Structured Data Problem Solving - Q&A

Geoff Kennedy
Published 18 December 2020

Structured Data Problem Solving - Q&A

Welcome to the third in the Sitebulb series of live, structured data Q&A sessions where our guest experts answer your questions on structured data. 

Our previous two sessions have covered:

In this third livestream we did a deep dive into what is possibly the most important area of structured data for SEOs: 'Problem Solving'.

Special Guests

We really have been spoilt in this series with the quality of our guests - and this time was certainly no exception. We were very excited to welcome two renowned experts and friends of ours: Aleyda Solis and Dave Ojeda

Alongside our guests, as usual were myself (Geoff Kennedy), and Patrick Hathaway, co-founder of Sitebulb. And again, Simon Cox was very kindly helping us out in the live chat, keeping the conversation going and keeping the more rowdy punters in check!

Watch the Full Livestream

Watch the full Q&A livestream recording here:

Please subscribe to the Sitebulb YouTube channel so we can keep you updated with future livestreams and webinars.

Jump Straight to the Questions

If you want to dive straight to one of the questions, use the links below which will take you to the point in the video where we covered that question, along with the full transcript:

  1. Is there a danger of schema duplication from different apps?
  2. Which structured data type should I use for which page?
  3. What is be the best way to get Schema implemented from a tech SEO/Google-ability point of view?
  4. Is it better to not use structured data code snippet than to use the wrong type on the wrong page?
  5. My Schema markup is correct and valid, but how do I ensure that the actual data is correct?
  6. How do you deal with changing or time sensitive structured data?
  7. Which structured data validation tool should I listen to?
  8. Why does Google recommend JSON-LD over Microdata?
  9. Has anyone seen structured data cause a manual action?
  10. How much should I worry about structured data warnings in Google Search Console?
  11. Have you found any content that you cannot use Schema for?

Resources

Here's some resources that were either mentioned on the livestream, or support what we talked about:

Full Livestream Transcript

Geoff:
Hello and welcome, everyone. My name is Geoff Kennedy and this evening, I'm hosting the Sitebulb live Q&A on structured data problem solving. Tonight on the live stream, we've got two of the most well-known structured data experts in SEO. We've got Aleyda Solis and Dave Ojeda. Welcome.

Dave:
Good to be here.

Aleyda:
Thank you for having me, although I wouldn't call myself a structured data expert at all, but I'm more than happy to be here sharing my, let's say, journey and misfortunes with structured data as an SEO, for sure.

Geoff:
Definitely, in the world of SEO, I'd say you're on the expert side of it. And from Sitebulb team, I've also got Patrick Hathaway here, who many of you will know.

Patrick:
Hello.

Geoff:
And on the live chat, we've got Simon Cox helping us out, so hello, Simon. This is the third in our series of structured data Q&A sessions. The past two, if you've not seen them, we did one on making a business case for structured data and implementation in the real world. Tonight, we're going to be looking at structured data problem solving, trying to tie up some of the issues from them last two that we've done.

Geoff:
Before diving into the questions, I'm going to ask Aleyda and Dave to give us a bit of an introduction. So, for anyone that doesn't know already, who are you, what do you do, and what is structured data to you? Aleyda, would you like to go first?

Aleyda:
Yes. I am an SEO consultant. I have an SEO consultancy called Orianti and I mainly tend to work with, whether SaaS, big brands, marketplaces, ecommerce types of websites, too. And the structured data, in my case, it has become, I will say, much more prominent in the last few years, even if we know it shouldn't be the driver of it, but it has. Google features support that are triggered by structured data.

Aleyda:
And, of course, we can talk about this later on in the Q&A but it realistically, yes, this has been the main driver in my case and the role that plays for me is how it can connect and align with the SEO process and efforts of my clients the strategy that has been set in order to maximize the visibility and ultimately the ROI and type of business-related goals that we are looking to achieve.

Geoff:
Yeah. A lot to cover, there. And Dave, how about yourself?

Dave:
I'm Dave Ojeda. I've been an SEO consultant for over a decade. My speciality now is doing structured data, whether it's for companies or white labeling for agencies. For me, structured data, I mean, I've been doing it a while and it's great to see that it's coming to the forefront like this time we're having here to chat about it.

Dave:
I think that, for me, it's two sides of the coin. It's the rich result potential that you have for your site and your site pages, but also that overall encompassing umbrella of understanding to build a quasi-knowledge graph on your site, relate the entities to each other, relate the entities to other pages and the internet as a whole and I think that's some really exciting times for us to be able to have the tools to provide it information to build these knowledge graphs.

Geoff:
Yeah. I think it's safe to say that most of us are just scratching the surface with a lot of this stuff. Once you get into knowledge graphs and things like you've mentioned, it's a whole different world.

Dave:
Yeah. Exciting.

Geoff:
Yeah. And Patrick, I'm sure most people know you now but give yourself a little bit of an introduction.

Patrick:
Yeah. Hi, everyone. I'm Patrick Hathaway. So, I'm the co-founder of Sitebulb. Yeah. You probably know me from spamming emails and stuff like that. And part of my role at Sitebulb is to help with the product development, and in the last few months, structured data validation across websites has become a big part of that. So, that is one of the reasons why structured data is become a big part of my job.

Geoff:
Yeah. All right. So, we've got quite a few questions to work through and we're also going to do a bit of a shout-out. If anyone's got any questions in the live chat, please let us know. If you can mark them with a Q at the start, that will help us pull them out and we'll try and drop them in as we go through and cover as many as we can. So, to get started, quite practical hands on one, this is about apps.

Is there a danger of schema duplication from different apps?

Geoff:
So, is there a danger of schema duplication from different apps? They give the example of Yoast and WooCommerce, but I'd assume this counts for any apps, whether that be WordPress or whatever platform you're on. Aleyda, you're nodding a lot there.

Aleyda:
Oh, yes. Oh, yes. Well, this, I have to say that I reckon that doesn't happen that much for my clients because they are more enterprise level type of companies and thankfully, whether their CMSs already have some sort of feature or we use an enterprise level type of website for example. I collaborated for an SEO process where they were using WordLift to create a whole knowledge graph of the websites. It was a very consistent, all the entities connecting to each other in a way that set up the right context regarding the meaning of the different objects, relationships, et cetera.

Aleyda:
On the other hand, however, I do have my own projects and these are the type of things that I think that is mostly smaller type of businesses might even struggle with that rely on WordPress, for example. That is a really good example, because yes, a lot of businesses use Yoast, but there are still a little bit of sometimes even restrictions of what they can, they cannot do or they are not only using Yoast but additional plug-ins that will add on additional structured data and some people will ... Well, there is room to specify even more. So, you end up duplicating sometimes the exact same different or assigned different type of types for the same content, things like that or duplicate the exact same property, things like that.

Aleyda:
So, I think it is important to have consistency to try to use a tool that will allow you to scale. So, for example, in my case with Remoters, We rely on WordPress. I use Yoast, but when we started implementing structured data for our job board, for the jobs structured data, it didn't supported it. So, we did run native sort of implementation. We ended up messing things around a little bit and now we are like, "Okay. Let's clean things up and rely only on this consistent tool to have a consistent implementation," a good implementation without issues.

Aleyda:
At the end of the day, this is the thing. Even if doesn't trigger any issues or might not trigger any issues or errors or warnings and Structured Data Validator from Google or in the Google Search Console at the end of the day is like, "Is this really an accurate representation of my content or of all the objects of my website or my entities?" And what we want to achieve there is to clarify as much to be able to give effective and consistent information to Google rather than mixed signals at the end of the day.

Dave:
And I couldn't agree more. I think that the thing is for small- to medium-sized businesses, schema duplication, it happens, and it happens without them knowing because they're just not aware of what's happening on their site.

Dave:
So, you can get a schema being created by a plug-in, by a theme. You could have it being implemented by GTM that you didn't know about, programmatic solution.

Dave:
So, the thing is that, to understand where is it coming from, and then from that point, understand, number one, should it all be there? In a perfect world, you'd like to have one source of truth or one creation scheme or markup, which sometimes you can't, but you want to look at that potential. But also, as we know, Schema mark-up can be microdata, JSON-LD, so then you have to look at another level of is it all coming in the right format?

Dave:
Now, Google can recognize mixed formats of microdata and JSON-LD, but in the end, I think it goes back to what that being consistent and knowing where it's coming from and having a consistent approach moving forward, regardless of the size of the site is the best approach to take.

Geoff:
Yeah. Yeah. That makes sense.

Aleyda:
I hear you with the GTM. No knowing where is this coming and searching through all of the features of your CMSs and plug-ins and themes and extension and then realizing, "Oh, the GTM implementation." With that static outdated structured data from, yeah, your CO. Yeah.

Which structured data type should I use for which page?

Geoff:
And next question I've got is quite a wide one. I don't think we can answer it in its entirety, but, "Which structured data type should I use for which page? I know it seems a simple question but I see the wrong type being used on the wrong page very time." So, without the context of what page we're talking about here, you can't really answer that, but have you got a process that you go through or any way of defining what should or shouldn't go on a particular page?

Dave:
I'll take it. So, the biggest thing for me is going and looking at the page and understanding what is the main entity? What is the main concept that is trying to be brought across when people are looking at that? And then, from there, looking at what are the other entities that exist on that page and how they're related to the main topic as well as the other pages and, like I said before, internet as a whole.

Dave:
So, now, certain pages are easy. A private page, a home page, which is an organization, a local business. Those are easy. Some of them are more nuanced. For instance, if you had an industry page for your SaaS product. That doesn't really fall to a perfect hole for Schema.org to represent, but you can use a catch-all in some of these cases, where it's like, "Well, this is information we want to get out there. It's relevant. Let's use an article because this is just information." And by using that article to represent something like that, you're able to represent that entity, fully define it, and even use external definitions such as Wikipedia to represent that industry related to the SaaS product and I think that's where you'd need to go.

Dave:
The other thing I'll say is that sometimes, with structured data, you should mark up only the information that is on the page but not necessarily mark up everything that's structured data. You have a choice there and there are some pages from certain cases where there just is not a need to put structured data on there. What is represented there is not important enough to build into this internal knowledge graph you're trying to build.

Aleyda:
Okay. Well, that is super important, because at the end of the day, it should be something meaningful for the context. For that particular context of the graph that you're building. It's definitely more straightforward in some scenarios than all this. So, for example, if you're in a product page and you have an FAQ at the end of the page or a Q&A. Realistically, if you ask any SEO, any developer for that matter, person in charge of the implementation and you ask, "Okay. If you want this page to be shown for something or to be recognized for something, for what it should be?" I think in 99% of the cases, the person will answer, "Oh, for a product. This is a product. The main value of this page should be a product." So, this is something that you will definitely prioritize.

Aleyda:
In other cases, it got a little bit trickier. So, for example, especially with informational content, Q&As, how-tos, FAQs, some FAQs can look a little bit like how-tos. And some Q&As can look a little bit like FAQs, to have a lot of hybrid sort of roles sometimes. But, at the end of the day is about verifying and validating and asking if I want to be shown for something or to be recognized this page for something, what is the most representative format or type for this particular page and be consistent. What is important is also being consistent, instead of trying to mix, for example, if you have 10 type of pages of this in type to be blending and switching from one type to another in every single one of them. Consistency is also important.

Geoff:
Yeah. I think that then key thing that you come back to is what is the main entity? What is the key focus of the page? That seems to keep coming up again and again.

What is the best way to get Schema implemented from a tech SEO/Google-ability point of view?

Geoff:
So, got one in the chat here. This is one we've actually covered in some of the other sessions, but it keeps coming up time and time again. We've already talked about apps.

Geoff:
So, Maret says, "What would be the best way to get Schema implemented from a tech SEO/Google-ability point of view?" Now, this is usually asked from the easiest way of something, but this slightly different spin on this.

Geoff:
So, and you mentioned before there the likes of Google Tag Manager. It's really interesting to get different people's takes on the likes of that, especially.

Aleyda:
Well, I think that I am very happy to see that there are more and more tools who work in better structured data. So, even before, we mentioned Yoast. Realistically, I have been surprised of how they have started to support so many more types of structured data lately and then we have tools like SchemaApp that are also supporting I think that most of the types that you will need and they have a WordPress add-on, too, so not only enterprise-level websites are able to profit from it or to leverage it. And then we have tools or add-ons like the ones of as WordLift.

Aleyda:
At the end of the day, I think that what will be important is to check depending on your own capacities and flexibility and resources is that what is the easier in your particular case if you're using a platform or a CMS where there is not native feature, what is the plug-in or add-on that you can use that it scales and supports and give you support for the structured data that you want to implement right now and that you might think that might need in the future, and may want in the future, too, to avoid having to switch or replace things like that.

Aleyda:
That is important but, yes, you might even leverage, you want to leverage GTM although it breaks my heart to see so much thing relying on GTM rendering and the impact that it might have also on so many other things, but yes, I will try as much as possible to implement them directly within the CMS or the platform directly itself. Yeah.

Dave:
I would go with understanding once again, what is the sophistication level of your team or the people that you're bringing in to implement Schema markup? I mean, if it's an enterprise-like solution, then you are looking at enterprise products such as SchemaApp or WordLift, something like that that you need to do.

Dave:
If the sophistication is lower, then maybe it's something, maybe they're simpler, more straightforward, so maybe you're just using a Yoast plug-in or something like that to get what you need because that's everything you need. And a lot of times, sites are not going to be scaling. They're a 50-page site, a 200-page site and nothing is going to change in the next two or three years so going with a simple plug-in solution is good.

Dave:
Now, with GTM, it all depends with the sophistication you have with that, but also I tend to look for other options other than GTM. GTM, I'm mostly confident in it, but not always is what I could say about that.

Patrick:
Confident in that it's being picked up?

Dave:
I'm confident in the consistency of it being able to work correctly and not breaking and they've made tons of improvements, but I've also seen some quirkiness happen over time. It's one of those things where it's like anything, you want to keep looking at everything you're producing, everything you're deploying to make sure it's still working with GTM. I'm more confident in other tools and other resources to implement schema markup versus GTM on a lot of occasions.

Is it better to not use structured data code snippet than to use the wrong type on the wrong page?

Geoff:
Mm-hmm (affirmative). Yeah. So, I've got another one here from Mark. Kind of a follow-up to his question about which type and which page. So, if you're not sure about which type you should be using, is it better to use the wrong type on the wrong page or not at all and what is the impact of using the wrong type?

Dave:
Well, I think that obviously we don't want to use a wrong type on a page, no matter what and if there are some questions on what that type is, maybe you look deeper into why is it we don't quickly understand what the major class is? There's a question there that goes just into the content of the site of what page represents that I think that, over time, if you make that investigation, you may start to understand what that main entity is or maybe decide that it isn't really one thing, it's several things. So, what should we do with it?

Dave:
I think that going with the option of putting something that you're not sure of, I would rather hold off putting any schema markup until you have a definitive answer that makes sense for you of what's represented on your site.

Aleyda:
100%, also because you can, not only that, you can end up with even getting a manual action, just because of that. It's something that you really want to avoid because then, the effort that you will do will go to the garbage anyway. So, I think that it is important that you confirm, validate well that it's actually the relevant type of structured data that is meaningful for that particular page. Yeah.

My Schema markup is correct and valid, but how do I ensure that the actual data is correct?

Geoff:
Yeah. So, jumping forward a bit. So, post-implementation stuff. "My schema markup is correct and valid, but how do I ensure that the actual data is correct?"

Geoff:
So, we were talking about this the other day when, after we got the question and how we've interpreted this is that it's the data type and everything in there is fine but, say, your template is pulling in a title into the description field or something. Is there a way of telling when it's valid; as far as Google's concerned, it looks correct. But how would you make sure that you're not populating it with rubbish? Is there a way of doing that?

Aleyda:
I think, at the end, this sort of type of the validation that you do in your SEO process with your own content, when you verify that your metadata is correct, that your titles are descriptive and meaningful and representative of the content. The same with the meta descriptions. The same with your headings. Then, you take a look at your structured data, of your schema implementation and see if it actually is updated or representative of the product, for example, or the information of the page where it is implemented.

Aleyda:
So, an active and proactive validation I think is meaningful and is important, especially when you are just releasing to avoid including information that is inaccurate or that is not relevant at all, for example, not representative of what is actually meaningful, et cetera.

Aleyda:
Yeah. I think that an extra step of validation is important to align with the editorial guidelines, not only the technical guidelines that Google has and I find that their documentation is really good. Actually, the examples, dos and don'ts. So, not only from a technical perspective but you have a checklist of validations that you do follow in their dos and don'ts and the specification for the guidelines for that specific type of structured data.

Dave:
I once again agree a hundred percent. One thing I would add, too, is then when you're implementing, let's say, like for enterprise level, that you're implementing it and you're just putting out a subset of what you did for the pages. You're putting 10% of the pages out there and then you're going and, once again, validating, reverifying everything on that scale, just to make sure that for the most part, as you're pushing this out more as a template level, that you understood that the template is sound, it makes sense and it's pulling in what it needs to pull in to properly represent what is on the page. So that would be my only addition.

Patrick:
Yeah. Templates is the key, though, isn't it? Check all your templates and eyeball them. Take the time to actually look at the data.

Geoff:
Okay. I think it doesn't matter how many tools you've got, it needs a level of manual review at some stage.

How do you deal with changing or time sensitive structured data?

Geoff:
So, another quite specific one. This one actually came up with Nick Ranger in our last session. She was saying one of the teams she worked with was having problems around time sensitive structured data. I think the example she gave was for products. So, I can't remember whether it was changing prices or it was stock levels. Yeah. Have you got an approach to dealing with that?

Aleyda:
Well, that is why it is important that you have a solution that actually populates and updates with the actual information that is shown all the time on the website rather than static option. Maybe for static page like an FAQ or a Q&A or something that never changes. It might be okay to hard code the structured data, but, if it is content that is dynamic and is changing all the time, very sensible to be changing the status. Then is important that the code populates from the actual information that is shown to the user and if the information that has changed, if the content, if the price of the product has been updated, if there is an update in the ratings of the reviews, if there is an update in the stock, that is reflected and updated also in the code and how it is populated and updated also accordingly, not only when you update that information with the information that is updated on the page, doesn't only update on the text, on the information, the content, but also on the structured data, the Schema that is generated for that page also.

Dave:
I think this goes back to - How are you implementing this? Are you implementing this in a dynamic way? Are you implementing as a way that will refresh and look at those pages constantly to make those changes? In our big ecommerce sites, you want to have that.

Dave:
Now, for smaller sites, you may not have the time or money to have a product or service doing that, so you may have to go back to literally you're the one or your team is the one that is manually making those changes and implementing them, keeping track of them, which is not the best way to go about it, but I realize that a lot of companies may not have the ability to invest in infrastructure or product to be able to make it seamless.

Patrick:
I think in this case, or actually Nick's example was one where they had a special price offer that expired on midnight of the 24th, for instance, but Google still had that version, the price displaying of the offer on the 25th, 26th, 27th, of people coming to the site and going, "Well, the offer that I saw in the SERPs was £15 but actually it's £20. What's the deal?" Is there any specific ways that, any specific properties that could be used, for instance, to mitigate that sort of thing?

Dave:
I think it's more of an issue of is your page getting refreshed? And at this point, I think we're totally out of control of even doing it on a singular basis now in terms of requesting indexing. So, I think that you're at the whim of Google.

Dave:
One other side note on that example. I think the other example is I think, Nik, you were saying about whether it's in stock or out of stock and why do you put something out of stock and then people won't go to it? And another little workaround on that is you pull, once again, you don't have to markup everything on a page. You pull that availability markup off of the page if it's out of stock, so that people don't see that. They will potentially go into that page and see that. And then they see it's out of stock, but now, if they're brand-centric, then you have a carousel on the bottom of what things to choose but also, you have an email, like, "Hey, it's out of stock. Put your email in here and we will alert the minute it is ready to go." And so that's another way to get them into the site and you to work with them to keep them as a potential buyer.

Aleyda:
This is great because it shows how this implementation shouldn't be insulated. It should align with a broader set of efforts within your technical list. SEO strategy for your content because, at the end of the day, this is highly tied with meaning and content of, even if the implementation is highly technical and this shows the impact of it, when to select, to show what, when to do a validation and an analysis of what is actually meaningful for you and you don't only do it because there's a checklist saying these are all of the attributes that should be shown. No. It is really required now. If it is not required, is it really something that I can profit on and it will be like something that will end up harming me or not benefiting me. So, you should assess that for sure, yeah.

Geoff:
Yeah. I mean, I've seen a few people that the solution of what they needed was to remove the structured data. They found that that was a decision, like you were saying there, sometimes it's not beneficial to have it on the page.

Aleyda:
But it's like, it's, again, you need to do an assessment and a validation from a strategic standpoint. In the past, I have seen and we can see a little bit of data how some people were relying on dynamic serving or independent mobile worksites and that of removing their structured data from the mobile pages to avoid adding more higher loading times of these pages entirely, rather than finding the more balanced type of solution and option. These pages now will have issues with the mobile first index.

Aleyda:
So, again, you can implement or select to do something in so many different ways. What is important is you analyze and validate and select the roadmap that actually makes sense to profit as much as possible and without necessarily implementing streaming solutions that will end up harming you even in the long run, right?

Which structured data validation tool should I listen to?

Geoff:
Yeah. So, next up is a question that I really like because we've had to look at it so much in the last few months. "Which validation tool should I listen to? They all give me different results!"

Patrick:
Sitebulb.

Geoff:
So, I mean even Google don't [crosstalk 00:28:05]-

Dave:
Sitebulb!

Geoff:
... the obvious, I'd say, yes. We all know that one, but even Google's own tools. I know they give different results.

Aleyda:
Yeah. Well, it's funny to see because, at the end of the day, the tool that Google is keeping and pushing and promoting is the one that is connected with the search features that they support. They actually informed yesterday that they will be maintaining but migrating the Structured Data Validator but in the Schema website, isn't it? Not in their own website, showing that, at the end of the day, what they will focus on is on those structured data that are tied with search features or that they will support at the end of the day.

Aleyda:
Realistically what I, as an SEO and I have to say, what I focus on and try to prioritize is the validation, the data that Google gives me whether through the Google Search Console.

Aleyda:
Then, of course, I use their own sort of features validator tool directly. I use SEO crawlers, too, to triple validate or double validate. Of course, Sitebulb is my preferred validator of choice in this case because how well is tied also and how it differentiates those structured data that will generate your features and which are not and how it connects also so well with what triggers issues and warnings with Google directly.

Aleyda:
So, yes I double and triple validate. I try as much as possible to avoid any warnings, any issues across all of them and if I will prioritize something, I will say what triggers the errors in the Google Search Console or in the Google tool at the end of the day, because, sadly, as an SEO, yeah, I am dependent of the big G here for my results, yeah.

Dave:
Absolutely, absolutely. In Google Search Console, that's kind of where you go for the absolute answer to truth, like, "Does this work for us, yes or no?" And then all the other pre-steps of doing, of going, of Rich Results Test, checking that out. Then, going to, well, a structured data testing tool I still use because I love it, but it'll soon be, I think it's validator.schema.org, which will be and going to check everything related to the schema marker that could potentially be on the page. I think that's important. And now, once again, that's only page-wise. And then we can go to something like a SchemaApp or a Sitebulb to look at site-wide, which, of course, gives us a whole bunch of other information of just how we're creating this as a whole for this knowledge graph and see if there's missing things.

Dave:
But I think that, once again, you're correct. You have to use all of these tools in concert but, at the end, it's going to be Google Search Console.

Dave:
But one thing I will say is, with all these tools, they are validating is the syntax correct, is the connection correct? But the reality is, they're not telling you did you properly take the main entity? Did you take that main entity and correctly nest a property type against it that makes sense? For instance, you have a product where you nest via the property type subject of for an FAQ page. Did you properly do that right? Did you do it when you should have versus did it and you shouldn't have?

Dave:
So, those are the bigger questions in terms of validation that is not just on the technical level. It's the level of the entity structure and connectivity and that's still ... There isn't a tool for that because that's where we come in, to make those decisions and help guide our clients for that, but there's just that other level as well.

Geoff:
Yeah. I mean, that was one of the things that we kind of hoped to try and do with Sitebulb, which is to facilitate that process, to put it into a format which you can easily tell whether that's right or not.

Dave:
I love the tool. It's great for what it does.

Geoff:
Get it.

Patrick:
So, information and Sitebulb is the answer to that question. Okay. Right.

Why does Google recommend JSON-LD over Microdata?

Geoff:
So, following on from that, from the chat, we've got one from Suganthan here. So, this is more about pandering to Google again. So, "Why does Google recommend JSON-LD over Microdata? Surely Microdata is superior as there's no chance of discrepancy between schema and text."

Aleyda:
Yes. This a really good question and again but in fact, then it really took my attention because I still remember. Oh, my god! I might tell my actual age here, when we were still using micro formats. And then, all of a sudden, Microdata came and it's like, "Oh, my god! This is great," but still HTML. And then JSON-LD is like ... So, I could see this push a much more reliance on the JavaScript side of things. It will be great. That is a question that you might want to ask to John potentially, too, to see why there is a preference in their case, because at the end of the day, yeah, with the schema vocabularies, you should be able to use different type of formats. I don't know if they have probably answered this in the past. I don't know they ... David, you know the answer here?

Dave:
I think that the biggest thing, to my knowledge of what I heard, is that with JSON-LD, you're extracting it away from the HTML. And so it becomes separate. And then, so you don't have to deal with what's going on in the HTML. And so, for a lot of people that don't have developers or in-house people, that not having to interact with HTML is a really good thing that you can bring in and slide this component in there that represents structured data and does not touch anything, because, as we know, if something goes wrong, they're going to be like, "Oh, you were in it. So, it had to be you," which is not always the case. But I think that that separation was the key that make that clean.

Dave:
Now, is it a duplication? Yes, it is, but is it easier in the long run for a lot of people not to touch the HTML? And I think that's the reason.

Aleyda:
It's true. It is. In an additional layer at that time because it's not within the HTML but, then again, indeed, you can manipulate it much more easily because it is not within the HTML directly. And yes, it's a duplication at the end of the day because the information is already in the HTML, too. So, yes, and then also, it's not that a coding in JSON-LD is necessarily less complex, that some few tags and added there. But, yes, it might be a cleaner implementation for sure and you don't need to be manipulating the HTML data entirely all the time, yeah. Might be right.

Geoff:
I know from personal experience when they first brought it in, I was working with a client and it took it from there was no way in hell they were going to be able to do it. I mean, it was a massive site to, all of a sudden, well this sits separately. So, yeah, we can do it. That's not a problem. It just needed to pull the data it needed.

Geoff:
So, that was really interesting how it was a complete no-go when they need to edit the HTML. Yeah.

Aleyda:
You know what? It's interesting, but call me a purist here, but if you give me something to implement in general, of course, I will use JSON-LD because it's the preferred format. They have mentioned it. But, if in general, if you give me different choices of doing it and if you give me the simpler to understand and better integrated one. If you give me the HTML way to do it, I will rather choose it and try to avoid any necessarily reliance. I think that there's so much reliance nowadays on JavaScript and especially the client-side render one and yeah. No, it has been going a little bit out of hands, but yes.

Has anyone seen structured data cause a manual action?

Geoff:
So, another one from the chat from Simon here. We mentioned manual actions earlier on. "Has anyone seen structured data cause a manual action?" I haven't.

Aleyda:
You know what? This is very funny. Potentially they said you focus on structured data much more. You have probably seen all type of cases and weird or not even, but in my case, thankfully, I have never been penalized for something that meaningful and in the case that a website I had access to was penalized by a former implementation, was something that ended up, it was a manual action literally flagging some particular pages that were not necessarily featuring ratings and were including some rates supposed to be right information and you don't like, and honestly, the impact was not meaningful pretty much because these pages were trivial literally for the website as a whole. It's interesting to see the stance though of Google with this sort of things, because I think it was a couple of years ago or a year and a half when, for one of the many elections that Spain went through 2018 or so, one of the main newspapers, the top one literally ranking for the election results in Google. Very big media website was using the event's structured data to be shown with the really nice looking event sort of features there and they were not featuring any events.

Aleyda:
And I called this out literally in Twitters like, "Why is this page featuring this structured data," et cetera, et cetera, "And Google shows that and nobody's doing that."

Aleyda:
And then, all of my SEO colleagues here in Spain that I know who works for other prominent media website and newspapers say, "Yes. We have all called out that many years ago and Google doesn't do anything."

Aleyda:
So, it depends on the site, I will say, what they end up doing from what I have seen with that example of the action that they take. But yes thankfully I haven't seen any case myself of a website losing even rankings or something that much of an impact because of a misuse because I am a very highly responsible SEO [crosstalk 00:38:43].

Dave:
So, I have seen manual actions and more so in the past than now. I think what has happened with Google is obviously a lot of it had to do with rating stars, obviously, but that abuse, they've cracked down on rating star abuse tremendously. And I think what ends up happening now is instead of manual actions or just actions in general is that they're just not giving it to you. You're trying for the rich result or you're trying to get the local business that is a subpage of an organization to have rating stars and they don't give it to you. And, in the end of the day, they're accomplishing their goal which is stopping you from algorithmic as opposed to doing a manual sort of action and I think that their ultimate goal is probably to do no manual actions if possible and do everything algorithmically. And I think that that's what I've seen in structured data is less of manual actions. And even then, they were far and few between and when they did come to me with that, it was real abuse. It was like absolutely hiding content, spamming rating stars and so forth, but I think it's less and less and they just have the ability to just not give you what you were trying to hope for.

Geoff:
Mm-hmm (affirmative). Yeah.

Patrick:
Geoff, we have a really good question from Andy Simpson. Do you want to pull that one out?

How much should I worry about structured data warnings in Google Search Console?

Geoff:
Right. While I go and find that, I'll leave you with another quick one here. "How much should I worry about warnings in Google Search Console?" So, a bit of a step backwards from the manual actions.

Dave:
Once again, it depends. So, once again, a warning is a warning, so you don't have to worry about it. But a lot of times, a warning is related to what they think should be on the page.

Dave:
Now, there's many cases where they're saying, "You didn't put this in here. You didn't put the tin code," and it's because, well, we don't have it on the page. There's not a need for us to put this information on the page. Therefore, we're not going to do it. And the priority has to be you're putting what you need to represent on the page for that product, for that service. And so the warnings are their guideline for what you should put, but the priority is always going to be what makes sense for your product and your pages in terms of the warnings. Errors are a different subject but even with errors sometimes it's like you get an error for something, but we just don't have that. For, let's say, a product and the product like, "We just don't put price." We don't feel the need, we do have to call us to talk to our sales people because it's $50,000 dollars a piece. We're not putting that out there. And so even errors sometimes, once again, it depends on what your needs are first with the product pages and with all the pages.

Aleyda:
I hear you, especially, for example, in my case, I have seen a couple of situations where, for example, job postings missing base salary. Sometimes I am not given that information. Sorry. What do you want me to give, misinformation to say, "Zero," or something like that?

Aleyda:
I've seen websites doing this and to get away with this. Unfortunately, what I have found and also it largely depend on what type of structured data but, in our case, for the job posting, we are not being pull and shown in the features box I guess because we have a few warnings like this. But it's because we literally don't want give out misinformation.

Aleyda:
So, again you need to assess here what's the best way to move forward, what you are looking to achieve, what is your goal, where are the trade-offs. In our case, for example, what we have chosen to do is to start gathering more data and we're adding more fields to whatever job posting that we publish on the website. So, we can support those fields in the future and populate those fields in the future with accurate information so we don't end up having any sort of penalization or issues in the future and being able to be completely open and available to be selected, to be shown in the job features box.

Aleyda:
But, yes, at the end of the day, it's to assess this. Another scenario what I have found that it actually bothers me and, sorry, I am sharing here my struggles. In the how-tos, why do you need an image for each single step? What the heck! I hate that. It makes sense. In other places, it's like, "Why?"

Aleyda:
And then what I have seen that many websites do is they populate every step with the exact same image. It takes you off the actual supposed to be goal of to represent each step with something meaningful. So, again, it's like, yeah, take it with a grain of salt and balance it out and analyze it and assess it in the best possible way that represents the content, trying to align it with the guidelines. The editorial guidelines are provided by Google and also to maximize the chances that you are gotten into those sort of features, because, by all means, I love that my content is more meaningful for Google thanks to the structured data, but I want to be included there, too. Yeah.

Geoff:
Right. So, time is ticking away on us so we've got this really interesting one from Andy that I think we'll probably finish up with.

Have you found any content that you cannot use Schema for?

Geoff:
So, Andy asks, "Have you found any content that you cannot use Schema for?" He does victim support pages for sexual abuse victims in LA and cannot get FAQ schema to show for any of the main pages. And he's also followed up with a few more deals there. They all rank on page one for related terms but no FAQ schema.

Aleyda:
Yes, I know that, for example, there are certain type of content like porn-related information that is not supported, for example. I think that there are some types that, unfortunately, of information that are not supported.

Aleyda:
So, yes it's important to check the guidelines at the end of the day and see how much you are able to get in or not. And I'm afraid that, if it is not supported, you're not able to do that, I wouldn't recommend you to do it anyway because you can end up getting a penalization because of that. It's important.

Dave:
And I think that with the FAQ schema, before it used to be almost a guarantee. You would apply it, you implement it and boom! FAQ. And then they backed off on that and they were just a little more cautious on making sure that the FAQ schema was relevant to the content, relevant to the search result. And therein, I've seen lots of clients who have lost FAQs that they used to have before because trying to be as aligned as they can.

Dave:
Now, I don't know the specific example here on how it relates or what's happening in the SERPs, but I think that it's a case of them being more cautious in aligning the content of the FAQ page to probably the main entity of that page and how it relates in terms of the search intent.

Dave:
So, once again, it doesn't answer the question and if I knew more details, I could probably answer it but, right now, they're more stringent in terms of allowing that rich result to happen.

Geoff:
Mm-hmm (affirmative). I mean, I do wonder as well with something like that whether it's possibly just being miscategorized along with likes of what you were saying, Aleyda, with porn and offensive content and things. I mean, when it's done algorithmically, you never know how these things are getting grouped.

Geoff:
So, that brings us to the end of our time. So, thank you to everyone for watching us, and thank you very much, Simon on the chat there, for keeping things rolling. Looks like we've had quite a few people in there. It's been good. Especially thanks to Dave and Aleyda. It's been great hearing your approach on this stuff. Could ask you a lot more questions, if we had the time.

Geoff:
To finish up, could I ask both of you to give one last tip for structured data problem solving? And this probably also a good time to let people know how to keep up to date with what you're doing and if you've got anything you're working on at the minute that you'd like to share with everyone. So, Dave, would you like to go first?

Dave:
Sure, sure. So, my tip would be just that, while you put a lot of effort into getting your structured data implemented out there and recognized, it also is important with anything regarding the website to occasionally check and verify. Is the schema markup still coming up? Is it still being deployed correctly because, depending on the site and the size of the company you're working with, there are other hands in the pot. They may be doing something on GTM or they're tweaking something with your microdata. And so, you want to periodically put that in with all the other things that you periodically test to make sure that it is working correctly. A plug-in updates and, for some reason, it's not working now, but you didn't know.

Dave:
So, just keep that in mind to just put that in your occasional checklist for the website. You can get a hold of me if you want. Let's see if I can do this right. Let's see. There on Twitter or you can grab me on LinkedIn. Happy to talk to anyone about the schema markup, and yeah. Thanks for having me.

Geoff:
Thank you very much, Dave. Aleyda.

Aleyda:
Yes. In my cases, make sure to align very well your structured data efforts within the broader SEO efforts that you do in your process. So, it pays off and makes sense and you can show results as soon as possible.

Aleyda:
So, for example, a way to do this is if I know that it will take non-trivial technical effort to implement structured data and the client will want to see results faster or they ask to make a case for them.

Aleyda:
So, I go and take a look at which are those are features that are shown in Semrush or Ahrefs or SISTRIX or whatever of these tools that are being shown in the SERPs where my pages already been shown in the top 10 results. And I see that all the players, all the competitors ranking along my own website are also even already leveraging. And then, I prioritize those pages to be implemented. The first ones. These are the ones that will start to get featured much more quickly, very likely the ones that will have a much higher impact, much more faster from a click-through rate perspective to and the ones that I will be using not to say, "Look! This is what we have done and that's it," but, "Look. This is my case. This is how we have implemented structured data across these sections or this group of pages, set of pages, and these are the results that we have got and this is how now we are featured before and after versus our competitors."

Aleyda:
So, this is why we should move forward with even more and get even more buy-in, more support, even if the initial implementation was potentially not necessarily the ideal in the first place.

Aleyda:
So, for example, I have had cases in the past where the initial implementation was we had to do it via GTM, I know not best, but it allow us to prove a case. And three months after, because of the results that we were able to show through this initial GTM-based implementation, we got the approval to implement structured data in a proper way directly through the templates on the website.

Aleyda:
So, I will say to try to be strategical about it, to prioritize, to build cases, to show results as much as possible faster because, at the end of the day, yes, we're in SEO and we need to show results and positive ROI of our efforts, right?

Geoff:
Yeah. We're all slaves to the clients [crosstalk 00:50:44]. Well, thank you very much Aleyda and Dave. It's been great having you. Thank you very much for everyone for joining us. Please subscribe to the YouTube channel. We're going to have more of these live streams coming up hopefully. And, of course, make sure you have a look at Sitebulb. As we just said, that's the tool you should be using. You can get a free trial on there. Let us know what you think and that's it from us. So, good night and thank you again, everyone.

Patrick:
A merry Christmas.

Geoff:
Bye-bye.

Dave:
Merry Christmas. Yeah.

Geoff:
And merry Christmas. Yeah.

Aleyda:
Bye-bye. Merry Christmas. Take care.

Geoff Kennedy

Geoff does marketing 'stuff' and other things at Sitebulb. He's been one of those SEO types for a long time now, so should know what he's talking about, but that's debatable, feel free to debate it some more.

Free 14 day trial.
Full, unrestricted access.
No credit card required.

Try Sitebulb for Free