Schema Story: Mike Arnesen, UpBuild

Martha: Hi and welcome to Schema Stories my name is Martha Van Berkel and I’m the co-founder of Schema App. Our schema stories bring you real people doing Schema.org markup, they share their tools and insights on how it affects their business and their agency. Today I’m joined by Mike Arnesen from UpBuild

Tell us a bit about yourself your background and your company.

Mike: I got a degree in history and fell into web development through WordPress. Eventually discovered SEO and found that it was the coolest thing I’ve ever had been exposed to in my life and decided I wanted to figure out how to make a career in that. So really, the key takeaway is that I geek out a lot and towards the end of my journey I discovered Schema.org. Semantic markup is the coolest thing because it has everything that I love – I can play with SEO, I can play with web development and I can make it mean something through applying this cool markup. Fortunately, I was able to play with creative ways of doing that, well outside of the norm, which I really like. Today I’m the founder and CEO of a company called UpBuild, and that was it. I guess you could say a large part was a result of digging into Schema.org, doing funny things with it, with like Google tag manager for example, which we’ll talk about later. But be able to do that, made me want to actually have a company where I could do that all day. So I started UpBuild as a technical marketing agency so I essentially hire people like me who love to geek out about this stuff and then let them do it. I still get to get my hands dirty pretty often as well.

Martha: In your own words, describe Semantic Search.

Mike: I would break it down as, the search of a couple years ago was very much based on the words that are contained within web pages and so if you’re looking for a specific thing in Google search, Google’s going to try to find the pages that are relevant to that specific phrase that you’re looking for. With semantic search, what Google is trying to do, and other search engines of course, is understand what you mean by that and understand that pages that might not even necessarily have that word or that phrase in their content, they want to understand that those pages are about that and would be a good resource for that. So, semantic markup is something that supports semantic search because we need to have machines be able to understand content, as well as, we do, as humans and that’s kind of the dream and the goal of semantic search.

Martha: How does semantic markup fit into your overall company’s strategy?

Mike: I would say that semantic search is a major part of the service that we offer to our clients. It’s kind of one of our key deliverables that we do as part of an engagement. So, we’ll of course, we hop in and we always want to look at the sites technical foundation and framework first, make sure everything is in line because it’s no use having semantic markup on the site if search engines can’t even get to the site or access its code. So, we’ll make sure all that’s cleaned up and then we kind of, start looking at, not just the keywords that we want to use on a site but the topics and the themes that the site’s talking about and from there we’ll still do some of that traditional keyword strategy but we try to identify what are the entities that we’re ultimately going to want to have on this site.

What other kind of topics, what potential kind of recognized entities in Google, are we going to be able to relate to that website, into like what Wikipedia pages are we going to be able to hook into this client’s content, freebase entities, things like that and we kind of start thinking about that all from day one and then eventually get really deep into a full fledge markup strategy, of ‘here’s what we’re going to do, through micro-data, through JSON-LD and stuff like that’ and then work with the clients to get added to the site.

Martha: Often people are looking at Schema.org tactically in search like ‘oh I want to get a rich snippet right so I’m going to add this structured data’ and to me there’s like a bigger strategic play that you can do with semantic search. Can you talk about your view on that?

Mike: I view it as a long-term strategic play and it’s kind of funny because we have a little bit of a divergence in how we think of it. At UpBuild, internally we talked a lot about what is this going to mean for the future of the web site or working on. Maybe today local attorney isn’t going to be something that’s going to generate a rich snippet but the hope and speculation is that it’s going to ultimately lend a better ability to understand that content to Google so we’ll talk about this stuff. Our ‘in’ with a client is to say ‘okay well this is going to help you get rich snippets, it’s going to get out of the ratings reviews stars and Google or it’s going to allow your price, and your items availability to show up, but also while we’re doing it, let’s do all this other stuff that, you know, may provide a benefit,’ and we try to get that all taken care of and use those kind of short-term wins as a way to usher in something much bigger that’s much more strategic and long-term.

Martha: What else do you use it for? So for instance, for search, I know you’ve been a thought leader around for semantic analytics. Can you give us some details on some of those things that you’re playing with, about how it can be used differently, for different intelligence or for different insights for a client.

Mike: Yeah, the idea of semantic analytics, essentially taking your semantic markup data and joining that with your analytics data, isn’t a unique thing, really, by any means but it’s kind of bridging the gap between two disciplines. You can have an SEO who is really great at semantic markup and just SEO in general, it doesn’t really have a good solid competency in analytics and that’s fine. Actually, and you can have an analytics expert, who can go through a site, and like tag everything, set up all the data like for Google’s enhanced e-commerce features, like there’s so much data layered in there that perfectly overlaps with stuff we do on the semantic side and this, the idea, really spawned as I, within another agency study prior to the one that I am at now, where we had an analytics team that did their own thing and an SEO team that did their own independent thing and we didn’t really talk as much as we should have or could have and I was realizing like, the analytics team is putting all of this code on the site to say like ‘okay, what’s the name of the product, what’s the price of the product’, and that’s just going straight into google at Google Analytics and then the other side were doing all this work to put it in that same exact data but in Schema.org – microdata format, and so why not bring the two together? So semantic analytics is about bridging that divide and kind of just being aware that, you know, there’s value, on both sides. You have this central core bundle of data and you want to be able to use it, ideally, for both search and analytics, so it’s all about bringing those two together.

Martha: Any other use cases? You mentioned Google tag manager, do you want to speak a bit about how google tag manager plays into Schema.org?

Mike: Google tag manager I think it’s really, really great tool for web marketers to have access to. It’s great if you’re working on analytics implementation for your clients but it’s also great if you just an SEO or if you’re both, that’s even better. One of my developer colleagues at one point, really broke it down, and I was like oh yeah he’s totally right. He really just said, you know, ‘Google tag manager is simply an abstraction of Javascript.’ So it brings a lot of the power of just JavaScript in general as a scripting language but makes it a little more accessible and kind of contextualized it for our needs as marketers and analyst. So using google tag manager some of the stuff that you can do is it was really cool is you can actually harness the power of JavaScript and grab things off of the page to pretty much anything you wanted and you know the, the original, I think, intention, when Google created that, was that you’d be able to pull things like the product name off the page and use that in your analytics. So one thing led to another, I was thinking, okay why don’t we take that and use it for a bunch of different things.

So we can actually, if we already have Schema.org markup in line, on the site we can say, ‘Okay Google tag manager, go look for you know, item equals name and grab that value’ and we can put that into analytics. If you don’t already have something like JSON-LD javascript markup, on the site,, we can take that value, and build out around JSON-LD files as well. So, lots of cool stuff there and a lot of potential for display with that data and bring it into your analytics with the goal of really, what I love, is being able to understand your pages semantically, to not just say, ‘Okay here’s the URL and it got 10,000 visit this month, this second URL got 5,000 visits’ and you don’t necessarily know why, especially because your data, of course is, much harder to come by then it was five years ago, but with semantic analytics one of the things we can do is bring in the different item types, so we can actually look at our content by the type of semantic entity it is, we can look at you know, ‘a specialty’ or ‘an about’ or an ‘about topic’ for a piece of content, sort by that.

A classic example is sorting blog post by authors bringing that information in semantically and then even things like, you know, looking at bounce rate by sorting based on availability on an e-commerce site. You can see in stock, out of stock, how does that affect conversion rate and of course its stickiness of the site, and we need to work on that? All these kind of things that once we get deeper than just at the top level page we can understand stuff that can actually be really, really insightful beyond SEO, beyond analytics. Like they can actually fuel business decisions, like ‘do we need to fix our supply chain because we have all these out of stock products that are driving people away from the sites because they can’t find anything they want.’ That kind of stuff is really, really cool and I think as an industry we were just you know looking at the tip of the iceberg here, we have a much deeper to go and there’s so much potential underneath.

Martha: One of the things I get asked often, and especially with agencies that are adopting Schema.org early in their journies, is how to articulate the value. You’ve talked a bit about the value from the business intelligence side and structured data beyond the SEO purpose. Can you talk a bit about when you talk to your clients, how you articulate that value.

Mike: Yeah, it’s a bit of a tricky one, because really when you’re working with a client, of course, your goal is to help them meet their business needs and to increase the metrics that they’re being judged against and we can always informally say, like, our goal is to make our clients look like rock stars to their bosses. It can be tricky to sell Schema markup as like ‘hey this is going to make the web a better place’. Usually if you have a marketing director, they’re not going to care, unless they’re awesome geeks. Actually funny story, we work with a law firm and the lead partner at the law firm is like, ‘This is the coolest thing I’ve ever seen and if I didn’t do law I would want to work with you guys, to like do, semantic markup because this is so neat!’ But yeah, semantic analytics has been a helpful tool, in kind of selling schema markup, so we can say like ‘Hey this is going to provide you greater intelligence on how your site is doing, we can you know pull better data from it to help you make business decisions, it’s also going to get you the short-term benefits of the rich snippets in search.’

The cool thing now that, of course, Google is doing, like having more and more support for json-ld markup. So things like recipe rich cards we’re starting to get into it. To be able to say like ‘hey you have recipes on your site or even you have like news articles on your site, we’re going to pull that data off of your pages, reformat it and kind of rearticulate it using google tag manager, push it out with json-ld to allow your site to have this new feature from google without having to even talk to your developers. We can do all that on our own and make that happen. It’s just such a crazy cool frontier still like even that I think is a selling point in and of itself, to a degree it’s like ‘hey we can get you something that’s going to benefit your SEO and you don’t need to spend extra money to have it implemented’.

Martha: When do you think semantic search got this going? We’ve talked about seeing that the big movement towards JSON-LD. Any other sort of trends or services like ‘This is where I think it’s going to be in 12-24 months’.

Mike: Yeah, I mean the thing,  the funny thing about semantic search is like if everything goes the way that we wanted to go as SEO, then as marketers, and it’s just maybe Google fans, is you know in five, six years, 10 years, semantic markup isn’t going to be needed. You know Google’s going to get to be able to understand our content at a deep enough level where like they don’t need us to tell them that like, ‘okay this website is like item type schema.org/organization’, like they know that it’s a company, like they’re going to find the signals and those kind of, unique identifiers that say like, ‘okay this is a company with a contact page, like here’s their location, we can like, you know, know if we’re at 99.9 percent surety, we don’t need necessarily the markup to do that, but right now, in the here now, I absolutely think that it’s such a benefit to be able to kind of support that, and I mean it’s always been my speculation that you know, long-term, Google is going to remember the sites that are doing things to help them increase the quality of their product.

I think there are a number of reasons of course, that  they kinda do this stuff even though in the long term it might not be something we continue needing to do. But I also see things that make me hopeful for just with the the melding of web development and SEO in general with you know html5 kind of limited set of what I would absolutely consider semantic elements. Just like the navigation tag or like a size and sections and articles in the actual HTML markup that’s not even something where you need to get into schema.org. It’s kinda interesting to see those two come together and I would anticipate, in the future we may get even closer to bringing those two worlds together, as just as kind of a facto new system.

Martha: Where  do you go to stay on top of trends and things changing in SEO, in the semantic space. Recommendations on homework, where people should go hunt for information?

Mike: You know, I’m not as in the weeds I used to be, which is a kind of a bummer, but of course I’m doing a lot of other things. When I want to figure out what’s going on and know exactly what changing I always go to Aaron Bradley’s semantic markup community on Google+. That’s a great place to check out just what’s going on, what’s changing. Not everyone has this privilege, but, I’ll email Aaron and say “Hey, so what’s up?”. He’s a semantic markup wizard, maybe you could say he probably taught me everything new I learnt about semantic markup in the last like, two year. Just a great guy, so check him out. That community on Google+ is like the best place, in my mind to talk about semantic markup,  to get your questions answered and of course learn what’s new in the space

Martha: Anything else you want to share?

Mike: I would just leave everyone with, hopefully, you know, if you’re not already doing Schema.org markup, or any kind of semantic markup for that matter, now is the time to do it and it’s it can seem really challenging if you don’t have a development background but once you understand those basic principles it’s such an easy thing to get into your tool belt. To say like ‘hey this is something I can do now’ and it’s such a satisfying and rewarding journey once you kind of go over that precipice and then you finally reach the point where, like okay, ‘it all clicked and now I’m just like coasting downhill with semantic markup’. It’s really cool and fortunately there’s some pretty cool tools out there (Schema App) that I think you’re going to make people jobs a lot easier and in the future as well. So check out some tools and some of the great documentation, as well, on Schema.org to get you started.

At Schema App, one of our core values is to always be learning and teaching. That’s why we love talking with other structured data experts!