about it. And of course, like, I have a lot of connections on the Search Console side where I can reach out to people who are developer advocates, but I don't have those same relationships with GBP, so I don't really have insights onto that particular metric. It's super frustrating though.
Joy: Appreciate the honesty. Yeah, so if you are looking at driving direction requests and you're looking at the month of January, you can simply ignore those. Dale, are there any particular metrics you think, that are useless that people should kind of ignore?
Dale: I don't think ignore is probably the right word I would use for any metric. I think that if you look take it with a grain of salt and understand that it provides benchmarks and things that could cause your tracking to be inflated or deflated is important. But I think in general we know that the data is not 100% accurate, so we're not going to be making, you know, strategic business decisions off one metric alone. And that we look at that as a gauge of an increase in performance and correlate potential changes that you've made or that Google's made and then look for corresponding fluctuations in that data. And then if you have the ability to try and understand what caused that, then that would be a a really good undertaking, but there's a lot of things that you could dive deep into and won't actually ever understand. Like the driving directions, that that is a great example of there's nothing you can do about it. All you can do is just be aware that that is a thing within the data set. but I tend to just look at all of them as a way to be able to to just gauge performance as business as normal until something drastic happens.
Joy: Noah, do you have any that you hate? I do. So, I'm just wondering if I'm the only one here.
Noah: Well, I love the question useless because I wanted to hear you go off on it, actually.
Joy: Oh my gosh.
Noah: Totally!
Joy: Yeah, my my number one, if we're talking like all data sources, is impression data in Search Console, probably the most useless metric. Don't use it at all, or I tell people not to use it because of the just absurd amount of things that can cause spikes and declines, and the biggest culprit of those is rank trackers, but we see that all the time where there's these massive spikes out of nowhere in impressions and it's usually due to rank trackers. But there's also, like you mentioned earlier to me today, Noah, maybe you can give that example, of what you were seeing in Search Console a little bit earlier?
Noah: Yeah, there was this weird situation a couple of weeks ago and I actually think I tweeted out about it. About fluctuations with site link data. And one website that I had seen a lot of their data, they had tons and tons of content and I was like, "Oh, no, is this like the first client that I've experienced the helpful content, like slamming them? You know, that update slamming them?" And what we saw was that, they started to lose the count of pages that were ranking, was dropping like crazy. And their average position was, was getting worse in a huge way. And what was actually happening was that Google for a period of like two weeks was consolidated consolidating the number of pages that were ranking over that period of time, which just meant that they were losing all their site links. And so instead of one main result with four site links underneath, it was just consolidating to the main result. Same amount of clicks, that main result was ranking in the same position. It was just that they lost all their site links. And so that had the follow-on effect of showing a massive drop in impressions and their average position was getting a lot worse. But the reality of the situation is that, no, their average position really wasn't getting worse. It was just that all that site link data was going away. And then we saw it recover about a week ago, which was super interesting to me.
Joy: Interesting. All right. Well, let's segue to kind of the main focus of the, the topic today, which is we want to know, like, what types of tools and things that businesses, marketers, what not, can use to help track just how things are going? Like are they actually getting an ROI on their SEO? What what things should they be looking at? So I know you both have really cool stuff to share on this. So Dale, let's turn it over to you to get us off with that.
Dale: All right, so I'm just going to quickly present my idea of ways to take the performance data and then push that into an overall metric. So, what I wanted to achieve with this is that we've got all our GBP data, obviously, you know, we have a product to push that into BigQuery, there's plenty of others as well. And so, I looked at this problem and said, well, there's 20 plus different metrics, how do we make this into something that a business owner can understand and and that we can report on across different profiles. So, this is what I've come up with as a sort of, quasi sort of score that you can use across the business profile. there is a little bit of sort of sequel behind this and I won't go too much into the weeds into how we do this. But this is an example of what you can do with the raw data. So, the considerations here that I've got is that we need to be able to compare profiles across different locations. Some have high traffic, some have low traffic. We need a way to be able to to compare their performance against each other, search vole, one of the things that I've been looking at with the search keyword data that Google provides for business profiles is that you can use the amount of keywords and the understanding of which are branded and non-branded to get a good idea of whether you're ranking for a a bunch of different terms rather than just brand terms. So that is a really good, healthy sign that you've got , you know, good dominating keywords across that local area, not all performance metrics are equal. So in the way that I've created this particular example, I've weighted them and then you can obviously adjust your own weights as you see fit. And, make it simple. So two metrics. So this is what I came up with. So this is a review score. So what I've done is got looking at the location rating, that is obviously a very big factor. So I've weighted that to be 40% of this overall score. total reviews, again very important. But then also the velocity of reviews. So I'm looking at the count of new reviews. So you can set this over a given time period and track this over time. And then the average rating of those new reviews. So is there an incline or decline in, in those new reviews that might affect your overall location rating. so I've skipped one here which was the engage-
Noah: Dale! So so what happened to the other 10%?
Noah: That's a good question. I think that the formula is wrong, from the slide. I'll fix that!
Dale: Just keep me honest there, Noah. so, and then, I'll come to the slide at the end that kind of shows this in, in Looker Studio as a way to be able to actually, visualize this, the other one is engagement score. So this particular metric looks at the insights data. So we're looking at the impression metric. So all of the caveats that we just discussed around which metrics are good, which metrics are bad, I kind of bucket them together and then you get a good sense overall as to how that profile is performing. So the action rate is there specifically to be able to to do that comparison. So, if you have a high, you know, conversion rate of impressions to actions, then that will show up. That's why that one's so heavily weighted. The search vole, so this is the vole that Google gives you that your profile is appearing. So, that particular metric that Google will either provide a threshold, which is a number normally around 15, it can go higher, as to how many searches need to be done per month in order for you to get the actual search vole. So if you're below 15, you'll just get the threshold. If you're above, they'll give you an actual number of the search vole. and then the keyword count. So, as I said, the spread of how many keywords that you have are is a good indication of the guess the prevalence of your particular profile in the local area because people are finding you for not just your brand term, but all of the actual services and categories that your business might fall under. so this is the sort of the smary of what that might look like in a report. There's obviously probably way too many metrics that I would actually advise on doing in an actual customer report. But this is an example of just sort of bringing all of that data in so you can either track this internally, and then you might just have an engagement score and a review score. So this was just my attempt into having, you know, some primary metrics that make up all of the underlying data set. there's a bunch of sequel to actually do this and I've got a blog post that explains this. So if you want to try this out for yourself, you can, and it uses all of the Looker Studio functions for dates so you can adjust your date parameters of your chart and then be able to change that that particular review metric or engagement metric over time. so that's that's my two cents on how to look at that data and and how to potentially make that useful.
Joy: So do you do you have an example of a business, like I'd love to know how you score, like did you actually score any particular businesses and and what were the the the takeaways from that? Like what would you do with that score?
Dale: Yeah, so as I said I like to use that score as a way to be able to compare a multi-location. So if you have multiple profiles and they're in different cities, you giving an indication of who is underperforming and overperforming. And that way you can really push your attention to any of those that are underperforming. Often, we kind of focus on the ones that, you know, might be a high-value to say the agency because they're paying you a fee. But there might be a better bang for the buck if we can lift some of the lower ones up. And then and then the ones that are higher, we can kind of just let that settle. Or we can do comparisons on do we have the same categories and services that are across those particular ones or the attributes so that you can understand what is actually creating that performance, and then is it just population alone. And then if you do have different websites or the one website, you can be comparing the local pages that those particular websites or website pages might link to from the profile.
Joy: So how would somebody go about getting a report like this? Like is is it something that you guys do at Jepto that they'd have to be a customer to get? Or like how would how would somebody run a report like that?
Dale: Yeah. So the we have a service that can pull down all of the GBP data and then push that into BigQuery. Very simple, low cost. You can also download that directly from the GBP dashboard into a spreadsheet and upload that into Looker Studio as well. It's a completely free option. we also do have a free tool for, basic metrics as a Looker Studio connector as well. So, plenty of options, and if you're interested, obviously we can discuss the scalability of that as in how do you take that across a large scale. we have the ability to use obviously BigQuery as that engine which provides us with that fast reporting, which is really the biggest issue with Looker Studio in that once you start crunching that across larger time periods or multiple profiles, it can be quite laggy if you're live fetching to the API. So by pre-fetching, you cut all of that out and add the flexibility of transformation and creating your own custom metrics.
Joy: So tell me more about this free tool. Where can people find it? Can we get that link, maybe send it to Richard to put up on the screen? So so everyone can check that out.
Dale: Yeah. Yep, I'll so just on the main navigation on our prepared for, for it's on jpto.com, and then under the main menu under features, you can see the two free tools there.
Joy: Awesome. Noah, I feel like you were going to say something.
Noah: Well, I love, I mean, what's funny about this is that we discussed this, these metrics on a call yesterday or the day before. And so this is, like, brand new. And this is kind of fun. Like Dale, if we could ideate for a minute, I would love to see that score and then maybe have AI insights showing which of the contributing metrics was the problem and then giving the user insights about what the next steps would be to, to sort of investigate that as an action. I think that could be useful for business owners.
Dale: Yeah, my brain goes to how how am I going to do that? so it's there's probably definitely a problem that we can take away and and work on because the I think the smary of multiple metrics becomes, complex in that people might have different objectives. And so what I really wanted to do with that weighting was give people the opportunity to change the weights as they see fit. Like, there is no golden solution for performance or reporting. Reporting is a very important part of relationships with stakeholders or clients and you need to find the right balance between your audience and what they can interpret the data and understand from that, along with any story that you might tell with that, as in we're doing x, and this has caused y. Now, correlation and causation, as everyone knows, is, is not the same thing, and that if you can have a very honest discussion around that, then I think there would be a really good case for relationships, which is probably the most important part for an agency is taking that raw data, making it meaningful to them, and conveying the value that you're bringing.
Noah: Can I, can I address a question that that Ben asked? Because I think it's, it actually is a great segue right here. Is that okay, Joy?
Joy: Mm-hmm.
Noah: Ben Milimas, "How difficult is it to get GBP API access?" Well, we can see it on the screen. great question. there is a form that you can fill out to request, API access. The process generally takes about five days to a week, and then, once you get that API access, then you need all the engineering chops to interact with the API directly. And, in terms of, you know, they, they provided a number of different what are called client libraries, so you can interact with it with a bunch of different programming languages, whether it's like Python or JavaScript or Go or whatever. Those are different languages. in terms of running a foul of suspension algorithms, you really just have to obey their different terms of services.
Joy: Yeah, I can definitely comment on that. I've seen companies lose their API access who are violating Google's guidelines. They're doing so usually by creating, like, fake listings, lead gen stuff, like, all kinds of, you know, shady things that Google tells you not to do. So you do not want to be running, or using the API if you're doing any of that stuff, but I think the other thing with the API that I've seen done wrong is I've seen some actual, I'm not gonna name them, but I've seen popular tools that everybody uses in the SEO industry, use the API in ways that I don't think they're intending to, because I've seen people delete listings out of a, a tool, and it actually deletes their profile. So stuff like that, I feel like implementing it is really important and I would say that's probably the number one thing. If you're implementing a tool, make sure that people can disconnect from your tool properly and it doesn't do anything crazy to the actual listing itself. That's, one of my biggest fears with hooking up listings to the API. I'm not calling Dale out. Sorry.
Noah: There's another great question, which is, "What are the benefits?" , and Kerry's answering it at the same time, but I mean, for me it's the ability to aggregate performance across locations, and it's also access to more information. And it makes it significantly easier to grab information like review data, which generally is really cbersome to grab any other way. Or Q&A data, also when we're looking at search keywords. If you're, if you're stuck inside the user interface, you can see whatever the top 20 or 30 or 40 keywords, whereas if you're going, via the API you can get many thousands of different search terms per GBP location. So, let's say you, you're a huge corporation that has many hundreds of locations, you can get thousands and thousands of keywords for each one, which is really neat. So you can kind of see information at scale across an entire enterprise landscape, that way. So it's, it's more information aggregated together, and it's information that you control. And what's really neat about it is that you can put the data up into a database like Google BigQuery, and what that means, which is really neat, is that you can then start to use tools like Google Looker Studio that are free, and you can use them not just to like look at some numbers on a page, but you can start to use that tool as like a tool for analysis, meaning it works super, super fast.
Joy: So, Noah, can you give some examples? I know you have some queued up of just like ways that people can use the API to get information on things that maybe you wouldn't know otherwise. So like in particular, like categories and services, those are really insightful things. I know you've built some stuff around that. Can you share a little bit about that?
Noah: Yeah. And actually Dale built a really killer tool, so that would be great if he could show that. But one of the things that's really, really neat via the API is you can use the country that you want information about, and you can grab all of the categories in your main language, , you know, whether it's English or Spanish or French or whatever, and you can grab all of the categories that are available for that country. And then based on that you can grab all the services and all the attributes that are available for each one of those categories. And that can inform strategy, that can help you understand, like, "Hey, I'm a plber, what am I missing out on in terms of the services or attributes that I should be taking advantage of if I want to rank for plbers in my market?" Does that make sense?
Joy: Yeah, is that one you have a visual on, or no?
Noah: Dale does.
Joy: Okay, perfect.
Noah: And it's awesome. He built this really cool free tool.
Joy: I want to know about this free tool, Dale. Tell us.
Dale: All right. we were literally up late last night, trying to get this to work, so, we, it should be there and live demos, obviously, so we'll see how we go. But if I rewind us to why we built this, the categories and services and the attributes matter. And what I wanted to do was try and have more of a, a visual interface as to how the different categories and services align. There is services that are repeated across categories, and we wanted to have a better way to discover those. So we built this, as a recommendation, tool. So, you can enter free text in here. So describe that, and then we use a, AI embedding model to, do the actual search. But might just show one. So you've got all of the different categories and services that Google define, in this list here. And so if you were to, to pick one, you're gonna zoom in onto that particular node and then you're going to get an AI suggestion as to the categories and services that are semantically relevant for your particular chosen text. So we've got a score here, and we've got the recommendation type. So whether it's a category or service. So if you've got a service, obviously, it looks at the service name underneath the category. if you were to say wanted to have a look at this, you can then click on that and then see where that is within the actual graph. So, if I just make this a little bit bigger and we're gonna play around with that. We can see that like all of these nodes here represent either a category or a service. So, the category is the solid circle, and the service is the, the hollow one. So you can see here how many of these particular services are related to another category. So you can really start to explore that and see sort of the patterns, within that. So if I just go back to,
Joy: So, I would love to, if people want him to to pull up, like if you guys are watching and you have a particular industry and you really want to see what it looks like, go ahead and leave that in the comments and we'll
Dale: Yeah. I'll do a free text one to show you that you don't have to just pick this one, you can just describe like anything to do with your business as well. So whatever the first one is, just yell it out, Noah.
Noah: Why don't we do garage floor painting?
Joy: Oh my god. Yeah, that's a good one!
Noah: Cause it's so random.
Joy: It is random.
Dale: So what's happening behind the scenes is we're turning that text into a an, embedding. So an embedding is basically a vector representation in numbers of that particular text or phrase, and then doing that semantic search. So, this is
Noah: So Dale,
Dale: Yep.
Joy: Keep, keep this up here cause I just, I'm gonna, I'm gonna try and and do my dbing down job here. So if I am working with a garage door company, I know exactly where Noah's going with this. So if they have a, service that, that paints garage floors, let's say. How in the world would you know what categories to use? Like you wouldn't, right? because there's no category for that. It's kind of a weird thing. So what this tool will do is when you type that in, it'll tell you, "Oh, you actually need the category flooring contractor because there's a service under flooring con- contractor that is called garage door, sorry, garage, garage floor epoxy coating. Like, how would you ever know that without a tool like this, right? That's, that's my takeaway.
Dale: Yeah, definitely. Yep.
Noah: Underwater basket weaving.
Joy: Okay, I gotta look at this one. That's just hilarious. Under, I don't even know what that is.
Dale: Underwater basket, is that a real thing? Underwater basket weaving?
Joy: I, I don't even know what that is. Somebody in the comments needs to, like, db that down for me.
Dale: Not sure Google knows what you're talking about.
Joy: All right, let's do roofing. That's a really interesting one.
Noah: Yes.
Dale: So we've got the option to pick one. I'm assing there is something relating to roofing anyway on the,
Joy: Probably.
Dale: But if you don't know, that would be where you'd put it in the
Noah: Yeah. Yeah. Okay.
Dale: Yeah.
Noah: Roofing contractor.
Dale: So, water damage. That's interesting that they have an underwater damage restoration service. Like, how does that relate to roofing? That's interesting. Crane service.
Noah: Well, what's also interesting is that a roofing contractor doesn't have a service, but there's roof installation underneath all those other categories.
Joy: That's really interesting.
Dale: Yeah, and you can explore that through the, the chart as well. Like, you if you kind of look around, they aren't placed on the chart, as in their semantic meaning, but you do get, if there is related services, you will see them grouped together. But the there is a lot of services and categories that just don't link to anything else. And you'll see that, obviously, all around the edges as well. So, all of those little dots that are around the, the edge, just mean that there is no relation to any other existing pre-defined category or service.
Joy: Yeah, and
Noah: Go ahead, go ahead. Super quick. The thing to take note of is the fact that this is not something that's static. These things change. Like, categories, services, attributes, like, all these things change over time.
Dale: Yeah. So I just wanted to quickly point out the little bit of the tech, again, try not to get into the weeds, so the, the model that we're using, is from Open AI, it's called Ada 002, Ada is the way that Open AI love to use people's names. So Ada Lovelace was the first recognized computer programmer, and so we will offer different models that you can then choose and see the different suggestions. So, some models might perform better for different languages. And that way we can add to this over time and then you can try out different suggestions and see whether the models produce different results. At the moment, this particular graph, is just showing English and the United States. So just to remind everyone that the category and, sorry, the the language and the location determine the categories and services that are available for the profile. So, I might quickly back to this one cause it does make sense, to talk about that as well, and that the relationship between the category, the service, and the attribute are inherently tied together. So the attributes are the combination of, as I said, the, the language location of the primary category, and then the services relate to the category as well. So, what this tool does is give you the ability to find all of the attributes and services for a given category, and then vice versa. So upon selecting an individual category, you can then see what are the pre-defined services and attributes available for that primary category. And likewise, you can go back. You can say, "Okay, given this particular service, what are the categories that exist for that service? And therefore, as a flow on, what are the attributes available for my profile, given that particular category?" And so you can filter and, and experiment with that.
Joy: If people want this, that, that QR code that you have up on the screen, they go there. That takes them to the the website. And can you clarify, is this try for free? So is there a limit or
Dale: It's totally free, yeah. So under the main menu, you can get here just through over here. we've just built this as a way to experiment with categories and services, and then you can also change languages. So, we've got initial support for different combinations of language and location, and we can add to this based upon feedback as well. So if there's, we tried to cover a few sort of more popular ones, but if you have another one then please, just reach out to us and request that and we can add to that over time, we kind of built this very quickly and so, we haven't had a chance to get everything sort of in there that we would like for now.
Joy: Oh my goodness, this is amazing, guys. Like as far as keyword research goes, trying to figure out how Google connects different topics and ideas, like there's just so much you can do with this information, from a strategy perspective. So, there's
Noah: Do you remember, we got lots of questions. Do you remember the Looker Studio template that I made that had, that had, like, category service attributes in it, Joy? He was, like, "Oh yeah, that's kind of cool stuff." But I think I can do better. And I was like, "Okay, fine."
Joy: Oh yeah, I, I use your tool all the time, Noah. So we, we use it for audits, and things to find opportunities for things that you're not currently utilizing. So super practical. I'm gonna save the questions till the end because we do want to get to them, we got a ton of questions coming in. But I know, before we get to those I wanted to make sure, Noah, you had some other visualizations to show, as far as using the API to help with reporting specifically. So can you show us some of that stuff?
Noah: Yeah. Do you want me to share my screen or do you want to do that so that, I don't really care. I can share, share mine.
Joy: Yeah, go ahead and, and share your slides, that'll work for me.
Noah: Okay, cool. Great, okay. And then, how do I, how do I share this as a, oh my god, I am so lame. There we go, okay.
Joy: Noah, no, hit slideshow.
Noah: Slideshow. This is embarrassing.
Joy: I know. I've been here all week!
Noah: So, this is an example of how to leverage GBP data. And we combine it with GA4 data at Sterling Sky when we want to do reporting. We want to keep stuff like super easy so our clients can understand what the heck they're looking at. And, Joy is really adamant about keeping reporting simple and really adamant about focusing on what matters. And things that don't matter are vanity metrics. The things that do matter are actions that are tied to business goals. And those things are phone calls, form fills, texts, chats. those are generally speaking the, the kinds of conversions that we want to show value of our work with. And so, we're going to show those over time and we're going to show trends by showing things like 15 months worth of data often times. And we're going to show how things are coming in from different organic channels, whether it's Google Business Profile or Google Organic. And this is like how we're leveraging GBP data that's being pushed to GA4. And then when we show it as just GBP metrics, we're showing, generally speaking, things like website, clicks. Not generally direction requests, but in this graph we're showing direction requests. And we're also showing GBP, call clicks, which again are not actually phone calls, they're clicks on that button to to initiate the process of making a call. What's really neat about getting data from the API is it enables you to also have being able to show data by city, which is really neat, or specific GBP locations. So if you're dealing with clients, or or you're part of an organization that, that has some scale to it, that has a number of different GBP locations, it's pretty cool cause it enables you to be able to say, like, "Hey, we're doing really amazing in Scottsdale, but we're really underperforming in Pueblo," or whatever, whatever the, the different, ones are. And I think what's really neat too, is we can talk about how to leverage the search part of the endpoint. Is that okay if I talk about that, or do you want to, do you want to hover where we just were for a minute?
Joy: No, it's okay. As long, like, I look at this and I'm like, "What does that mean?" Methods, locations, it's like a bunch, a bunch of Greek.
Noah: Oh, no. I, I love it. So, with the API when we're grabbing our search keyword data, Dale and I had a, had a discussion the other day. We were trying to understand like, when you look at the search vole metric that comes back via the API, we were trying to understand, "Is it via search and maps or just search?" And the docentation makes it pretty clear here, where it's like, it's , the search keywords used, for people trying to find a, a business, in search and maps. And that's why if you look at your search keyword data and you're confused about why some of those terms are in there, it means that potentially someone searched for another business that your business is in the same map window. And they might have hovered over your location to pull up your pin point, and that would have given you an impression for that particular search vole. And, and the reason I'm bringing that up is that when you make that API call and you get the data back, you're gonna see all kinds of stuff that you're thinking like, "I have no reason to rank for any of this stuff." , stuff that's completely topically not relevant, other brand names, stuff that you would never think you'd see yourself in the, in the local pack. Why is this stuff valuable? Is that it enables you to do lots of different types of analysis on your business. And, it, and it's even something that you can use to make really interesting, and I would only look at this stuff directionally, to understand like how much revenue might be at play if you're thinking about building out GBP locations. You know, we got reached out to by a brand where they wanted to understand the value of building out department listings for, for different, , different departments. And so, we were able to take something like 1.6 or 1.7 million different search terms, and throw them, each one of those different search terms, from hundreds and hundreds of different GBP locations into different categories. And we did that using, a machine learning model, and what's crazy is I was able to build all of that machine learning with the help of ChatGPT, which is shocking to me. And I was able to, like, work on the model and make it better and better. And what was neat was that I could put all the search terms into these different buckets that we were thinking of building out, GBP , department listings at scale for. And based on that search vole, that 251 million searches on the left, we were able to determine, direction requests. And thankfully we were working with good data, IE, before when, when all those challenges happened. And we had a sense of what, what that company's in-store, , conversion rate is when people come in. We also had a really clear sense of what the average sale vole was. And we also knew what the break down of searches were based on the API between mobile and desktop. And then we had these sliders up top left so we could determine like in a good scenario and a bad scenario, what percentage of people would come in via mobile, what percent would come in via desktop? And then we were able to kind of put revenue numbers that are only directional. They're, they're not meant to be accurate. And, it was just like a neat exercise. You can't do this without data that's coming out of GBP. You can do stuff like this with search console data, but you can't get granular down to the, to the location level, again, this is directional only, but it's the kind of analysis that you can't do at the local level unless you have data that's coming out of the API, which I think is really neat, and then Dale and I quickly, oh, almost one last thing. Dale and I had kind of a competition to see who could build a better model, which is really fun. And I had done it by doing something called supervised learning, which is where you basically label, I think I labeled like 8,000 search terms, which is insane. I completely wasted them. But, I labeled like 8,000 different search terms, like, "Hey, this goes in category one, this goes in category two, this goes in category three, category four, whatever." And then used that to train the model to then categorize out 1.6 million terms, and it was really fascinating to watch how these, Open AI models competed against the supervised learning. And, we both learned some lessons along the way. And, you know, the takeaway there is supervised learning wins.
Joy: I have to tell you, Noah, that spreadsheet is where you just lost, kind of went over everybody's head, I guarantee it. For those of you that made fun of me for being a nerd, this is what Noah does for fun. He builds things that I don't even know what they are, I get made fun of on a regular basis
Noah: And then I, and then I do huddles with Zoom at 9:00 at night, my time, with my wife. Like, "What are you doing?" Nothing.
Joy: My, my spreadsheet for Mario Kart is like not as nerdy as this. I would just like to say that. So, my role here is to basically tell everybody why, they should shoot holes in some of this data. you guys are making it sound awesome. It is awesome. I just want to play devil's advocate for a second and share a couple examples of things people need to know before they completely trust it, cause there are, there are flaws in some of the data points. So, I'm gonna share that. So, this is a perfect example where, I was looking at this earlier and Mar Vista, okay. I'm not, like, an expert on California, neighborhoods, but it's like a small neighborhood in LA somewhere. There is no way that these are real han beings searching this stuff. Like, "Mar Vista Condos real estate agents" is not one of this business's top query. This would be what we call bad data. But it's clearly, coming from bots. Coming from rank trackers, right? And how you can tell is, like, how similar some of the numbers are, and the patterns are similar. We see this all the time. So something to keep in mind before you're using this data to pick all your keywords, is if you start to see patterns like this that don't make any sense, there is a good chance that a rank tracker could be picking up on things and, and skewing all your data. So that's just something for people to be aware of, another thing that I feel like I, I definitely want you guys to comment on because you know a lot more about this than others, when people are looking at date ranges and they're comparing data, especially like in the current month. I don't think a lot of people realize that there's a delay in Google business performance data. And I've been told the delay's three days, five days. I pulled this screenshot this morning, it only goes up to I think February 19th or something. why is there sometimes like a difference in the days and the number of days that Google is behind on this data?
Joy: Do either of you know, because I don't.
Noah: I don't either. And I've reached out to several people about this, and the delay seems to be changing over time. I thought it, I was seeing stuff that was like a, a day or two old, and now it seems like there's anywhere from a five to 10-day delay. What do you got, Dale?
Dale: Yeah. On average it's four days. some can go longer, though, yeah. And, and it's not uniform. You'll see some metrics can be pulled from a four-day period and some might lag to say six. There's no consistency there. It's varies all the time.
Joy: And this is a good question that came up in the, in the chat about this. Just to clarify, I'm looking inside the dashboard. Is it the same in the dashboard as what you get pulling from the API?
Dale: Generally yes. The, the main difference if you are using the API is when you last did that sync. So if you do a daily sync, which most people do cause that's as granular as you can get for most metrics, then you, depending on the time that you ran that versus when you look at it in the dashboard, may be a slight difference there. But if your tool allows you to do a refresh, then they should match. Yes.
Joy: Okay, perfect, one other thing to keep in mind, guys, is if you're running Google Ads, we have an whole article about this that, we can share in the chat. But essentially, Ads data shows up in here, especially if you're running Performance Max campaigns. So this was a storage company that has all these queries for restaurants, and coffee, and it's like, "Why would those be the top queries?" we break it down and explain it in the article that we'll get shared in the comments here, but, essentially, keep that in mind if you are running Ads, it can really skew your Google Business performance data. It does pull in data from Ads, and I feel like that is a very uncommonly known thing, cause I've gotten this question several times before.
Noah: So, I don't know if you've seen this, Joy, but in the questions, tons of notes about direction requests. And that the, like, the takeaway from that I saw from all the messages was that, we should expect direction requests to be down, but that's not what we're seeing, right? We're seeing like many multiples up.
Joy: Yeah. And Google did confirm that was a bug. So, I don't think they fixed it yet, though. Like if they are going to be, like retracting it later on, it's, I haven't seen it yet. So all of our reports are still showing these insane lifts in driving direction requests which are inaccurate. So it's good to know when the data you're looking at is wrong, and that would be one example of it for sure, the last example I'll quickly share here cause I know I want to leave some time for questions. this is something we noticed that I talked to Dale quite a bit about. And it's when Google shifts, the local pack results from what's on the left here to what's on the right. this makes a big difference in how your metrics show up. So in this case, we noticed last year for dentists that Google switched from showing a traditional three-pack to, I don't even know what this thing's called. I think Mike Blenthal had a name for it. But it's like where there's images, but you'll notice there's no website icon, no directions icon, and then it looks even more different on mobile. Instead of having a call icon, it has nothing. Like you, you basically have to click on the image before you even get the call button. and this change happened for most dentists last year around March, I believe. End of March. And so when we were looking for a few clients of ours at non-branded traffic, you see like essentially clicks to the website kind of fall off a cliff. like they go down significantly. and just to clarify, we can isolate this because we are using UTM tracking. and this is non-branded searches, specifically. So it's resulting in less people going to the website, more people staying on Google. and what was interesting is when Dale ran it across 4,000 dentists, cause he can do this at scale, given all the profiles that are connected. you definitely see the hit to website clicks when this happens, and then you also see the hit to calls. The big mystery question is, like, "Where did that user behavior go?" Like, I actually don't have an answer to this, we didn't see a drop in leads. I think people are still gonna need a dentist, they're still gonna need to contact the dentist, but this was really fascinating to see both of those metrics dip, when this happened. So it's something where you really want to keep track of, like, what's showing up in the search results and understand how that's going to affect your reporting, this is not something we published yet.
Joy: Do either of you have anything to add on that?
Noah: it was fascinating when you, when we did the analysis. I thought it was really interesting. Yeah. So, are, are we having a webinar next month? Is there something we should be mentioning?
Joy: You're so funny, I was getting to that, yeah, I do want to call it our, our webinar next month. I'm gonna be interviewing Mordy from Wix and our own Carrie Hill at Sterling Sky. We're gonna be talking about, websites, especially since Google's getting rid of their free website option. it's, you know, most businesses we work with, they have a full-blown website, but there's a lot of businesses out there that don't. They're relying on Google's, free website. So Wix does have some free alternatives that are great. And we're going to be talking about, like, what businesses should do, also just in general questions about onsite practices, content, things like that. So make sure you sign up for next month's webinar, or subscribe to our YouTube channel as well. You'll get, an instant alert when that starts, if you're subscribed. So make sure you do that. And now we're going to do, like, some rapid-fire questions. I hope you guys are ready, cause we've got a lot of them and, I'm gonna start with this one. Okay. "We've been using Local Falcon and can get most of the information, but there's a lot of AI recommendations you can get on there. Have you tested out the recommendations and seen any impacts?" , Noah, I'm gonna throw that one to you.
Noah: No.
Joy: Okay! Dale, you got this one. Rapid fire, right? You wanted to,
Noah: I mean, I've, I've seen them.
Dale: I want to collaborate with the Local Falcon team to get those delivered via API, like that's a personal goal. But I've not tested them.
Joy: Okay. next question we have, "How do you distinguish data from Google Search Console and Google Business performance properly? Like, how do you, what do you use those two for properly?" They're very different.
Noah: Ooh, Dale, do you want me to take it and you flush it out? Okay. So, Search Console to me is like learning all about your search landscape and getting really great insights at the connection of page and query data so that you can make all kinds of onsite strategic decisions. Google Business Profile data is the only way that you can see stuff at that granular city level, where you can really see how you're performing locally, they each have their, they each have challenges. Search Console data is only available at the country level. Business Profile data doesn't really let you get out of that local to then get to the national or international level. You wanna flush it out? Flush it out?
Dale: I think UTM is really just the answer there. cause you clear as god.
Joy: Yeah. There we go. Good answer! next question, actually, is one I saw when I read it earlier today and I thought it was really good. So this person said they've noticed a discrepancy in Google Business performance data from one week to another, despite the fact that they're exporting it for the exact same time period. So they said, "For example, on February 18th, they looked at data from the 11th to the 16th, and they found a total of 3565 clicks. And then they went back today for the same date range and found 4190 clicks." I'm assing this is the delay, right? Like, is that, that's what I would asse is the answer, is that correct?
Noah: That sounds like 1/7th of the date of the data, roughly. 500 clicks, yeah, it is right about 1/7th. So they were probably missing a couple days of data.
Joy: Yeah. So, we actually made a change here at Sterling Sky recently to, to make sure we weren't sending clients data before it was all there. So if you send out reports, for example, on the first of the month and you're using Google Business performance data, you're actually missing like, potentially up to a week, right? So it's like you're telling them, "Hey, you got 500 clicks in January" when in reality they got 600. so you're kind of cutting yourself short. So that's something to really keep in mind if you're using this for reporting. You won't have the full month until about a week-ish after the month is over.
Dale: That's the same for the keyword data. So any of the keyword data that we discussed today, Google provides normally around seven days after the month is ended for that previous month. So you do have to wait for that.
Joy: Okay, I think this is a question specifically for Dale, can you explain the color coding on the tool? I, I asse they're talking about the tool that you were demoing.
Dale: Yep, random. Completely random.
Joy: Okay. All right! and then they want to know, and somebody else wants to know if the, the data is U.S.-specific or international?
Dale: for the graph one, it is U.S. English. for the other tool you can pick the language and location.
Joy: Okay. somebody else wanted to know, "Are we analyzing, are you analyzing the connection between the quantities of services and products and the performance metrics?"
Dale: Yeah, that, that's been a pet project of mine for quite a while in changing those. I think the, across the board, no one has seen any penalizations from adding more. So it really is just a matter of picking the right ones and making sure it's relevant, that you're not ranking, you know, not attempting to add things that aren't relevant to you. But in general you will see an uplift. It obviously is going to decay as you add, you know, five or 10 more.
Joy: All right. And another one for Dale. is this what clients of Jepto can see, like or do they need a template in Looker?
Dale: so the way that we handle the reporting side for, for clients is that we grab the data from Google, push it into BigQuery, and then you use the BigQuery connector, just like you would any other connector, to create any Looker Studio report that you want, we've got templates and stuff. Hit us up later.
Joy: All right. I will put this one to, Noah. does, does everyone use call tracking numbers? Cause I agree tapping to call misses the mark.
Noah: Yeah. Yeah, it's something that we almost demand at Sterling Sky. And we think it's incredibly smart, and we think it's incredibly smart to not only use it but to be granular with how you use it. Meaning, on your GBP number that someone could click to call, we'd like to see that be its own phone number. And then, information, you know, phone numbers associated with organic and Ads, and also GBP onsite are all separated so that you can have those levels of insights as well. Did you see? Does that, do you agree with that, right? Am I, am I, am I, that's how,
Joy: I, I agree that call tracking is really necessary. What I did find, I did a study years ago comparing call tracking data with the Google business data, which is only clicks to call, right? And the trend lines were the same. So it is good to know that, generally speaking, if you see an increase in one, you would see an increase in the other. Like you can kind of trust those increases and decreases. But the numbers are way different. so I, I,
Noah: Yeah, they're horribly different.
Joy: For sure, there's a few more really good questions here that I will hopefully get to in the next two minutes. Just a really quick reminder, guys, if you find this stuff helpful, these webinars are, are completely free. But all we want is for you to subscribe to our channel. So, it really helps the YouTube, the YouTube algorithm, see our videos more and, we just want to ask everyone to do that, I'm gonna put up this one cause it's very different and it's interesting.
Noah: Oh, I want that one! I want that one!
Joy: Okay. All right, you got it.
Noah: Okay, so, so, yeah, basically, this gets to the question, the core part of this question is third-party data, which is available in any SEO tool like SEMrush or Ahrefs, versus the data that you get in Search Console. And the deal is you want to always work with what's called first-party data. And when we're getting data from Google Business Profile, or we're getting it from Search Console, it's straight from Google. It's not some third-party tools modeled data, IE, data that they think is accurate based on some kind of, like, advanced algorithm that they've produced in their own tool. You want to use stuff that's coming straight from Google. And that's Search Console data, or Google Business Profile data, or Google Analytics. And if I'm making strategic decisions, I'm living in Search Console data all day long, and that's just how I, I do it. But like Joy, Joy and I approach this a little bit differently. Like, like often times, I start from the query, and I do strategy stuff from the query, and Joy, you work from the page, right? So you're living in Analytics to start, right?
Joy: Sometimes. I don't know. I'm in all of it. So I, I literally use all, all, all three: Analytics, Search Console, and Google Business performance data, cause they all track different things. And I think just understanding that you kind of have to look at all of them.
Noah: Yeah.
Joy: all right, I want to be respectful of people's time cause I know we're at time here. So, we did not get to all the questions. If we did not get to your question, I would encourage you guys to post it on the Local Search For. Both of these guys are active users on there. I can tag them, flag them down to answer any questions we did not get to, so feel free to post them on our for over at the Local Search For, and we'll make sure to answer them there, and again hope to see all of you on our webinar next month with Wix. Thanks, everyone.
Noah: Yay! This was fun!
Thanks a lot for your feedback! If you’d like a member of our support team to respond to you, please send a message here
Oops! Something went wrong while submitting the form.
