How can a simple technical SEO fix increase organic traffic by 62% and the number of pages that rank in Google's top 10 and top 100 results by 100%?
This week on The Inbound Success Podcast, Huckabuy Founder Geoff Atkinson breaks down the topic of structured data markup, covering what it is, why it matters, how it works, how to implement it, and more.
Geoff's company Huckabuy works with companies large and small to implement structured data markup on their websites, and he has seen incredible before-and-after results.
Some highlights from my conversation with Geoff include:
Huckabuy is an SEO platform that helps to translate websites so that Google can more easily understand them. It does this through structured data markup and something called SEO Cloud.
Structured data markup was created 10 years ago as a way to help Google parse the data on websites in a way that is more easily understood than HTML.
The open source movement behind structured data is called schema.org.
The benefit of structured data markup is that if Google is able to more easily and accurately understand your website, that increases the chances that it will rank well in search results.
Structured data markup can be used for a wide variety of things, from products to individuals, events, recipes, diseases, and more.
It is most commonly used in the ecommerce industry where having product listings render correctly in search engines is essential for business success.
Geoff says that it is surprising how many otherwise very sophisticated companies are not using structured data markup.
Geoff's average client that implements structured data markup sees a 62% increase in organic traffic within 12 months, a 100% increase in the number of pages that rank in Google's top 100 results, a 100% increase in the number of pages that rank in Google's top 10 results, and a 62% increase in clicks.
Implementing structured data is a bit complex and if you get it wrong, it can really hurt your search engine rankings, so Geoff advises working with an expert or using a platform such as Huckabuy.
I'm your host Kathleen Booth and this week my guest is Geoff Atkinson who is the founder and CEO of Huckabuy. Welcome Geoff.
Geoff Atkinson (Guest): Hi, Kathleen, great to be here. Thank you..
Geoff and Kathleen recording this episode together .
Kathleen: I’m happy to have you here. You have such an interesting background. I'm going to have you talk a little bit more about yourself but the highlights, when I looked at your profile, I was interested to see that you were the senior vice president of marketing, CRM analytics, and demand management for overstock.com, which of course is a household name.
That sounds like a lot of job for one person which we can get into if you want. You've been SVP of marketing for another company, Portico Club, which was acquired. It's a really interesting deep marketing background. So maybe you could tell my listeners a little bit more about yourself and how you came to be where you are now and also about Huckabuy.
Geoff: Yeah, for sure. So I actually was a ski racer in college and when I graduated I had full intention on just being a ski bum and I was moving to New Zealand and that was my plan. And my mom made me take one interview and that interview ended up being Overstock who's based in Salt Lake City. So I could end up having this somewhat real job and still be able to ski a lot.
So I ended up at Overstock on the ground floor. I started in email marketing, worked my way up. I was fortunate enough to put some numbers up on the board and the CEO, the founder, and CEO, Patrick Byrne, sort of took me under his wing. He was my mentor and eventually climbed up and ended up running marketing and analytics CRM. I also ran buying for a certain amount of time.
So yeah, it was kind of an incredible early career path for me, which was very fortunate. And now I have my own thing at Huckabuy.
Kathleen: Yeah. So tell me more about Huckabuy.
Geoff: Yeah, so Huckabuy is SEO Technology. We're a software platform that ... what I noticed in the SEO industry is that, I always looked at SEO as very technical problem. Websites are built for human beings, but I argue the most important visitors, this Google Bot, and that's really a technical issue. And yet the industry's almost entirely services driven. So there's a lot of consultants, a lot of agencies, they do great work. But there's hard adjustments that need to be fundamentally made to the site technically.
And so we take a very software, technical approach. And basically what our software does is it helps translate websites so that Google can understand them. And we do that in two ways. One is through a language called Structured Data Markup and the other is through a a new product called SEO Cloud. And essentially it just makes it easier for Google to understand, the more Google understands the better the SEO results.
Google and SEO
Kathleen: I am so interested to talk to you more about this because this, I think too many marketers feels like a black hole. You know? I mean, the biggest complaint I hear from every marketer that I talked to is they just don't have enough time to stay on top of all of the changes that are happening in the world of digital marketing. And there is a lot and one of the biggest drivers behind that is just Google itself and how quickly it changes its algorithms and the AI behind search.
And it was interesting to me that you mentioned that websites are designed to solve for people, and you said the most important person is Google. I always think of it that way too. But I always think like Google is trying to solve for people. So it's basically doing the same thing. It's just doing it in a much more technical way.
And I think for most of the people I've spoken to, would say that it can be very intimidating even for professional marketers. Like the whole topic of structured data is intimidating.
Geoff: I think SEO in general, most marketers find intimidating because it's a lot different than other channels of marketing. It's less touchy feely. And much more technical and pretty hardcore and Google is moving a million miles an hour.
What I say to sort of overcome that fear is to just be directionally correct. So if you're aligned with Google and the general direction that they're heading, you'll get results as they update the algorithm. If you're sort of back using tactics from the 1990s or whatever, that's a problem. But if you're directionally looking ...
But yeah, a lot of people do have fear around SEO and that's, that's too bad.
Kathleen: Yeah. I always think about it as going back to the people thing. If you're trying to game Google, ultimately that's going to come back to bite you, because the technical ways that it does things are constantly changing. But if you stay true to that goal of solving for the person, that will never go out of style. I mean, there may be ways you can improve upon that and refine upon that, but it's never going to hurt you to solve for people, right?
Geoff: Oh, for sure. Yeah. To an extent. So I would say you're absolutely correct. If you're fundamentally solving problems for users and making a better user experience, Google is going to figure that out, and give you the appropriate ... you're gonna end up getting back links. They have ways of figuring that out.
But there are some things that if you do make a sort of major technical mistake, no matter how good the human interaction is, Google's going to be left scratching their heads. And so that's really what we're trying to do is say, "Hey, you can do whatever you want to the front end, we're just going to ensure that you have this wonderful interaction with Google because you're going to put a lot of money and time and energy into what the user experiences. And that's wonderful. Let's make sure it's as good for a search engine." And so that's really what we're trying to solve.
What Is Structured Data Markup?
Kathleen: So for the marketer who doesn't really understand structured data well, can you explain that? What is it? How does it work? What is it used for?
Geoff: Yeah, yeah. So, for years and years and years, the way that Google and other search engines understood websites was through crawling HTML. And if you've ever viewed HTML, it's very confusing. They're only so much in HTML that's actually useful for them.
So years ago, over 10 years ago, search engines and academics came up with this language. They said, "There's gotta be a better way for us to understand. And so let's build a structural language that'll take unstructured websites and make some sense behind them."
So they created this language and you can really use, it's called Structured Data Markup. The sort of open source movement is called schema.org. And you can actually have markup for almost anything. So you can't represent anything on a page that's not there. But if there's information on the page that's about a human being, if it's about an event, if it's about a sports score, a recipe, almost anything that's out there can be represented via structured data.
Once you do that one, Google understands each and every page a lot more clearly because it's now receiving the information in a structured way as opposed to in HTML. And they understand the site more and more and then you get the appropriate search exposure for that understanding.
But also they use it in very interesting ways. So as you see now within the search results, they're what they call Rich Cards. And Rich Cards are enhancements to search results that are coming, at Google fast and furious. So if you search for a recipe and the recipe just shows up, that's being powered by structured data. If you search for a movie time and the movie times just display, that's structured data.
And essentially they're, the reason that they can now leverage it is that structured data is authoritative, so it's factual and because it's being presented in this nice format for them, they can actually now integrate it into search results.
What's also important about it, is that if you think about voice search, if you ask a question to Google and it gives you the answer, the answer is coming via structured data, because it's basically just reading their search results, which now has this structured information to it.
So it's becoming really important in the voice search. As voice search gets adopted as more comfortable for people to use, whether they want to buy a movie ticket or they want to order something off Amazon or whatever it is, structured data needs to be in place for that interaction to happen.
So it's becoming a very universal language that Google loves because it makes their search results better. They understand the internet more. They're a big proponent of it. Probably the biggest, and it's getting adopted a much faster pace than it ever was.
Kathleen: It's interesting to me that you mentioned voice search because I definitely can see where, especially if you're transacting over voice search there, that's so important to it because there's a degree of specificity that's really important for voice commerce. But also the thing that I've been thinking about a lot just for our own marketing is as we move more in this direction of voice search, we're going from an environment where there are multiple answers and you're trying to be either at the top, or close to the top, to an environment where there's only going to be one answer.
Kathleen: In voice search.
Geoff: That's what's wild about voice search. Yeah. They give you one answer. It's, easy, but you're either number one or you're not.
Kathleen: Right. Winner takes all, like the stakes are really high. If you do it right and you land that spot, you're gold. And but there's no second place, right? So it is scary and that's why it's so important to get it right.
What Are the Different Types of Structured Data?
Kathleen: So talk me through some of the different kinds of structured data that there are. Because you rattled off a few, but I know that it's a pretty long list. And I don't expect you to list all of them, but I just think for people listening it would be interesting to get a sense of the breadth of options for how you can use it.
Geoff: Yeah, I'll give you some of the important ones and the ones that they're using a lot.
Product markup is probably the most, it's the most adopted markup across the Internet. So products being able to show price and skew number and all that. There's a lot of fields around product markup.
Individual markup. So a person, talking about a person, where they're from, what their job is, all that information.
Event Markup is probably the second most leveraged. And so when you look at ticket sites, and race sites and companies that are based on events, it's sort of that, one winner wins all type situation where they're taking up everything above the fold as being powered by structured data.
Medical diseases and, man, I mean weather, pretty much everything is able to be marked up.
One of the important ones for us is software application. So if you think about e-commerce sites are already really structured the way that they're built. SaaS sites are not all that structure goes away. There's not really nice categories, product pages.
So when a search engine comes to like a Salesforce or an SAP who, we're fortunate enough to have as customers, they know the site's really important, but there's none that structure for them to be able to really understand what they do. And so even though they're huge companies, Google's left sort of scratching their head. And software application markup ends up being really powerful to communicate what a software company does to a search engine.
So there's a few examples. There's some fun ones, like recipes are used all the time, and that's when you can ask and it will give you step by step by step directions. That's actually being done by structured data, kind of across the board. And it grows. It keeps growing, which is exciting as well.
Kathleen: Yeah. You know I have a funny story about this. You talked about the person markup. And we actually had somebody helping us with our structure data recently and they discovered something which I already knew about, but I didn't know quite how to fix it, which is that if you Google my name and go ahead and do it, if you're listening. Because this is actually pretty funny, you'll find results that are me, but you will also find a result, I believe from Wikipedia that tells you that I invented one of the first computer programming languages in the 1950s or it was one of the first of the modern era. I wasn't born then.
Geoff: It says you're 97 years old here.
Kathleen: Exactly. I found out about this through a Google alert I have on myself. And because somebody wrote up an article about the top women in the history of computing and they tagged me on Twitter, and I tweeted them back and I said, "Well I would love to take credit for this because it's actually way cooler than my real life accomplishments. Like I was not born at the time."
So the folks helping us with their structured data, we're like yeah, I think there's an easy way to fix this if we just use some markup language. So I thought that was hysterical. Part of me doesn't want them to fix it because it's sort of this novel, interesting thing. But no, accuracy is important.
Google's Knowledge Graph
Geoff: If it's helpful, I can tell you exactly how that panel is populated. It's actually called Google's Knowledge Graph that's there ...
Google's Knowledge Graph panel is a fascinating thing. Google figured out that humans think in terms of association. So like, I'm from Boston, I think of Boston, I think the Patriots, you think of the Patriots, you think of Tom Brady and Bill Belichick. It's just how we think, we have context and it's how we think.
Google realized they were missing all these connections. And so they built a thing called Google's Knowledge Graph. And the only time you see it is when you see that box in the top right panel and it's populated. It's manipulate, it's not manipulated, it can be populated very easily using structured data.
But it often gets it wrong, because it'll make these associations. And then so, in Google's mind you're the same person but that's not reality. And so there are ways that you can fix it, but their Knowledge Graph is pretty impressive actually. Just the fact that they thought of that and they're working on it is pretty cool.
Kathleen: Yeah. I mean that same concept is how, when we talk about content strategy with my team, we're always talking about, not to get into like overly technical terms, but latent semantic indexing. Like what are these related keywords?
Because it used to be that you had your keyword and you just hammered the keyword home constantly over and over. But that led to some really, like, awkward writing and reading experiences. And now it's nice that you can be freer and looser, and write - and as though you're talking or write like a real person would write, if they weren't writing for a search engine and the search engine is more attuned to picking up on what it's really all about.
Geoff: Yeah. That is a part that I feel is a bit of a black box is what content they like and don't seem to be changing a lot. And I hear our customers complaining about that a lot, where they'll say, "Well, we're worried that it's old content." I just, I don't know the answer to that.
Kathleen: Oh, if you think if you figure it out, you're going to make millions. Because I will tell you, I have an awesome team of people who do content strategy and I think our content is really great and it's always, like, funny, but at the same time terrifying to watch the fluctuations that happen when the algorithm updates occur. It's just, there's no rhyme or reason. So I would love-
Geoff: I think they're testing heavily right now in that arena, their language processing, which is ... they'll ultimately probably figure it out and then be a little more stable.
What Impact Does Implementing Structured Data Have?
Kathleen: Yeah. Hopefully the tests come out in our favor. So back to, to markup language, structured data. I think the big question, two questions on everybody's mind when they hear a conversation like this is one, so what will happen if I do this? What's the result going to be? And two, how do I do it?
So let's start with that first question. I mean you've worked with lots of companies. What kind of impact have you seen when these companies have either implemented structured data for the first time or cleaned it up or improved it? What have you?
Geoff: Yeah, so I'll give you a stat. Our average customer, after 12 months, their organic search traffic is up 62%. Now they do some other things, but a lot of that is structured data.
So in the early days of structured data, Google would constantly say, "No, it's not going to do anything to rankings. It's not going to do anything other than give you potential rich snippets." What we've found is that it does impact rankings, and it also gets you ranking for a lot of new keywords. So non-branded, sort of mid to long tail terms where there just wasn't an association between website and the keyword in the past, because you layer on this authoritative language, they'd start picking it up.
So our average customer in terms of ranking keywords, that means keywords that rank in the top 100 of Google, they grow over 100% on that metric in 12 months.
The next metric is if they have a good domain authority, those links will start showing up in the top 10 and start getting eyeballs. Again, that grows over 100% in 12 months on average across our customers. And these are some big customers. I mean, they're not just, small sites, they're just growing crazy.
And then the third is, yeah, the clicks, which is ultimately what we care about. 62% growth. And because it's mid to long tail terms, that's sort of the bread and butter of any good SEO strategy, where you're picking up very specific queries that convert.
And so we want to look at you, then look at revenue. We don't actually have the exact stat around revenue, we're kind of working on being able to pull that. But it's again, up here, in triple digits because you're getting the keywords that you really care about those non-branded, mid to tail terms, new customer, hype keywords that that really moved the needle.
Kathleen: Yeah, that makes sense. They're very bottom of the funnel if they're searching terms that specific. And so it should be easier to close them as customers.
Now when you say, when you cite these stats of improvement, as you said, these are smart companies with really good marketing teams. It doesn't sound like these are companies that have like never played with structured data before, correct?
Geoff: You'd be surprised. It's kind of wild what we see. So some companies have never even heard of it, that ended up becoming customers. And some of them are some really big names. So the level of sophistication, I think you talked a little bit about this in your last podcast, where the e-commerce world for example, where the margins are so much thinner and if you're not doing SEO, you're just sort of good luck, you know? Versus software, where the margins are so much bigger. And if you have a good sales team, you could go without doing SEO for 10 years.
So we do run across some really big companies that are just starting to figure this out. And that's really where we come in. So we work with companies that have in house IT, SEO teams, they have agencies, they've kind of everything firing. And then we have others that might have one person part-time on SEO, but they realize it's very important. They realized structured data is very important and they just get it. And so they buy Huckabuy.
How to Implement Structured Data
Kathleen: Cool. I mean it's impressive to hear those results. In terms of how to do it. I know that there are ways to DIY it. The code is available. I've been to the website where you can see, it's in the public domain, like some of the different Structured Data Markup language. But I also know from having visited that website that it's not entirely intuitive.
Talk me through the different options for how to put this in place. Obviously one of them is to use your platform, but it's certainly something that that you don't have to buy your platform to do. What kind of technical expertise do you need to do this?
Geoff: Yeah, I mean it's usually done best by engineers, software developers. If it is done in house, it is a pretty complicated language. The fact that it has to be correct, and if you do have errors or warnings that you kind of get dinged for that makes it ... You want to be kind of careful with it. You don't to just throw it up there and like, let's see how it does. You definitely want to have someone that knows what they're doing.
And so we see sort of all levels. There are some sites that it's built in. So if you're using Shopify, for example, you get product markup. You might not get it in the level that Huckabuy would provide, but you're getting it, and you're getting reviewed markups. So you're going to get those fields being populated on search results.
And yeah, there are some ways to go about it by going to schema.org. And usually the first step, usually it gets initiated by an SEO person that's into it, and is doing their research and sort of figures out that this is a big thing. And then they start working with a front end developer and they figure out sort of a programmatical way to do it or they start looking at outsourced options.
The reason that ... and we see it's pretty rare for us to see a company that's really doing it well internally. If they are, they're typically like your Overstocks of the world. They're in the e-commerce space, they've been doing it for years. They have developers that have worked on it for years.
Well what's funny is, in the software space, SaaS, it's very rare that we come across a customer that's doing it. Because if they don't have developers, usually the really good developers are dedicated to the product, not to the website. They're usually WordPress sites and they just don't have the in house expertise to sort of pull it off.
So yeah, I mean I'd recommend looking at schema.org just get started and then, either working with a developer or pursuing some of the third parties like Huckabuy, that can do it at a world class level. So it is an important part, I think a very important part of SEO and getting it right, makes a big difference. And you can change your results pretty tremendously.
Maintaining Your Structured Data Markup
Kathleen: And is this a dynamic thing? And what I mean by that is, let's say I go in tomorrow and I implement structured data across my site. Is it like "Check, I can move on now to the next thing"? Or is structured data kind of like Google's core algorithm updates where the milk and the grocery store, as soon as you know where it is, they move it. Like is there an element to it that needs to be maintained?
Geoff: Yeah, unfortunately it is the latter. They are changing the target quite a bit. It's in a good way in that it's developing and getting better, but it is not something that you can just do and have that box checked. You know, it might be checked for six months or something, but usually not even that.
But they are changing it and it's getting better. And one of the nice things, one of the unique things about Huckabuy is that we're literally having hundreds of thousands of conversations on behalf of our customers with Google each day. And so we really know ... That's a unique position to be in as an SEO company. Because really know other SEO company does that. So we know when they change, and what happens when they change. We can adapt to that really quickly.
So yeah you do have to keep an eye on it because it is a moving target, for sure.
What Does It Cost To Implement Structured Data?
Kathleen: Now of course the next question that I have that I'm sure anybody else listening will have is like what does it cost to do this? Is this accessible? Like, who is this realistically accessible to in terms of the investment needed to do it?
Geoff: So for, at Huckabuy sort of ranges from on the low end, around $1,000 a month all the way up to some bigger customers are quite a bit more than that.
Usually it's in the range of $1,000 to $4,000 a month. If it's multiple domains, it gets a little bit more expensive. And you are really kind of buying the Ferrari. There are cheaper options out there, but with Huckabuy, you're getting world-class structured data and that's guaranteed.
We also have a bit of a strategy behind how we do structured data. So if you're SaaS site, there's a different playbook than if you're an e-commerce site or if you're a travel site, there's a different playbook.
So, but yeah, that'll give you some range. So, pull $1,000, $2,000 out of your PPC budget, plunk it down into the structured data and I think you'll get a much higher ROI.
Kathleen: Which arguably though you could easily spend that much hiring somebody to do it, if they were really doing it correctly. Or if you look at the time value of money, if you're doing it yourself, you could probably be spending a lot more than that is my guess. Diving down that rabbit hole of trying to figure it all out.
Geoff: Yeah, you definitely could. It takes some time though. Yeah, it'd be a commitment.
Does Your CMS Affect How You Implement Structured Data?
Kathleen: Yeah. Now speaking in kind of more technical terms, there are so many different types of websites out there. You mentioned WordPress, you mentioned Shopify. We happen to have our website on HubSpot. There's a million CMSs. There are hard-coded HTML sites. And if somebody wants to implement structured data in general, is it something that's easy to do on any kind of a site?
Kathleen: Great. Yeah, and I know on HubSpot you can do structured data. I mean you definitely have access to be able to hard code things. So I'm sure it'd be a pretty simple exercise if you've got a good developer.
So anything else you think somebody listening should know about structured data?
Geoff: No, I think we've pretty much covered it. I think it's going to be very interesting to see how it continues to grow.
You know, we talked a bit about voice search, I think people are going to get a lot more comfortable with voice search to the point where they'll actually, when we said, it's either win or go home, people will actually be able to through talking in a different way, "Show me option two, show me option three." I think it'll get where people get sort of trained on how to do voice search.
And so what's interesting about Google is they're very good at dominating markets and sort of changing trends. So as they are just kind of pushing people into voice search, there's going to be these trends that end up developing and structured data's gonna be right smack dab in the middle of that.
So it's going to be an interesting few years ahead around structured data and I think the importance of it is just going to grow. But no, I think we covered it pretty well.
Kathleen's Two Questions
Kathleen: All right. Well then I have two questions for you that I ask all of my guests. So it is that time.
The first question is, when it comes to inbound marketing specifically, is there a particular company or individual that you think is really killing it right now?
Geoff: Yeah, I do. I would say one is Concur, the software company. They get an enormous amount of inbound leads, via SEO. And it has scaled that business like crazy. I mean it's over 50% I believe. And they're feeding like 400, 500 some account executives. So they're super impressive.
They're a customer, and I've just been like, "wow, you guys are doing amazing things." It's rare to see a software company invest so heavily in SEO and it's just working for them like crazy.
And then the other one would be a travel company out of Florida that's growing nationwide now called 360 Blue. And they started when we started. I mean, this company has grown so much since we started with them. It's just amazing. I mean it probably tripled in the last two years.
They just get it. And they have a really smart CMO and he thinks about things the right way. And it's just really cool. They're doing acquisitions now and just the businesses going crazy. So those two really come to mind. We get to sort of see as I have customers and, I get impressed. And those two definitely impressed me.
Kathleen: I can't wait to check those two examples out. So that was Concur, and did you say 360 Blue?
Kathleen: All right, well I'll put those links in the show notes. So if you're listening and you want to check them out, head to the show notes for that.
And then the second question is, we were talking about this before we started recording, and being in digital marketing today, it's like drinking from a fire hose. So personally, how do you stay up to date and on top of all of these changes?
Geoff: Yeah, that's a great question. So at Overstock, I actually had two really smart people working for me that is all they did, was just figure out what's going on. Now I have a CTO that really knows this stuff and instead of ... I've kind of gone a little counterculture on this. So instead of going to all the conferences like I used to and having my ear to the ground, I am taking a much more technical approach now.
So we just listen to Google, Google I/O, we listen to all the talks, we really scour their technical documents. And we're sort of figuring things out by just really listening to them. Like there's this crazy phenomenon now that really, not a lot of SEOs know about called dynamic rendering, where you can set up a whole different version of your site just for Google.
And that's what our SEO Cloud does. You just don't hear about it, but we paid attention and now you can do that. And it's like you can create that Google's perfect world version of your site.
So that's kind of how we do it. CTO is very smart and he watches them closely, and it kind of feeds me ideas and then together we collaborate on what the next product looks like. So that's how I do it. Probably a little counter-culture.
Kathleen: No, but that is really great advice. If you only looked at one source, Google is a good one to choose from. Great.
How to Connect With Geoff
Kathleen: Well if somebody's listening and they want to learn more about Huckabuy, or they want to reach out and connect with you, what's the best way for them to do that?
Geoff: Yeah, so the best way is just go to huckabuy.com, fill out a contact us form or for your listeners, just contact me directly, if you're interested. It's just Geoff with a G. G-E-O-F-F @huckabuy.com. And yeah, I'd love to hear from your audience, hear what they think, hear their experience with structured data, and if they see it as an opportunity, obviously we'd love to help them.
Kathleen: Great. All right. I will put all of that in the show notes and thank you so much Geoff. I've learned a ton about structured data is definitely one of the more complex topics, so I appreciate you breaking it down for us.
Geoff: My pleasure. Thanks for having me, Kathleen.
Kathleen: Thanks. And if you're listening and you like what you heard, of course I would love it if you would leave a five star review on Apple Podcasts, or the platform of your choice. And if you know somebody else who's doing kick ass inbound marketing work, tweet me at WorkMommyWork, because I would love to interview them. That's it for this week. Thanks.
Want to stay updated when the podcast is released?
Drop us your name and email address below and we’ll send you the show notes every Monday!
Here Are Some Related Articles You May Find Interesting