528 episodes | 520K+ downloads

Supporting image for Getting Technical SEO Right: Structured Data Markup – Geoff Atkinson
Getting Technical SEO Right: Structured Data Markup – Geoff Atkinson
The Agents of Change

Getting Technical SEO Right: Structured Data Markup – Geoff Atkinson

By now we’ve all learned that appeasing Google’s authoritative and ever changing algorithm will go a long way in helping us rank higher on the search engine. Structured data takes a technology-based approach to SEO, where it helps to better organize the data on your website so that it’s more easily processed and analyzed by Google, and they can have a better idea of what your website is all about.

Pioneers in this area, such as Geoff Atkinson from Huckabuy, guarantee that taking this approach will see your website ranking higher, you’ll see more organic search traffic, and search engines will better index your pages for keywords more relevant to your business. Most importantly, this is a best practice that both large and small businesses should be adopting.

Rich: My guest today is the former Senior Vice President of Marketing at Overstock.com. During his tenure they grew SEO from a $0 to a $300 million channel in a few short years. It really became a core competency as they battled with e-commerce giants like Amazon. During this span they came to the realization that SEO is really a technical problem that revolves around improving communication with the most important visitor to a website, the Google search bot. 

He went on to found Huckabuy, which was initially based on an affiliate marketing play, it ultimately morphed into a SEO software company with two products; automated structured data markup, and the SEO Cloud, which were inspired by key Google initiatives to encourage a perfect world for their search bots. They are a fast growing startup now and current customers include Salesforce, SAP, and Concur. Welcome to the podcast, Geoff Atkinson. 

Geoff: Thanks Rich, great to be here. 

Rich: So let’s start with something in your bio that seems to be contrarian to what a lot of SEO experts say these days. That SEO is a technical problem that revolves around improving communication with the most important visitor to a website, the Google search bot. If I’m understanding you correctly, you’re saying write for Google and not for the end user? Care to explain?

Geoff: I’m talking less about writing for Google, I think you do end up writing for the end user. What I’m saying is that most sites have a technical gap that they’re slightly built for human beings, and there’s a lot of money spent on UI and UX, and what the human experience is. And really most sites don’t pay attention to what Google’s UI and UX is, what’s their experience when they come. And that gap is really a technical problem and that’s what Huckabuy is out to address. 

Rich: Ok, alright, so there’s these technical issues, and we often talk about there’s on page issues, there’s off page issues, and then there’s these technical issues. One of the big ones is this idea of structured data markup, which I believe a lot of people call “schema”. If they’re not the same thing, what’s the difference? And if they are the same thing, how do you explain structured data markup to the layperson?

Geoff: So they really are the same thing. The reason that some people call it “schema” was that the first open source movement around structured data was on a site called schema.org. And so they are the same thing. The actual repository that Google uses, however, is not the repository of schema.org, it is something called JSON LD, but that’s really just a name difference. 

Structured data is a language. So for years and years the way that Google understood websites was they would crawl HTML. They still do that today for 95% of the internet. But if you’ve ever seen HTML it’s quite complicated and confusing. So search engines and academics came together and said there’s got to be a better way for websites to communicate information into a search engine. 

And so they established this language called structured data that allows you to communicate what’s happening on any given page. So there’s structured data around if there’s a person on the page or a bio, there’s structured data for that. If there’s a recipe on the page, if there’s a product on the page, there’s structured data for all these things. So it just allows Google to understand – and all search engines – more easily what’s happening on any given page. And then they actually end up using this information within their search results. So as you search for a recipe and it just shows up, or a sports score, those enhancements to the search results are powered by this language, structured data,

Rich: Ok. So on some level it’s kind of like back in the old days when we put alt tags on everything – and still that’s best practices – on images because that’s trying to make it easier for the search engines to understand it. One thing that structured markup data does is it makes it easier for Google and other search engines to understand the content that’s on our webpages.

Geoff: Yeah, exactly. So the image Meta data – or just Meta data in general – is very similar to structured data. But the difference is that Meta data is really suggestive, so it’s saying this image could be a dog or whatever. With structured data it’s authoritative, so you’re telling them officially this is the product and this is the price. So that’s really the only difference is, it’s authoritative. 

Rich: Not that we need to worry about any of the other search engines out there, but are other search engines like Bing also paying attention to structured markup data?

Geoff: They are, yup. This is a universal language that’s adopted by pretty much every search engine. But the big ones especially, Google’s on the forefront but Bing has made a very big effort to leverage structured data because it does enhance the search results. And so if they’re not using it, they’re at a disadvantage to Google. 

Rich: Now I believe there’s a number of different types of structured data, you mentioned one like the bio, can you give us some examples of ones that we might run into or might be using?

Geoff: So the most common structured data is product. Product, specs, descriptions, reviews. Product is by far the most used, but you really can do almost anything. Next it’ll probably be like real estate, there’s a lot of structured data around real estate and hotels and vacation rentals. There’s a ton of structured data around recipes. 

You can actually kind of notice where structured data is the most prevalent by how much the search results are enhanced. So as you search for a recipe and the search results are really enhanced and they’re showing the actual recipes or showing reviews, those are typically queries that are in industries that have a lot of structured data. 

But it’s almost anything. So humans, the medical world, the legal world, sports, it really covers almost anything. The cool thing is that they’re always updating it, too. So one of the big pushes just in the past week was the movie structured data got a lot more enhanced and complicated. And this is an effort for Google to basically be able to be a marketplace for movie tickets and for people to actually be able to buy movie tickets using voice search. So voice search is actually also powered by structured data, so it keeps changing and getting better and better. 

Rich: That’s very interesting. So is Google then the group pushing forward on these initiatives, or is it part of this non-partisan group or whatever who are saying here’s where we need to improve communication on the web?

Geoff: It is really pushed the most by Google. The other non-profit organizations help with the standardization of it, but unless Google adopts it, it really doesn’t matter. So it’s Google that’s really pushing this initiative. 

Rich: You mentioned that the movies one had just been enhanced. I think I read recently that there had been some new ones added. What are some of the new ones that are now making it to the marketplace?

Geoff: Oh man. I’d say it’s kind of constantly evolving. I think the best way to sort of monitor it is there’s actually a GitHub repository that Google uses. That’s what our team, our development team uses it. We use the same GitHub repository that Google does, and it’s the JSON LD repository. And JSON is basically a way to connect data, it’s a coding language that allows you to make these connections between pieces of data. And that’s really the best way to sort of monitor the enhancements and what’s being changed.  

Rich: So Geoff, you’re obviously in this a lot, is there a category that you don’t understand why they haven’t done this yet? Is there one on your wish list that you wish they’d come out with that you think is relevant? 

Geoff: I don’t personally have something that I’m rooting for, but I know our tech team does. They made suggestions to these groups – particularly the JSON group – and some of them have been adopted, which is pretty cool. So we are part of the conversation in contributing, it’s an open source community, anybody can really contribute. And it’s cool to see you can make a suggestion that actually turns into something. 

But I’m not rooting for any type. There’s certain types I wish were adopted a lot more, because we’ll say like software application – a company like Salesforce – obviously we do a lot of software application markup. But because not a lot of companies in the software world use structured data, it’s not adopted as much by Google as say recipe structured data where the whole industry has basically adopted it because if you’re not doing it you’re not getting any ranking. So I’m leaning more on the industry to adopt structured data that’s already there.

Rich: Interesting. Can you give us some examples of companies that have instituted structured data markup, and maybe what the outcomes were?

Geoff: Of course, yeah. So SAP is a great example. The results can be pretty incredible and I’ll try to tell you this example pretty quickly. So if you think about how well a search engine understands Overstock or any e-commerce player like Amazon, the sites are pretty straightforward because they’re structured. There are these very well described categories, the categories are all the same format, so when Google comes in they’ll say, “This is men’s watches” or “This is patio furniture”. Then they get to a product page and those are all the same format as well, there’s the product name and price and description. When they come and crawl a site like that it’s relatively easy for them to understand.

If you take an SAP though, they’re a B2B software player, really all that structure goes out the window. Google knows that SAP is really important, SAP has super high domain authority, they want to know everything they can about SAP, but the site is just not organized in a way that lends itself for them to understand. And so when you layer structured data on top of a site like that, it really goes gangbusters.

Our average customer after 12 months just on structured data, grows 62% in terms of organic search growth. There are some that are well into the triple digits. So it can really move the needle.

Rich: So if you’re a…you’re talking about stuff like SAP, you came from Overstock.com. Should small business owners be paying attention to structured data markup, or is this just too far advanced for those smaller companies and really it’s more about the big guys?

Geoff: No, I think they really should. It benefits really any business. We do have a number of much smaller customers, it just seems that we mostly hear about the big ones. Of course smaller ones – especially local structured data – so describing I have a business, it’s located here, here are the hours of operation, here are the reviews, that can be really impactful for a local business to be able to grow the amount of traffic or lead flow, or whatever their business objectives are, to really just grow that channel.

So it is definitely a best practice for small businesses. They don’t typically have to do it to the extent of some of these big sites, but it’s a good thing for them to be aware of and trying to get it live on their websites.

Rich: And that makes a lot of sense because I think you know there’s this idea that there’s a flat playing field on the internet, but obviously there are things that pop up all the time that maybe tip the advantage to big corporations. So if you’re not taking advantage of some of these opportunities, then you may be left behind.

Geoff: Yeah, I always say that small businesses don’t have to be small businesses forever. They can grow.

Rich: Right.

Geoff: And you’ve got to start somewhere. You know the startup world – which really is still small business – all these companies that we work with, including ourselves, have big aspirations. And part of that growth curve is organic search.

Look at brands like eBay and Overstock and Wikipedia, that’s all based on organic search traffic. So it’s a great way to scale a business from small to large.

Rich: Now your company, Huckabuy, offers automated structured data markup. Now at the risk of you asking me why should a company outsource their website design, I’ll ask you why should we consider outsourcing our structured data markup. Why not do it ourselves?

 Geoff: It’s a pretty complicated thing to do. It’s hard to find structured data experts that can come and get your site done, or have it internally. I do encourage people to try, for sure. It’s better to try it and do some structured data than do nothing at all.

It is one of those things that does lend itself to be outsourced. So it’s a relatively complicated task, the target is always changing, we monitor across 50-60 customers in real time the health check of their structured data and making sure it’s dialed in. And that’s just typically not something that can be done at the level that we think is important internally.

So it does lend itself to be outsourced. Even Huckabuy is working on a sort of premium–type model where people can get 50%-75% down the field and by themselves, and then to get that added bonus you sign up. But it’s just a complicated thing in a target that’s always changing. And that’s sort of why we felt it should be outsourced.

Rich: You mentioned earlier that recently the movies structured data had changed. I assume that if I have a lot of recipes on my website and the structured data changes for that, then I don’t need to worry about it if I’m using an automated outsource service like yours. Correct?

Geoff: That’s correct.

Rich: Ok. So outside of structured data markup, what can the average small business owner or marketer do to present their content to Google in a more Google friendly way? What else should we be doing when it comes to that technical SEO?

Geoff: Probably the biggest mistake I see happen with smaller businesses in regards to SEO, and it’s something that can be fixed pretty easily, is the optimization of their navigation. Oftentimes we create the links and the header of a site based on what we think they should be, and not really looking at what do people search for. I call this a demand centered navigation versus basically just calling things what you call them internally.

When you start doing keyword research and you actually see what people search for relevant to your business, and include those links in your header, not only are you making the site more friendly for SEO and getting those pages indexed and having the site be understood by Google, but you’re also basically making the site really user friendly by calling things what they call things.

It’s a great thing to do in terms of just making your site better for SEO, changing your navigation and keywords and relevant stuff about your business. But it’s also a great way to learn about your business. We’ve seen companies change product categories because they see there’s a ton of demand for x and there’s very little competition, so let’s start building products that service this area. And then the site really chases demand for something down the public’s road. It’s a much easier way to come to market than inventing something and hoping that people will take to it. Does that make sense?

Rich: Yeah. And you know if people are using – whether they’re doing their own markup or using a service like yours – let’s just use movies as an example. So we’ve got movies we need to list. What does it look like from the admin standpoint, or the editorial standpoint, of creating new entries? Is it kind of like I’m filling out an online form for movies or recipes, and then the information gets added to my website? Or is this kind of like something I put on afterwards? What is that process like when you’re creating new pages and you’ve got some structured markup data already in place?

Geoff: So the best way to go about it is, if you just look for what fields are required when listing a movie. Like if you’re going to have a movie on your site what do I need to be showing Google to recognize that it’s a movie and have all the appropriate fields built in. And literally you’re just adding that content to the page. The times, the location, the title, all the things that they’re looking for. And then the structured data is not visible to the user, it’s just typically in this little packet of information at the top of the page, or just in line in the HTML you’re actually adding the code that then populates those fields using the content on the page.

You never want to represent something that’s not visible on the page. You just go about it by adding the movie and all the appropriate fields to the page and then adding the structured data.

Rich: Alright. Listen, while I’ve got you, what is the SEO Cloud?

Geoff: So SEO Cloud is what we call Google’s perfect version of any given site. So this was a big change in the last year for them. There’s something called “dynamic rendering”, and it’s a very important – we think – development in SEO that not a lot of SEOs even know about.

Dynamic rendering just means that pages can load dynamically or differently based on what calls them. So if I go to my mobile device and I go to a URL I’ll get one experience. If I go to that same URL on my desktop I’ll get a slightly different experience. And that’s best practice and Google supports it. The big change was that now they said you can give us a version. And the reason for that is that sites have just gotten so complex and the front end coding languages have gotten so difficult, that Google often can’t even crawl. So they’re just saying give us a version that we can at least understand.

That opened the door for SEO Cloud. And if you think of what is Google’s perfect world, it actually looks a lot like Wikipedia. So it’s flat HTML pages, they’re not complex, they load really quickly, and they include structured data. And that’s what we do with SEO Cloud, we’ll take a look at a website, our software converts it into a flat HTML version, we add structured data to the top, and we host it in what’s called a cache [inaudible]. Which basically means it’s posted in memory and instantly available. So we’re just queuing up sort of the perfect experience for Google to come crawl a site so they can understand it.

Rich: At the end of the day, Geoff, it sounds like what you’re saying is that you’re still writing for the person you’re selling to and the person that buys and consumes your products. But if you want that message to get through to them, that you really have to understand what Google is looking for, and as you say, what is the UI and the UX for the Google bot.

Geoff: Exactly. Keep writing and making products for humans, but there is this technical conversation that every website has with Google and it’s sort of ignored. And so what we’re saying is make sure the technical conversation goes well, and all that content writing and all the great stuff that you’re doing will get due credit with Google and you’ll get the proper exposure.

Rich: It reminds me of my earliest days of sales training when they’re like, “Always treat the gatekeeper, usually the person at the front desk, with respect.” Because they’re the ones that are actually going to get you in touch with the VIP you want to meet with. And in this case, Google bot is the gatekeeper.

Geoff: I love that analogy. That’s a really good way to put it, yeah.

Rich: Feel free to steal it.

Geoff: Be nice to the person at the front desk, because that’s who’s going to get you in the front door and get you to meet the important people. The same is like Google is the front gatekeeper, treat them well and they’ll give you the right exposure.

Rich: So Geoff, just to wrap up, where can we find out more about you and Huckabuy online?

Geoff: So the best place is just huckabuy.com. The best way to go about it is to fill out a ‘contact us’ form and I’ve made a promise to this podcast that anybody that fills out a form and says that heard about me on Rich Brooks’ podcast, I’ll actually personally get in touch with you and make sure that we have time to talk.

Rich: Awesome. Geoff, thanks so much for swinging by today, I really appreciate your time.

Geoff: Thanks Rich, that was great. 

Show Notes:  

Geoff Atkinson and his team at Huckabuy have mastered the science of structured data markup and are helping countless business – both big and small – rank higher and see more organic traffic.  

Rich Brooks is the President of flyte new media, a web design & digital marketing agency in Portland, Maine, and founder of the Agents of Change. He’s passionate about helping small businesses grow online and has put his 20+ years of experience into the book, The Lead Machine: The Small Business Guide to Digital Marketing