In this interview, Blake asks Geoff about his background, how he got started in his career, and how he grew into his role as SVP of Marketing at Overstock.com before starting his own SaaS business. They then dive into discussing a more technical approach to marketing and SEO optimization. This means optimizing for humans, but also for search engines. Along these lines, Blake learns how to use structured data (also known as schema) and dynamic rendering to succeed in SEO.
Click here to listen to the full episode.
Blake: Let’s talk about you when you were just starting off in your career. Was Overstock the first marketing opportunity you had or did you work for other companies before that?
Geoff: I grew up in Boston, I went to school in New Hampshire at Dartmouth and I was a ski racer. I got to travel one year on the U.S. ski team, which was awesome. From traveling all around the world, I really fell in love with the West and I did want to do something in marketing.
After I graduated, I knew I wanted to be a ski bum. I was actually moving to New Zealand, and my mom, after four expensive years of me going to Dartmouth, was like, “You have to do some sort of interview before you leave.” So I got an interview to work at Overstock in Utah. When I came out to Utah, I had one of those sort of really rare bluebird powder days at Snowbird when I visited, and sort of fell in love with the place. So I started at Overstock really at the ground level, but I put in the work and that really launched my career.
Blake: Then you end up going on to have several roles at Overstock. So you’re not only starting off as an email marketing manager, but then beyond that you actually grow into roles like the SVP of Marketing. How did you do it, and what advice would you give to others that are trying to get to that SVP level in-house?
Geoff: That’s a great question. I’ve never asked for a promotion and I’ve never asked for a raise, and I think you really have to almost over-qualify yourself to get promoted. I was really young. And so, unless I had a track record of results where everybody in the room — everybody in the company — were like, “how is this guy not getting promoted?” then I really didn’t think I deserved it. I believe in a meritocracy. Fortunately for me, at the time, the culture really was a meritocracy. So, for example, I wasn’t even an email marketing manager — I was just on the email team, but we grew that thing from like 50 million to about 100 million in a year or something. And at that point, the CEO started to take notice. We became friends. He’s also a Dartmouth alum and he became my mentor and then we just consistently drove growth. I was always on the customer-facing side at first, so I did website optimization — obviously our SEO is a crazy good story, we went from zero to 300 million dollars in revenue. Ultimately, nothing drives me more crazy than people that expect promotions without really doing something. My advice would be to make it so obvious that you’re driving growth and driving the company values that it’s just a no-brainer for you to move up.
I was the youngest SVP ever there, and when I got there, I really did feel like it was deserved. You know I hadn’t cut any corners or received any favoritism. I do tell young professionals, you really got to think about finding an opportunity at a company — typically that’s a little bit larger because you’ll learn a lot more — that has a career track. If you do well, you’ll move up. Also, you should have a really key mentor in place. I think people don’t really think about that when they apply for jobs and get jobs, especially early in their career, but having someone that really cares about you as a mentor that wants to see you succeed — someone that wants to grab you by the bootstraps and pull you up — is probably the biggest break you can get in your career. It really doesn’t matter how much money you make until you’re like 40, so don’t worry about the money as much as what experience you are going to get and what are you worth in the open-market by the time you hit 40. That’s sort of how I coach young people. Think about the experience, find yourself a good mentor, and then do a really good job that moves the needle.
Blake: Sure, so you mentioned the story of growing the team from zero in SEO to 300 million. What was it originally that really caught your eye about SEO that made you kind of fall in love with it?
Geoff: I remember I took a phone call with a guy named Paul Bremmer that was recommended to me to speak to, and he basically just explained what SEO was — I had never even heard the term. At the time, we were spending something like $15 million a year on advertising at Overstock, and so I was like, “Wait so if you optimize your site, then you’re gonna get all these organic search visitors and they’re actually better visitors than any other channel.” And so that was really appealing. That was a technical problem, but if you got it right, and you had this great communication with a search engine, then these amazing results would come from it. Then the analytical side of my mind was running the ROI numbers. It was like, well, if you get to 100 million I don’t care how much you’re spending on the team, the ROI is gonna be so much better than any other channel. So, once it sort of dawned on me and I kind of explained it to Patrick — who was the CEO — we both just sort of got it. And fortunately he invested in me, and invested in our team, and we got all the resources we needed. And we really did almost everything you possibly could do to try to just grow, and grow, and grow, and that’s what happened.
Blake: You do a lot with structured data and schema nowadays at Huckabuy — but I’m wondering what role that played, if any, in the success there at Overstock. Was that something that was on your radar at the time?
Geoff: It was. I felt like, you know this guy Paul, he was so ahead of his time. He was thinking about SEO and how it worked. And even in like 2007, he was saying to me, “Geoff, structured data — this is the future. This is how Google wants to be communicating with websites.” And so we implemented it. You know we had full structured data probably before almost anybody. And it moved the needle pretty well, but it didn’t move the needle as much as it does now. Google cares so much more about it now than they did then. When you’re a really early adopter, you don’t get the benefit as much as if others are adopting as well.
So we saw it as more fundamental than just the fact that it helped us grow. It was sort of in my head that this was the future. This is how Google wanted to communicate. There were so many reasons why it was better than them just crawling HTML — they could use it in interesting ways, they could give people answers right on search queries — so just the recognition that it was sort of the future of where Google was going I think was really how I got excited about it and ended a building a product around it
Blake: Now you’re doing your own thing at Huckabuy, but I want to talk about the foundations of that and where the idea came from. If you remember, where were you when you had the idea and started developing it?
Geoff: Yeah, so I don’t really remember the exact moment that I came up with the idea, and my original idea was a terrible idea, which I think is actually a great lesson for entrepreneurs. You do have to be open to the fact that your brilliant idea might not have been so brilliant.
But while pursuing that first idea, we had built some really cool sort of SEO automation technology, and people that knew about our company wanted to start licensing it, so I didn’t even really understand the value of a software company and the recurring revenue model and just how great that was. It was just the fastest path to revenue so we pivoted two and a half years ago, into this software business. And we were like alright let’s build a structured data product, there’s nothing out there that’s good, it’s hard to do internally, it’s getting more and more complex and Google’s asking for more and more requirements. This is something that lends itself to being outsourced, and we came up with a really slick solution to do that. So that was our very first product. We have two products now, but our first product was built to automate world-class structured data for any given website, to do it as quickly as possible, and to maintain that on an ongoing basis. And the results were just great.
Blake: Would you mind briefly just explaining some of the different ways that structured structured data can be used just in case there’s somebody in the audience here that has questions?
Geoff: Yeah, I can give you just a real high level overview on what it is. Essentially, it’s a language that allows a website to authoritatively talk to a search engine. So, for years and years — still today for like 90% of the internet — the way that a search bot understands a website, is through what is called HTML. HTML is unstructured, it’s very difficult, if you’ve seen it, it’s kind of, you know, just a massive blob. And so search engines essentially invented this language to allow a site to actually tell them what’s going on in any given page. You can have structured data around a person — it’ll explain that this is a person, this is where they work, this is where they went to college and how old they are. The most common thing structured data is used for on the internet is product structured data — so describing a product and its price and dimensions and all those things. But there’s structured data for almost anything — including events, movie times, medical structured data, legal structured data — almost anything that can be represented on a webpage you can communicate via structured data.
Structured data has two main benefits. First, it helps Google understand just what’s happening on any given site and the more they understand, the more search exposure they end up typically giving a website. The second benefit is that Google uses structured data in lots of interesting ways. For example, when you search for a movie time, a sports score or a recipe and it just shows up, that is powered by structured data. That’s all being powered by structured data.
And it’s becoming even more important because it’s actually what powers voice search. So you know when you search online and your desktop, you get 10 blue links. When you search via voice search, you just get one answer, and that answer is essentially just reading back to you what’s called the rich card — powered by structured data — at the top on any given page. And so, it’s become very important, it’s a really big part of their algorithm at this point, and almost anything can be marked up and communicated to a search engine.
Blake: Are there any new features of structured data or new capabilities that you would love to see added to the next Google update based on your experience?
Geoff: Yeah, they’re pretty open about it. I mean by ourselves we have actually submitted enhancements to structured data. One misnomer about structured data is that the open source movement is actually schema.org. Schema.org really was the starting point of structured data, but the structured data repository that Google actually uses today is called JSON LD, and it’s a technology that allows data to be associated with one another. It’s what you can submit to if you have new things that you want to mark up and it gets updated almost monthly at this point so there’s new structured data, new structured data object types, and requirements that change pretty regularly. So if you identify something that really would help structured data and help Google, usually it will come to fruition.
Blake: Why do you think that structured data doesn’t get enough attention compared to the value it provides?
Geoff: That’s a great question and I think about it a lot actually. So, I don’t think it just applies to structured data. I think it actually applies to SEO. So SEO kind of gets ignored by your general marketer, you think of your sort of general CMO — they kind of are fearful of SEO. If you think of all the other hats that they wear, none of them are technical. So all the other stuff falls under traditional marketing skills like budgeting, measuring ROI, looking at where you invest dollars, and then you come into SEO, and it’s like this totally different skill set. It’s very technical. And that’s scary to a lot of marketers.
Even when you start down the SEO conversation, there’s this sort of fear and nervousness because the skill set, I think it’s just so much different. What we really try to do is to eliminate those fears. We say, “don’t be fearful of the technical side, our solution helps you tremendously in the technical conversation between a site and Google and we’ll tell you how it works.” It’s not overly complicated and any marketer can learn to understand it. I think people get intimidated by it, and as a result they under-invest and under-research in these more technical areas.
Blake: UX and SEO are melting together more now than ever. How do you personally toe the line of “Yeah we’ve got to optimize our site for human beings still, but we also have to optimize your site for search engines.” How do you tow that line and keep it balanced?
Geoff: Yeah, Blake, you’re asking great questions, by the way. These are so good. So here’s my experience. I don’t understand the friction between UX and SEO and I think you’re totally correct that good SEO is also good UI/UX. Any website that I’ve seen that gets optimized for SEO, the conversion rates go up, because you’re organizing a site based on what people search for and you’re calling things what people call them, instead of what you call them internally. And then you’re creating clickable links and pages about those things.
Page speed is one of the best things you can do for conversion rates, and it’s also one of the best things you can do for SEO, because fast sites convert well and are sites Google loves. So, I think they’re the same thing. That’s a totally different take than you’ll probably hear from anybody else, but when it comes to UI/UX and SEO, I see that they help each other.
Blake: If you had to take your wildest guess as to what percentage of SaaS or B2B companies actually use this, use structured data in their strategy. What would your guess be? And as for e-commerce?
Geoff: For SaaS or B2B, I’d say like 3%. Most of them would be Huckabuy customers.
And as for like e-commerce, I’d say about 95%. So, it’s a little disparate — but it goes back to those margins. You can have a great sales team, and you can grow into a pretty big software company, without ever doing SEO. You can’t get anywhere in e-commerce without being an SEO ninja. So, software companies discover SEO later in their lifecycle than e-commerce sites do, and e-commerce sites are way more competitive, because their whole business — or a lot of it — is based on whether they are winning or losing search out there. When you talk to, like Backcountry.com — they’re a Park City based e-commerce site that’s gotten pretty big — when you talk to them about SEO, they are dialed. When you talk to Salesforce about SEO, like they often won’t even have someone working on it. I mean they do, but it’s different.
Another one that’s even worse is financial services — so your Fidelity and Bank of America type websites — they have nothing in terms of SEO. They’ve never heard of structured data, no one in their organization has heard of structured data — that’s probably an even bigger offender than the B2B world. They spend a lot on marketing, I mean it’s in the hundreds of millions, but they don’t do any SEO.
Blake: Yeah, those are some of the worst websites out there I think in terms of SEO. Do you think that’s because companies are afraid of the time it takes for SEO to really work?
Geoff: For sure. Yeah, it’s not that immediate sort of paid channel, you know, we call it the crack pipe, like immediate hit. Yep. Yeah, I think that’s a big part and then to earlier in our conversation, the marketers get scared of SEO because it’s so technical and now I hear all these rumors around that “SEO doesn’t matter anymore,” or that “Google’s too smart.” It’s just ridiculous what you hear. I think it’s a combination of things, but more traditional marketing — they’re slower moving industries, the margins are sort of big so they can get away with paid channels and the mass works by just doing paid channels. I think there’s a lot of different reasons for it, and I think the smart ones though will figure it out, that it’s there. The smart software companies are all over SEO, they get it, and you can only get so big without it, and then you really have to invest in SEO and so I do see the Salesforces and Concurs and SAPs of the world, all of which are customers, they get it, and they’re going big on SEO.
Blake: If Google was actually to have their ideal website, what would that look like? How can we make websites load that way? I would love for you to dive into what your vision was behind the SEO Cloud, and get a little more technical as well by talking about dynamic rendering and things like that.
We take a given site, we convert it into flat HTML versions, we convert all that dynamic content and we strip out everything that doesn’t matter. It ends up usually being about 30% the size of the previous site — which shows you how much code bloat is out there — it looks exactly the same, and then we post it in the cache somewhere. So we have a partnership with CloudFlare, who does edge delivery, and so the site is basically instantly available for Google to crawl. It’s much lighter weight, it contains all the content and information that they still care about, and it has structured data stored at the top, so it’s cached it’s very fast. We’re trying to basically build what we call Google’s perfect world, let’s just give them what they want. Google’s pretty open and honest about what they want on the site.
Blake: You and I both know that page speed has SEO factors behind it. You can rank for more keywords and get more organic traffic if you have a faster website. Google incentivizes it so much that they’ll actually prioritize your site over slower sites. So why do you think that people are so slow to really take action?
Geoff: I think the issue is that it’s often a technical problem and it’s a difficult technical problem. A marketer usually can’t influence page speed, unless they start ripping out tags and stuff. So whatever CMS you’re using, whether it’s WordPress or whatever, there’s often a lot of stuff built in there that’s slow. And if you think about how a site loads today, what happens is, if I call a site from Australia and the servers are in New York, I make that trip to New York (like the internet trip), I call the site, the site typically will have something dynamic — whether it’s personalization recommendations or a chat box or tracking pixels — and those all have to get called from their different locations around the world, and it all comes back together and makes that lap back to Australia. So there’s just a lot of stuff going on. The business requirements that are put on a site to have a chat box, analytics, dynamic content, and personalization cause a lot of difficulty when it comes to page speed. And for any organization, solving that problem internally, is a very big technical problem to solve. And I very rarely see it even being addressed quite honestly, because it is so technical.
Think about big sites that load slowly. These companies are doing all these content marketing efforts that probably aren’t even getting indexed. So we look at a lot of crawl stats — with SEO Cloud you can actually watch them crawl and see what they’re indexing. A lot of companies that have indexation problems — meaning Google can’t get through the entire site — have no idea that it’s even happening. And so what we’re trying to do is basically make sure that Google could come and get everything in like 30 minutes and be done with hundreds of thousands if not millions of pages.
Blake: Overall the key takeaway for me is, if you’re not a technical SEO, or if you’re not a technical marketer, try to take the steps to get to a point where you’re comfortable with it because it’s really overlooked, and especially in the B2B side of things right now, there’s a huge opportunity to stand out because so few people are actually taking advantage of it. And then on the flip-side, if you are a more technical person, try to build relationships between development departments and marketing departments. That can be something that really just benefits everybody involved.
For the in-house marketer that’s listening to this podcast, what’s the number one skill that most of them don’t have, but really should?
Geoff: The quickest win is just to get good structured data live throughout the site. One thing in SEO that is very often overlooked is the navigation and authorization of their navigation. It’s probably the most important, it’s the very first handshake that you have with Google for them to try to understand what your site’s about. And usually it’s not organized in a way that makes any sense at all to them. So those two things are important.
Also, as you said, getting a developer excited about SEO is one of the biggest things that a marketing department can hack into, because they can actually make the changes that will lead to growth. They can make your content more effective. We had a bunch of really good developers working on SEO that were passionate about SEO and they cared about the numbers, too. It’s pretty rare for a developer to work on something that they can actually see really impact the business. And so I think getting them sort of excited and involved into the numbers and watching things grow is a great step in the right direction. And of course, if you don’t want to do it, we’re happy to do it for you here at Huckabuy.
Blake: Great, well yeah, give us your pitch. I’d love to give you a chance to talk about Huckabuy, what you can offer, and why people should check you out.
Geoff: Thanks, Blake. We are obviously a bit different than your typical SEO agency. We are not an agency actually, we are a technical software solution — which I think is really refreshing to a lot of marketers. Think of us as the company that really handles the technical conversation between your site and Google. So, the structured data will be world-class, it’ll be live all the time, as your site changes it gets updated and you don’t have to think about it. Then with the SEO Cloud product, if you have kind of a more complex website it can really be a game-changer. Our average customer — and these are some big, established customers — grows about 62%. So you’ll see some pretty significant numbers.
Advice from Geoff Atkinson: Make it so obvious that you’re driving growth and driving the company values that it’s just a no-brainer for you to move up. Find an opportunity at a company — typically that’s a little bit larger because you’ll learn a lot more — that has a career track. Also, you should have a really key mentor in place. Having someone that really cares about you as a mentor that wants to see you succeed is probably the biggest break you can get in your career. It really doesn’t matter how much money you make until you’re like 40, so don’t worry about the money as much as what experience you are going to get and what are you worth, in the open-market by the time you hit 40. That’s sort of how I coach young people. Think about the experience, find yourself a good mentor, and then, do a really good job that moves the needle.
The issue is often a technical problem and it’s a difficult technical problem. The business requirements that are put on a site to have a chat box, analytics, dynamic content, and personalization, cause a lot of difficulty when it comes to PageSpeed. And for any organization, solving that problem internally is difficult. A lot of companies that have indexation problems — meaning Google can’t get through the entire site — have no idea that it’s even happening. One solution is to outsource technical SEO or you could hire a developer with technical SEO expertise.