In this Episode
- [01:06] – Why did Marcus decide to switch from affiliate marketing to running a software service company?
- [05:18] – Stephan steps in for a moment to share an origin story of his own.
- [06:55] – Marcus shares his thoughts on what Stephan did, which he believes was a smart strategy. He and Stephan then go on to discuss their companies and work experiences.
- [11:27] – Stephan brings up the idea of surrounding yourself with people who are smarter than you. Marcus agrees and gives an example of having done this.
- [14:17] – We learn about how private SEOktoberfest is, and how it’s structured in terms of experts and attendees.
- [16:52] – Marcus and Stephan engage in a role-playing exercise by giving each other cool information, the way they might at a mastermind or think tank.
- [18:02] – Stephan’s first contribution involves YouTube searching on Google Trends. His second is about Christoph Cemper’s research on 302s being better at passing the SEO benefit over time (versus 301s).
- [20:36] – Stephan shares his craziest conference story
- [25:01] – We hear about Marcus’ and Stephan’s thoughts on the pill-pushing game.
- [27:22] – Certain black hat techniques stopped working in 2007 or 2008, Marcus explains.
- [32:49] – Stephan and Marcus talk about the featured snippet, or instant answer, on Google.
- [34:23] – We hear more about SEMrush, and a tool it offers related to featured snippets.
- [35:37] – Marcus takes his turn for sharing ideas. He talks about TF-IDF analysis, a major topic in German SEO circles.
- [38:41] – Marcus offers a reverse example of what he’s been talking about.
- [45:11] – We hear an example about the kind of thought process Marcus has been describing, from Stephan this time. He talks about Homesteading.com and the fact that one of their articles outranks the home page on Google.
- [48-31] – What does Marcus think of latent semantic indexing (LSI)?
- [52:42] – Are there any free tools that give some actionable insight into TF-IDF? Marcus reveals that a limited version of this tool is available in the free version of OnPage.
- [54:31] – Stephan and Marcus touch on the problem of using the disallow directive instead of no-indexing pages.
- [55:22] – Marcus talks about some common SEO screw-ups that OnPage can find.
- [60:14] – Marcus’ company doesn’t offer any consulting services. Here, he explains why.
Hello and welcome to Marketing Speak! I’m your host, Stephan Spencer. Today, I have the distinct pleasure of inviting Marcus Tandler onto the show. Marcus co-founded and runs the enterprise SEO software company, OnPage.org which is ranked in the Technology Fast 500 in Europe. He’s a former super affiliate, in fact, he was in Commission Junction’s Top 5 earners in all of Europe. He hosts a super exclusive conference called SEO Oktoberfest. It’s so exclusive I can’t even get in. Marcus, it’s great to have you on the show.
Why switch from affiliate marketing when you were so successful as a super affiliate and doing arbitrage and all that sort of stuff to running a software as a service type company?
That’s a good question. I got to say, affiliate marketing just changed a lot. Back in the days when I was running very good in affiliate marketing, that’s a long time ago and it was a whole different playing field, especially in Germany where I would just go to American conferences, I would fly into US conferences all the time, San Diego, New York, just pick up marketing tactics that were almost old for US peeps and just imported that into Germany. These days, the playing field is just completely different. I think affiliate marketing is totally dying. It’s just not really a sustainable business model for me anymore. But, with OnPage.org, it was not that we really wanted to, hey, let’s do a software company, it was actually we didn’t want to be a software company for the longest time. OnPage.org was a tool we have coded for ourselves, we were working for a couple of mostly larger clients the Deutsche Bank, the Deutsche Telekom, the Disneyland of sorts, mostly very big sites. Working with these big sites, we just had very unique problems that no other software could solve for us at that time. I’m a big field fan, I’m a big Screaming Frog fan but, especially with the frog, everything beyond a million URLs gets quite hard with Screaming Frog. Of course now, we can write it in a cloud in multiple instances and stuff like this. It’s getting better. Still, I didn’t like the feel of the proprietary software. Also, what I didn’t like was always basically writing a report, exporting it to Excel, writing another report, exporting it to Excel. I always ended up working in Excel, I just hate working in Excel. We want a better software solution and this was what we did internally. Of course, the software was nowhere to be in the front that you can actually sell it to someone. It was running on a shell, right? It’s very rudimentary. We had a lot of people, a lot of friends that just hey, you should release this software, other people would like to work with it as well and we are always like, no, not going to do it because we were quite happy being just four people running a boutique shop and being able to fly around the world half of the year. Doing a lot of mileage in the air, this is just nothing you can do when you have what I have now. 50 people sitting in the office, I cannot just say, okay, I’m here today and tomorrow I’m there, right? You have to show them that you have to be in the office. And also, we knew we would need to get support guys, we need sales guys. I suck at sales. We would need bigger IT and all that stuff. It wasn’t really our first choice to become a software company but then we did a beta version and said, okay here’s 100 accounts, €89 a month each. Let’s see how that scales, let’s see if people can work with this software, let’s see if people understand how to work with this software, how to really use it effectively. We were thinking we would sell like 10, 20 accounts to close friends and then it’s a great software so maybe like 5 more people, 10 more people. It’s B2B software, right? It has the sexyness of an Excel spreadsheet if you’re not into this game. It turned out completely different. We actually sold 300 accounts within 24 hours and we saw ourselves being profitable just one week after we started and we all looked at each other and we just saw that we got to do it now. We burned everything else down, okay, now we got to focus 110% on this and never look back. Now, we are 48 people, we have a huge office here in Munich, a very nice office, very Google-y style and we have over 300,000 users from around the world using the software. We have over 2,000 paying customers, huge enterprises that are using our software. It’s been a crazy ride. And again, just bootstrap, no external capital, just started with $6,000 start capital and never looked back, profitable ever since, just growing out of cash flow for the past 4 1/2 years.
That’s amazing. It’s a similar origin story or there is some similarities in your origin story to mine. I had this gravity stream SEO technology platform that I’d mentioned to you in the past. This was something that was born out of working with clients. I remember back in the day working with calls and just being frustrated. They wouldn’t trust at anything that SEO would be good for their brand so I try to show it to them without having any access to their servers. I would use a proxy server to inject SEO changes into a proxy version on their site just to show them that we weren’t going to ruin their brand with SEO. And then I realized, we could use this in a production environment. We created something that was just very rough, I’d written it in parole and PHP and some rewrite rules and it worked. We were able to use it on production environments and we are able to quickly ramp it up and make it a very profitable side of our business. It was just born out of doing stuff for our own purposes at the beginning. I just tried to solve a client problem and yeah, ended up being very successful at charging on a cost per click basis for SEO and that’s mainly the reason why we were able to sell the company. It’s pretty impressive that you were able to be profitable within a week. How long have you run OnPage.org?
We started in 2012. Ever since. But I think it’s a very smart move what you did with the proxy server, especially working with large clients. It’s always the problem with, hey you got to fix this. Even if it’s just a very small change. For example, fixing something in the robot’s text right? It will take them a month to change one single thing and this is very frustrating if you are an SEO consultant and you tell them all this stuff and nothing gets fixed, right? If you don’t do something, there’s not going to be anything good to come out of it. Very smart move to just, yeah, we just got to do it ourselves, you don’t have to do anything and we got to charge it in the click to click basis. Holy shit! That’s awesome! Fantastic. Really smart move.
Yeah. It was really lucrative. Most consulting companies, agencies, we’re able to charge maybe $15,000 a month for monthly retainer, we were able to get seven figures a year from some clients.
Crazy. You have pay for performance, right?
There’s all pay for performance.
That’s it, right? This is the best deal. I’m going to give you 10 figures if you can make me more money with it, right? Because on a paper forms deal. That’s the best kind of deal for both sides, that’s a definite win-win. Much better than having a retainer and not really knowing what would come out of it, especially if I as the client have to do some stuff and have to have IT resources, have to have people that actually do the stuff that you tell me. I think on a pay for performance level, this is the fairest deal for everyone.
It works up to a point. At $.15 a click which is what we’re charging. It’s scaled to the point where you could get up to seven figures a year in revenue from one client and then a client would be like wow, this is a lot of money. I really should build something myself and we’ve actually had clients that built their own proxy based platforms to replace us. At seven figures, that makes the CFO pay attention.
Yeah, definitely. They probably pay even more per click on adverts, right? At $.15 a click, if you’re in a very high margin space that’s not a lot, right? They end up paying $.40 per click in their normal adverts account. It’s a much better deal. Although I totally get the point if you spend over a million, you might as well just put the money to do it yourself. Although it’s quite weird that they would do a proxy.
I know. It is crazy.
It was a terrible version, they did a terrible job building, whatever. I’m no longer involved with the technology because that went with the acquisition. I sold my company, the tech went with it. That’s fine. It just struck me as a similar origin story that you’re just out there having a successful client focused business, you created a tool to solve a problem, it was really rough just for your own purposes and then you realized that hey this is actually a viable business in its own right.
This is always the best business to start. We are very much in the startup scene here in Germany as well and I have a lot of people, I want to do a startup, I want to build a startup but I don’t have the idea yet. They’re scanning ideas, what’s going on well in America and what could I bring over to Germany? I couldn’t do it, I know I couldn’t do stuff like this. Because for me, SEO and then running OnPage.org, it’s like a hobby. It still feels like a hobby because this is what I love and this is what I’ve done for the past 20 years. I still love this shit, I love this game. Just doing something because it’s successful in other country and just bringing it over, I don’t know. You really got to love being a manager or being masochistic, that’s the right English word because as you know, as a business owner, you got to do everything, right?
I always wanted to build a smart team around me so that I didn’t have to do everything. Even though we bootstrapped too, we took a very small amount of money, like $250,000 maybe 10 years into the business, quite a long way down the track. We were bootstrapped the whole time, pretty much, and I wanted to surround myself with people who are smarter than me. I moved to New Zealand and started an office there, we were a block from the beach, it was amazing. We are able to hire a whole team of 35 people within a pretty short period of time. I started with the general manager. Somebody who was amazing at managing and the stuff that I didn’t enjoy and the stuff that I wasn’t as good at. I’m sure your 48 staff, you have some rock stars that are way better than you at different aspects of the business and that’s just the nature of running a good business.
Absolutely. Same strategy, early on. We really pinpoint on some key people, we really want it. For example, the guy running the business operating process right now, he used to run online marketing for match.com. He’s very well versed in the business, running on large scale, large custom acquisition type of stuff. Really having people that have done this stuff before, that are very good at the stuff, where you could learn stuff. I think this is exactly what you need to do, you have to surround yourself with people that are smarter than you and it will just make a much better business, definitely.
Yeah. You’ve been in the SEO world from the very beginning. You were optimizing for AltaVista even before Google existed.
We also have Fireball in Germany. Fireball was a very big German search engine, and other stuff of course.
Let’s talk about your evolution. Now you’re in the corporate world, you’ve got thousands of customers and some of them are very large enterprise companies but way back in the early days of Google, you were doing Black Hat SEO, I’m guessing right?
Of course. We didn’t even have a dedicated team. Google never had dedicated team for Germany. Black Hat was the way to go because everybody else would be doing Black Hat. Why would you use White Hat techniques where they just don’t get you anywhere because everything’s Black Hat. Here in Germany, for the longest time, very early on, it was only Black Hat, of course.
Let’s talk about that. The Black Hat world, because it still exists and it’s still working for some people. You run the SEO Oktoberfest, what did you call it? A Mastermind or a conference?
Think tank. SEO think tanks.
Okay. Think tank, I’m sure you have a lot of Black Hat discussions in there. It’s all super secret, everybody’s under NDA and can’t share the stuff outside the event, correct?
Yeah. We’re very proud that actually nothing, no content ever leaked from SEO Oktoberfest. But the way that the concept goes is as follows, it’s every expert, half of the people are experts, half of the people are attendees, although we have so many so many attendees who are regular speakers or even keynote speakers at different events. For example Christoph Cemper, you know very well, I guess. He’s an attendee at SEO Oktoberfest but not because he couldn’t be an expert, it’s just also a lot of people enjoy not having to present something. As an expert, what you need to do, you have 15 minutes of fame, as we call it. In this 15 minutes of fame, you just tell something or you show something that you think most of the people in the room have not seen or heard before. This is how the game goes, and then people vote on you and the more votes you get, the more likely it’s going to be that you’re going to be invited for the next one. This is how it’s evolved over the years. This is really the thing that you ideally hear at least 28 people talking about stuff that you just didn’t think of yet, and this might also be stuff that you’re just not into. For example, Robert Hansen, big heck of a guy who runs Black Hat in Vegas, he’s one of the most famous hackers who ran hacker.org back in the days. Of course, most of the stuff he talks about, I understand only 3% of the stuff he’s talking about but 97% is just plain cruel. That’s the stuff, right? It’s more like a little bit of mind opening. Hear how other people, what are their problems, how they solve different problems, what they are working on and stuff like this. It’s not always that it might be this is exactly the tactic that I’m looking for to do something actionable but more like getting to the mindset of different people that are very successful in this business and just learn from everyone.
Awesome. That sounds similar to, I was just at the War Room Mastermind. As you know, we were just talking about it and the whole segment of the event, this is a high level mastermind, you pay $25,000 a year to be in it, they meet four times a year but they have this part of it dedicated to the attendees coming up on stage and sharing their best stuff in just a few minutes, you could do a short one for two minutes or you could do a longer one for six to eight minutes and they call it Wicked Smart. It sounds very similar. Let’s role play this because it sounds really fun and I’m sure our listeners would enjoy it. Let’s do an SEO Oktoberfest Wicked Smart kind of thing right now. How about I’ll start, I’ll give you something really cool and then you could follow suit and we’ll just keep going for as long as we want. This is stuff that you don’t feel uncomfortable sharing in a more public environment because this is going to go to who knows where on the internet tube. Don’t share anything super secret that you don’t want getting out. Sounds like a plan?
It’s hard. Especially if everybody can hear it. There’s cool stuff going on but as soon as everybody knows, it’s not going to be very long, how long it’s going to continue to work.
How about we do this? How about we focus on stuff that is already out there but just not well known? So you don’t feel at all reticent to share it because it’s cool stuff. I’ll give you a few examples. I’ll just think as we’re going here and we’ll see where we go and then you can chime in if you want. Here’s one that’s surprisingly not very well known. It’s super simple, it’s part of the free Google tool called Google Trends, you go to where it says web search and you choose YouTube search instead and then you can do a keyword research based on people searching on YouTube instead of on Google. That’s I think pretty darn cool and everybody should know this. Everybody should know this. A lot of people know about Google Trends but very few of them know that you can do YouTube keyword research with Google Trends. So, there is one. Let’s see, Christoph Cemper who we just talked about a minute ago came out with a really interesting research about 302s being better, passing the SEO benefit overtime versus a 301. He released that at SMX East and actually had a client test that. One of the sites that they acquired, they switched from 301 to 302 and they got a nice lift. The idea of having the rankings benefit transfer to the destination URL for a long period of time seems like the 301 is only a certain number of months, just a few months and then the rankings benefit trails off or drops off. It’s like the context of the link doesn’t continue to transfer, the pagerank does but not the anchor text and context. That worked. It’s not just something that is a theory put out there, it works in the real world. That’s pretty cool. A lot of folks don’t know about that either.
What’s on the anchor text also needs to be on the target page. I think this is coming back from the old days, doing a lot of 301s with expired domains. Worked like a charm, right? You just pick up a couple of linked expired domains to shoot them on a target and you can even have those 301 log where you can just point to another domain, put up until it gets burned and then you just shoot it until the next domain. I think Greg Boser is very well versed into that. I even believe that the internal Google tool that finds the flight, this is named after Greg Boser. I’ve heard this from several sources, trusted sources, so it could be true.
Before we go on, I got to share an amazing hilarious story about Greg Boser and Todd Friesen. They would speak at conferences, they do sessions together a lot.
With SEO Rockstars.
Right. So they had their own show together too. They were doing an SEO workshop at Web 2.0 Expo and I was speaking later in the week during the main conference. This pre-conference workshop, I was just curious. I knew the guys and I just wanted to see what they were going to talk about. I go into the session and I start hearing stuff and seeing stuff that’s kind of familiar to me. And then there’s a bunch of tools on one of the slides and includes a very esoteric tool that very few people knew about that I developed. The Netconcepts’ Link Checker, right? It was one of the tools on the slide so I raised my hand and Boser and Friesen, they get all weird. They were like, “Okay, first of all, this guy is an SEO and he’s just trying to sell you stuff just like we are.” “That said, go ahead, Stephan.” And I’m like: “Oh, Todd, I just want to say thanks for the mention for Netconcepts up there, that was my agency up on the screen.” And then they act all weird again and they were like, “Okay, we got a confession to make, these are Stephan’s slides. It was late, we were out partying and stuff and we didn’t have a slide deck yet and who has great slides but Stephan Spencer. We just lifted one of Stephan’s decks and I’m glad we put the time in to tweak it a bit so that it wasn’t just a complete copy and paste.” They admitted to the whole room, 400 and some people, that they had lifted my deck. Now, at that point, I stopped sharing my decks. I was very generous. I had dozens and dozens of PowerPoint decks that I’d shared online and I stopped doing that from that point forward.
And the funny thing about it is they were proud of it. They thought that it was hilarious. At the next show, it was SMX Advanced and we were all in the Give It Up panel which was my idea, like you would bring your best secrets and tips and stuff and share it. I wanted that stuff to stay secret in the room. Danny changed it so that there is a month embargo and then you could release it to the world so that kind of sucked. I brought amazing ninja stuff, I brought this thing of how to figure out back when Google did host crowding, they grouped results from the same sites together and I wanted to figure out what the intended result true position was. I would do num=nine, num=eight, num=seven, etc. in the Google search URL to see when that second result, that intended result dropped away and then I knew the true position and I could go after indented listings that were weak, that were near number 10 by promoting some stuff that was on the top of page two. But only if it was weak, if it was number 10 or number 9 and I could just go after those indented listing. That was the sort of stuff that that I was bringing and one of the things that I brought to that session was how to find force to research reports that are normally thousands of dollars just with a Google search. And I showed it, I’m like you do file: PDF and endnotes great find force to research etc. and you’ll find multi thousand dollar force to research reports for free just by doing a Google search. And then Friesen, he’s so funny. He’s like being a smart ass and says: “you can find great PowerPoints that way too!” I’m like: “Dude, now you just opened yourself up, now you got to tell the whole audience what you mean by that” and he did! He was so proud to tell everybody what happened at Web 2.0 Expo, hilarious! I couldn’t resist but to share that one. But they’re cool guys, I really enjoy whenever I see them at conferences and chatting with them. They are always on the leading edge, testing out all sorts of crazy stuff. Now Todd is heading up SEO at Sales Force. No more Black Hat stuff.
Yup, interesting days. I never got into that space but I have a friend who ranked in the first half of page one for Viagra and made millions of dollars just off that one keyword.
Very nice. I was actually not in the game as well but it was actually during one of our podcasts. I think that was called Weapons On The Roof on Webmaster Radio. Doing one of those shows, I ran a search for Viagra and on the tab one result was a parasite hosting thing. Some parasite host who got just pawned by a Viagra spammer, I don’t know whatever subdomain or directory it was back then, something like tripod or blogger.com or wordpress.com or whatever. It was ranking number one but it was always suspended, so you click on it and it was already suspended. I saw this right on the show. I’m like, holy shit! The only thing you need to do is go to that host, register for free, something that Viagra in the URL’s log. Viagra for free or Viagra 24/7, and then just point a lot of links in that and back in the game you could still use [00:30:53] network to get a lot of RU links or whatever and point a lot of links in there because you can break this personal host, right? Because it’s a very trusted host so you can’t break it by shooting ultimate links at it and Google will replace the result because it has already been suspended and then Google will think hey, is this something new with Viagra, so let’s have this one right there, like they do it normally as you can see in Google switch console, just breaking multiple results from a domain to see which is the best results for the user. I actually said that right on the show. And actually, a friend of mine picked it up right there and two days later, it was ranking number one for Viagra and selling some stuff with it. Only for a month though, I think it was like three weeks to four weeks or whatever. But it was like, holy shit, this is so easy and nobody picked it up but him and he ended up ranking for a couple of weeks for Viagra number one. That’s crazy. Most people just sell fake pills and I always stayed away from selling fake stuff or trying to scam people because I think there’s a fine line between Black Hat and just being an asshole. I don’t want to scam people. Black Hat back in the days was Black Hat. I started doing Black Hat meaning white font on a white background, just keyword spam like what you needed to do [00:32:16] and days. And then now these days, it’s all about the hacking stuff and injecting XSS exploit kind of stuff. My stuff gets hacked too, all the time too. You just don’t update a WordPress log once and you just get pawned by an automatic exploit. It’s just horrible. It’s very annoying. There’s no real Black Hat these days anymore which is like [00:32:49], it’s mostly Gray Hat techniques and that’s definitely not my ally.
That sort of technique of driving low quality links to a high authority site that you could host on, so parasite host. That works until when? When did that stop working?
I’d say 2007, 2008, maybe, just a good guess. This is when it worked really well. When you have the wordpress.com, blogger.com, the big URLs, Myspace even worked well. Blah, so much stuff. All these free blog hosts, like the big blogging craze where you had so many blog hosts where you can just open up a blog, maybe even automatic like a WordPress model user on platform and just upload keyword databases and end up with thousands of blogs and just pull links to it and they will rank immediately because it’s such a strong host. Google was not very good at finding that stuff back then but all in all it was a tough challenge because it was just everybody was doing it and such a massive scale, it’s just a lot of stuff to clean up.
I wasn’t doing it. There was at least one person not doing it. One thing that I think worked for even longer, at least from what I heard was, if you want to get your YouTube videos to rank, you could drive low quality links to your YouTube videos because of the strength of the domain of youtube.com. Google owned it and everything. You could really boost your rankings, your Google search rankings for YouTube videos by dropping low quality links to your YouTube videos. Did you ever do that?
Nope. I actually have not been on that YouTube game, I have a couple of friends who have been doing this excessively. When Google started doing the universal search stuff and really put out a lot of YouTube videos, especially YouTube videos, right in the serps, they had four poker bonus codes and stuff like this so they would rank for all sorts of stuff and then they just upload this video where it shows a code and they were killing it. It was really crazy but no, never been into that game. Back in the days when Google started showing frugal, product search. We convinced Google to rank products which weren’t products like payday loans. It’s a German, it’s not English stuff, it’s German stuff, but like long search queries. For example, loans for self-employed people or something like that. You would convince Google that this is actually a product which costs $.1. What ends up happening is that you just go beyond, immediately above the number one results because Google wanted to promote frugal at that time and it was just phenomenal. It was working so well. Actually told the story, I was invited to speak at Google on that story, it was huge. But again, this is like 10 years ago.
You get invited to speak at Google and you told them how you spam the heck out of their search engine?
It wasn’t working at that time anymore. It was just showing how to stay on your toes when you’re in SEO. You have to see opportunity, you have to see something and we were seeing all these frugal results popping up in the search and especially above number one, you had very competitive commercial queries and all of a sudden you would be before a number one result. Holy crap. That’s awesome, right? We were trying to, okay, how do we do this? And how can you convince Google to also rank “other products,” which are just not products but you just pretend that they are. And this is basically very easy like you have to have a couple of merchants that actually sell this product. They have to have different pricing, so there’s definitely price comparison component to it and stuff like this, it’s very easy to gain. Didn’t work for too long, like all the good stuff, but it was crazy because you can really, next day, rank above number one for better competitive terms. Holy shit, it’s awesome.
And now these days, you got featured snippets so you can out rank the number one organic listing again. There’s a whole set of strategies and tactics for that too which we can get onto in a minute.
This gets heavily abused. I get results, everybody has five stars, right? It’s not going to be something that will help you in regards to improving click through rate anymore. But if you’re the only one that has five stars, and then have a future at snippet, holy shit, awesome!
But what about position zero? The featured snippets? Instant answer essentially to the problems, that gets a much higher click through rate.
We have a couple of those with OnPage as well, for SEO related terms, what is SEO. In Germany, we have amazing results with those. It’s just really crazy. You might think that, oh my god, everything is already there, why would somebody click on it? But click through rates are just going through the roof. It’s really, really cool. Something really worthwhile to spend your time trying to get these featured Snippets.
Does OnPage show featured Snippets that you’ve acquired? Does it allow you to see anything about your competition, if they have featured snippets?
No. OnPage, the focus is really just on your website. We put your website in the focus. There’s no research component to it. As I said before, we did it because there was no other tool that could do what we wanted. I’m using Search Metrics each and every day. They are amazing for research, the research component for backlink research, competitor research, keyword research. We left this completely out of OnPage.org and even with the rank tracking, we focus on Google search console. Google search console data is the best possible data you can have. It’s much better than scrape data. Since we put your domain on the focus, we only help out about your domain, I just want to use Google search console data because it’s the best possible data you can get.
I use Semrush as well.
Yeah. One of my favorite features in there is related to featured snippets. That is you go and list your competitor’s keyword lists, which keywords they’re getting rankings for and estimated traffic volumes for those keywords, and then you can filter down to just those keywords that your competitor has featured snippets for. And then you go through that list looking for featured snippets that are particularly weak, like paragraph snippets are particularly weak because those aren’t usually the best answer to a query, to a question. If you are asking how do you boil an egg, that really should be an ordered list, a numbered list. If you find your competitors have some keywords that they have attained featured snippet position, zero status for and those are paragraph snippets that they have, that you can swoop in. If you have good click through rate for your position in the search results and you’re somewhere reasonably ranked on page one for that keyword, you could overtake the featured snippet. That’s a nice low hanging wicked smart idea there. Let’s turn it over to you for some wicked smart ideas.If you have good click through rate for your position in the search results and you’re somewhere reasonably ranked on page one for that keyword, you could overtake the featured snippet. That’s a nice low hanging wicked smart idea there. Click To Tweet
I got something. This is very big in Germany. It’s actually about TF-IDF analysis. TF-IDF analysis has been the most dominating topic in the German SEO scenes for years now. Interestingly enough, it’s not being used in English speaking countries. I see that time and time again, there were articles, for example, Dr. Pete wrote something on Mars but it was a very complicated approach doing this with Excel again. In Germany, we have six, seven tools alone that focus on TF-IDF analysis. We actually happen to be the first tool for TF-IDF analysis. That’s pretty cool. Actually, Germany has now six or seven tools that do what we’re doing. It works, it works incredibly well. It’s just not used all too much in US or even in the UK as well. I think it tunes in quite well with the German engineering kind of approach. We’re not very creative, Germans as a whole. For the Germans, it’s very nice to have an algorithmical curve that you can write your content for or with. It’s like having a guidance of what’s good content, but people always miss that you can still write shitty content with good data. When you get to the US, they’re just creative. They just think okay, what does the user want to hear? What does he want to read? What’s the best possible answer you can give them? And this is what you write. It has a more of medical approach to writing content but it works, it works amazingly well because especially in long tail environments, with long tail queries, Google favors holistic content because if I don’t know what you’re really looking for, what’s the very heart of what you’re looking for and I haven’t seen the search request a lot of times before. I don’t really have experience data from usage using data and all that stuff. And then they just favor more holistic content because they just have a bigger, better chance that you actually find something you were looking for within this content. I’ve also seen stuff going from page three to page one with no new links and I think just because you offer the most holistic content, and also that you do not rank for certain stuff, especially local results when there is certain key points missing. For example, this can also be found out with the TF-IDF analysis. This is what amazes me, really that in UK and US, not a lot of people use this yet although it’s a very powerful weapon to just write better content, or just at least have keyword inspiration, so to speak. Find more stuff that might be interesting to include in that content.
Right. So you find related keywords that are not just direct synonyms because of this analysis and then by incorporating that into the copy your pages have more holistic answer to the user search query and then you’re going to rank better.
I can give you an example. Reverse example in this case. I had a big agency, most of the big SEO agencies are using our product. A very big design agency from Munich showed up in our office and we showed them this software. He’s like, “Oh, Mr. Tandler, now that I have you here, I have a question that’s burning on my mind.” This is why when they were searching for web design agency in Munich, it’s actually just a two-word combination. They were ranking only with their contact page, right? It was the contact page and not their root domain. What you would think, you would think that the root domain would be ranking and all the other rankings were the root domain of web design companies,. We were showing them also our TF-IDF tool and then we ran the report for web agency in Munich. We actually found out that there’s a high correlation for 089 which is in fact the telephone area code for Munich. And interestingly enough, their root domain didn’t have an instance of 089 on the page, it was only mentioned on the contact page. Only there they had the telephone number. This makes a lot of sense if you think about it that we have a lot designing agencies, a lot of SEO agencies who just have a myriad of landing pages. They are an SEO agency in Hamburg but they would also drive to Bremen or Lubeck or Munich to acquire those clients. They have those landing pages, SEO agency Munich, SEO agency Bremen, SEO agency Hamburg, SEO agency whatever. Google doesn’t want to rank that stuff. If somebody is searching for an SEO agency Munich, you don’t want an SEO agency from Hamburg only because they have created SEO landing pages for that combination. They are looking if there’s an instance of 089 on that page, meaning that if you have the area code, if your office has the area code 089, you are actually an agency in Munich. This is what actually happened. I’m not saying you have to, have to do this, but when Google had to decide which page they’re going to rank, they actually went for the contact page because they’re never sure this is about Munich and they weren’t sure about the rest so this is why they were ranking for the contact page and not the domain. They changed it, put it on the root domain, that’s when root domain start ranking. Could be a huge coincidence but I’ve seen these cases over and over again, I actually ran a lot of reports with local search queries like this and I actually have somebody who’s ranking number two for SEO agency, in this case it was Bremen and he redesigned his home page and he didn’t have an instance of the telephone number anymore, it was like a visibility hidden and then you hover over and then you see the telephone number. It’s very agency page, the redesign that they did, and they couldn’t see that telephone number anymore and they actually dropped out of the top 10, dropped to page 2. Again, could be a huge coincidence, that stuff happening, but nothing else happened. It was just that kind of stuff and all the other pages actually had their area code on the page. When you told me that problem, I looked at back links, I looked at social signals that the contact page has. Significantly more links or social signals than the root domain, whatever reason Google could think that this page is more relevant for the search in that instance, but there was just no indication and it was really just TF-IDF analysis, where it was like eye-opening, like, oh, that makes a lot of sense. I have hundreds of cases like this where you can really see a correlation by leveraging TF-IDF analysis. Change the stuff, rank on the first page, it works. Again, lots of examples. As I say, it’s a great weapon in your arsenal for keyword inspiration, for analyzing serps for competitive analysis. You could phone 100 SEOs from Germany and 99 would know about TF-IDF analysis, huge topic here but not in the US. Actually, [00:50:15] New York has spoke about it. They we like, “Hm, what are you saying? Why don’t I just write what people want to hear, what they want to read?” Yeah, of course. Perfect. That’s how you should approach this game but it’s also great to do that kind of stuff. Actually, TF-IDF is like one of the oldest concepts, it’s been the third ranking factor with the Yandex, it has been one of the earliest ranking factors with Google and of course, not for ranking domains, but for example, for identifying homonyms, right? Word has multiple meanings and depending on other words being used on that context, you can find out what that word means. Like the good old example of Java being used in the context of beaches and Bali, the island of Java. And then with C++ programming, it’s the language, the program language Java. This is reverse why it is so important for Google to do this kind of stuff. This is what you do with TF-IDF analysis. It’s not voodoo but good old reverse engineering how Google looks at certain keywords or keyword sets and keyword in dds and stuff like that.
You got to start with the problem. Let’s say with that design firm, they were not ranking anymore for that particular keyword. They were targeting web design firms Munich, whatever it was. And then you try and reverse engineer what’s going on, or reverse engineer the Google algorithm, try and diagnose the situation and it could have been something else. Turned out, it was that this related keyword was no longer present on the page.
It could be. It could be a huge coincidence as well. I’m just saying. It’s a pain point. For them, it was very frustrating because the contact page was very stripped out, it was just a contact page, it was a very bad result if you looked for an agency and you just don’t see their portfolio and the agency and stuff like this. The just saw the contact page. It really bugged them and they were like, why doesn’t the root domain rank? Why? I just offered one explanation why this could be possible. This is what SEOs write all about. This is what I love about this game, the detective part, really trying why is this guy ranking there? What does he do? How can I replicate this? How can I use this? I love this stuff.
Here’s an example, having that kind of thought process. I was at War Room, which is run by Digital Marketer. Digital Marketer owns a bunch of ecommerce websites, like homesteading.com and Survival Life and a whole bunch of really interesting ecommerce sites. I was looking at the search results for homesteading as a keyword, one of their brands is homesteading. homesteading.com is their domain.
I’m going to look up real quick what steading means. I’m lost with translation here.
Homesteading is making your own laundry detergent and being self sufficient, kind of living off the grid, growing your own food, preserving the food and peanuts and jars and that sort of stuff and being able to live off the grid, make your own electricity, and what have you.
Not a bad thing in the age of Trump, right? Be prepared.
Oh please, let’s not go there.
I know, I know. It’s just very sad to see this in Germany, having so many friends in the US.
I can’t even go on Facebook anymore. Oh god, I just need to take a multi month break from Facebook because it was just so depressing seeing all the devisiveness and stuff on that. Yeah, it’s not good.
I’m sorry for the remark.
That’s okay. But with this homesteading keyword, I was trying to figure to why is it that an article on homesteading.com is the thing that’s ranking versus the home page. Because the home page I would think, being that that’s the domain, that’s the name of their site, I would think that maybe they’re getting penalized for the home page, maybe they did too aggressive link building with rich anchor text or something and they’re getting a keyword level penalty for the home page for this particular query, because there is no site links either for homesteading.com. It’s just an article, a really good article, it’s 133 Homesteading Skills For The Modern Day Homesteader but that’s raking number two, Wikipedia’s ranking number one, and where the heck is the homesteading.com homepage? It should be in the top 10 and it’s not. I found that really interesting, okay, they’re not a client, I was just curious because they run War Room, they run Traffic and Conversion Summit, huge conference every year, I’ll be going here next month. What’s going on? Why is this happening? Doing that detective work and being like a Sherlock Holmes trying to figure it out. It’s always fun.
That is the SEO game. This is what I love, that’s the cool stuff.
Much better than rewriting snippets. This detective stuff is more fun.
Yeah. Like let’s rewrite some meta descriptions. What do you think about LSI, Latent Semantic Indexing? Where do you think that fits in? What do our listeners need to know about LSI?
LSI has been around for a long, long time. I think this plays a little bit into this, right? Like Latent Semantic Indexing goes a little bit into this. I haven’t heard this in a long time, I really got to admit. If you stumbled across LSI a lot…
Here’s the reason why I asked, because the Digital Marketer folks are really big into LSI, they talk about it at Traffic and Conversion Summit and may or may not talk about it in War Room. I can’t talk about what happens in War Room. It’s like I don’t hear that term very much and the Digital Marketer folks are really talking it up in giving tools to use for LSI. I’m curious. I don’t talk about TF-IDF either over here in the US and I’m not talking about LSI either. I’m talking about like entities, I’ll talk about the Topic Explorer from Search Metrics.
By the way, it’s awesome, the content explorer tool sensation. Very cool.
And the content editor, just the whole content experience suite is amazing. I love the tools. I’ve had Marcus Tober from Search Metrics on the show as well. It’s great episodes listeners, be sure to check out Marcus Tober’s episode as well.
Great guy, one of the best SEOs around, definitely.
We geeked out on SEO in that episode as well. Just curious, any thoughts on LSI from your standpoint?
LSI and TF-IDF, they are sort of like the same base. It’s an old concept, that’s why I think LSI’s also like 1990 or something like that. It’s a very, very old concept but I think TF-IDF is going a lot further because we’re actionable. With LSI, I’m yet to find a practical approach, a really actionable approach besides doing a lot of calculation, right? It doesn’t help me. TF-IDF actually helps me. But it’s the same concept that you calculate a term document matrix and you weight the results and this is what I like about TF-IDF because it actually gives you an actionable result. It would actually give you something you can work with. We have a TF-IDF tool, we also have a content editor, we also give you the suggestion and that kind of stuff but it fails with the search intent concept. The deck you have depending on the query that you have, just a different intent, like if you have a navigation query, information query, a commercial query. It’s just different user metrics that Google will expect from this and also it’s going to be different how to rank for these terms. Of course, it’s totally different ranking for an information term than for a commercial query. I think this is just what’s not within this very broad concept of LSI. It also doesn’t help with [01:00:30], it’s just not actionable. TF-IDF really do the fact that they basically extract the most important thing, they extract all terms, they extract the unique terms and then I weigh these unique terms and then I also apply a deepening factor that can prove keyword filter to filter out for example brands or stuff like this so that they might have a stronger relation, they show a very strong correlation but it’s only due to the fact because they are a brand. I need to calculate this out of the equation. This is what I can do with TF-IDF and LSI will just give me a huge data set where I can just take my head and try to find something useful. As I said, I’m yet to find something actionable and useful with LSI.
The TF-IDF analysis stuff, that’s a big dent to OnPage.org?
Are there any free tools that gives some actionable insight into TF-IDF?
In our free version, OnPage Free, you can also use the TF-IDF tool. Actually, you can use the full power of OnPage.org, it’s only limited in size. You can only crawl 100 URLs and the usage of TF-IDF tool in our knife crawling engine is limited but it’s for use.
And it’s not just a free trial for a certain length of time, right?
No time trial. It’s a free version.
Yeah, of course. Because as I said, I got a small web master with a website with less than 100 URLs. He’s not going to pay €99 to use my product. Before he pays less money for a worse product, I give it to him for free. Why not? I’d rather have him evolve in the process and we actually see this now more and more that people, for example, started two years ago with the free account and now they work as an in house SEO at a company. They say I now can buy it. This happens all the time. It’s very cool. To see people, they work with the software, they like the software and where they end up over the years, they might also take the software with them. It’s a great tool for us. Especially being somehow like a bootstrap, not having any VCs or anyone breathing down my neck, I can do that stuff. I can just give it up for free. I don’t always need to make money with it. I just want people to use the product because it’s the best solution out there, it’s the leading technical SEO software, I just want everybody to use it.
Awesome. We need to wrap up here shortly but what would be some of the common technical SEO screw ups that your tool can find? Like for example, can it identify when people are using the disallow directive instead of no indexing the pages? They’re trying to keep the spiders out and the page is still showing up in the search results?
Why is this? Oh my god. We have that, in fact every day. We have someone with this, hey I have the noindex tag on there but it’s still in Google. Yeah, you disallowed it so Google can’t read that SEO index, so they’re going to index it. They just don’t get the concept of robots and no index.
Yeah or they’ve done both. They have a meta robots tag for no indexing the page and then they still have the disallow directive in robots.txt so Google can’t see that there’s a no index because of the disallow.
Maybe I said it wrong. But yeah, we see this every day. Very common problem. Yeah, of course, don’t forget content. We also check the external links, we can also see for example, broken external links and you can even use regular expressions that you find cool stuff. I can even look, for example, with regular expressions, I can use RegEx to find me all the pages that I link out to that has an instance of Viagra or poker on the page, so that might be a potentially shaky page, or show me all the pages that I link out to where they have links with the anchor text which includes Viagra or poker so they might be a link [01:05:22] and stuff like this. We even go one step further and also crawl all the stuff that you link out to. It’s very comprehensive.
That’s very cool. What would be another example of a common technical SEO screw up?
Redirect chains, always a big thing of course that.
So too many redirects in a row?
Of course, of course. Invalid use of canonical tags which is always stupid because Google is just not going to follow your recommendation with the canonical tags if you screw it up, if you just do it the wrong way. You have a different URL and the meta and then the header, you have just broken links in the canonical. Also HREF link, we can do an automatic HREF link validation if everything is correct. This is the English one, it can do the German one. The German one can do the English one, we can do this automatically, low times with every domain, with every page, we crawl. A lot of tools can do this kind of stuff. This is not magic, right? This is crawling a domain and showing you all the stuff that’s gone wrong but I think the strong part about OnPage.org is that you have 100% flexibility with this data. You can filter everything in real time and even with millions of pages. For example, we have our largest client, it’s a 4TB crawl. It’s tens of millions of pages, a hundred millions of images. In this case, 1.8 billion internal links, we also save the whole into the link wrap so we can also show you all the link [01:07:05] and stuff like this. It’s a 4TB crawl, and with OnPage.org you can actually work with these 4TB of data almost in real time. Meaning, basically, you can just filter for something, you can search for something, show me all the pages, all of the tens of millions of pages that don’t have the Facebook like button, stuff like this and it’s just one second and you have the result, it’s right there. I think this is really the strong suits. 4TB of data, two years ago, you couldn’t even put this on a hard drive. With OnPage.org, you can basically work with this amount of data almost in real time without exporting something to Excel and watch it break slowly. While you want to see that data, it’s right there because it’s cloud based, it’s right there. It’s very powerful. This is, I think, the most important part that you have big websites and you can actually find stuff very fast, very easy. I have tens of millions of pages, I have this former employee, where is his name still mentioned on the page? Where is this telephone number? Is his email address still mentioned on the page? There it is.
I love that you can do RegEx. I’m a big Regular Expression geek.
Nice, nice. You can also [01:08:16], we actually have a lot of clients that use this for competitive intelligence. They have online shops and they will use xpath for example, product name and the price, and then everyday they crawl basically their competitors, of course with a Googlebot in disguise. And then they just have their huge lists of the product name and the price and basically get an automated system go where they see the pricing, the different online shops, so they can act accordingly. Very easy because it’s just xpath, right? That’s the only thing I would crawl from this pages, so it works incredibly well for stuff like this as well.
Awesome. Can your tool check for the bidirectional links between let’s say a separate mobile site and the desktop site, the real alternate pointing to the mobile site and the real canonical point…
HTML, of course the same thing, same logic applies because you can do custom fields so you can basically do anything. You can set up your own arguments and so to speak for example, payment with HTML, I want to check and then we can check it bidirectional if it’s used correctly.
You’ve got an amazing tool set there.
Yeah, you will see.
Awesome. Let’s of course leave our listeners with the website URL to go to OnPage.org, definitely get the free version. Also if somebody wanted to work with your company or with you on a more involved basis than just going online and signing up for the free version or buying the paid version, what if they wanted to work with your company on a consulting basis? Do you offer any kind of services like that?
No. No consulting. We have 100% SAAS revenues. This is just due to the fact that most big agencies use us so I think it wouldn’t be fair to offer SEO services because we are partners of the biggest SEO agencies. We see ourselves a partner and not like that they have to be afraid that we’re going to steal data and clients and stuff like that. Why get into this game? We want to a make cool software much better. I don’t scale very well.
Yes. I have that same problem. Although I do take on consulting gigs. If you need consulting for SEO, you can come to me. If you need an amazing tool set, go to Marcus’s OnPage.org tool. Thank you Marcus for all the insight and the sharing of exciting war stories from the Black Hat days and ninja tips and techniques that still work today. This was a lot of fun. Listeners, do grab the checklist from this episode because this will be a checklist you definitely want to implement. I do a checklist for every single show. Top 10 things, takeaways that you can implement based on the content of the episode. Definitely go to marketingspeak.com for the show notes, for that checklist, the transcript for this episode, all on marketingspeak.com. This is Stephan Spencer, signing off. We’ll catch you on the next episode of Marketing Speak.
- MarcusTandler on LinkedIn
- @mediadonis on Twitter
- Marcus Tandler on Facebook
- Screaming Frog
- Alta Vista
- Black hat SEO
- War Room Mastermind
- Christoph Cemper
- 11 More Things You Didn’t Know About Links and Redirects by Christoph Cemper
- Greg Boser
- Todd Friesen
- Marcus Tober on Marketing Speak
Your Checklist of Actions to Take
☑ When you’re making hiring decisions, look for people who are smarter than you — especially when it comes to the work they’ll be doing in the position.
☑ Do keyword research on YouTube as well, not just on Google. Google Trends lets you do this by switching from “Web Search” to “YouTube Search” after you’ve entered your search term.
☑ Use 302 redirects instead of 301s. Christoph Cemper has found that 302s deliver more consistent performance in terms of SEO.
☑ Put some time into deciding what information to keep private and what to give away. Adding value to others is important, but so is keeping some secrets to yourself.
☑ Stay on your toes as you’re working in SEO. Instead of being devoted to tried-and-true methods, be ready to see new opportunities and take advantage of them.
☑ To get featured snippet positions on Google, use SEMrush to find weak answers by competitors, then create better content yourself.
☑ TF-IDF analysis is a major topic in German SEO circles, but underused in the United States. Dig deeper into this concept and work on applying it to your marketing strategies.
☑ Work on providing holistic content for your users. Especially when Google doesn’t know exactly how to answer the question, it favors holistic content.
☑ Incorporate related keywords (that aren’t just direct synonyms) into your copy to provide a more holistic answer to searches, and therefore to rank better on Google.
☑ Be careful when you’re using disallow. This doesn’t prevent Google from seeing your pages; no-index pages instead for that purpose.
About Marcus Tandler
Marcus Tandler cofounded and runs the enterprise SEO software company OnPage.org, which has the goal of helping people’s websites rank better in search engines. OnPage.org has ranked in the Technology Fast 500 in Europe and as one of the Top 100 Innovative Companies in Germany.
Marcus, a native of Munich, Germany, is a former super affiliate who was at one point in Commission Junction’s top 5 earners in all of Europe, also runs a super-exclusive conference (or think tank, in his terms) called SEOktoberfest in Munich every year. If you haven’t heard of it, this event takes place during the more famous Oktoberfest, and offers a place for many of the best and brightest minds in SEO to gather and exchange ideas.