Chiang Mai is a quiet place compared to other parts of Thailand, minus all the hustle and bustle. I stayed in the northwest corner of the old city, which was central and enabled me to get around easily. It’s a simple layout. A square, surrounded by a moat and ruins of the old walls, surrounding a centre full of Buddhist temples, markets and many cool spots (shout out to SP Chicken and the Cutler).
For the most part, I tried to avoid the tourist traps and blend with the locals. But towards the end of my trip I veered out east to a more westernised area to attend the Chiang Mai SEO Conference and this is where I spoke with Tim Soulo from Ahrefs.
Tim is their Chief Marketing Officer and Product Advisor and is forever in pursuit of publishing engaging data-driven content. We dive into our chat from his early days as a DJ and go on to talk about his social plugins (which he promoted through content marketing in a very niche market), right through to all things links and developments within Ahrefs.
Grab your favourite beverage and or/popcorn, get comfy and read on!
How did you get involved in SEO?
Tim: So back in my early days when I was still studying at university, I was a DJ. But once I graduated from university, I understood that by DJing I wouldn’t make a career for myself and I needed something more interesting, more serious and something where I would have a career later.
Woj: What kind of music were you playing?
Tim: Well, my passion at the time was drum and bass music.
Woj: I like drum and bass. That’s my favourite.
Tim: Awesome.
Woj: I’m a junglist mate.
Tim: Yeah, yeah, yeah. I’m a fan of Hospital Records.
Woj: Hospital Records. I like Critical at the moment.
Tim: Oh, okay. Okay. I’m kind of more into mainstream drum and bass, not the dark thing, not the neurofunk.
Woj: I dig neuro.
Tim: I was playing at a nightclub, but I had to play house music because you don’t play drum and bass every weekend. Once I graduated from university, I understood that I needed another career path. And just by chance, one of my friends told me that there’s a thing called SEO and that it was getting popular. And that there’s a guy in his company who was doing SEO and he’s making nice money.
Woj: And this was back in the Ukraine?
Tim: Yeah, it was back in Ukraine and it was like 10 years ago, I think. And I thought, “So the guy’s making nice money doing SEO. Let me tap into this and see what it is.” I found a Junior SEO Specialist position, basically with almost no knowledge. But because I was an active person in terms of creating different projects – before I joined the company, I had built different websites, music-oriented local forums for our German-based Ukrainian community and all that stuff.
I knew a thing or two about hosting domain names and I also worked in technical support for a hosting company from Canada which was outsourcing its technical support to Ukraine because our people are good engineers and generally know nice English. [This is good] if you don’t need to have calls [and can] communicate with them in chats. So, they cannot even tell if you’re Ukrainian or from America.
What did you learn at Hostopia that has helped you today?
Tim: One of the best things I learned was doing technical support – and for quite a long time as well. When I came to Ahrefs and was put in charge of their marketing, I did a lot of support work. I just communicated with customers and I was trying to figure out what tools they use in Ahrefs, how are they using them, what kind of problems they ran into. I was also helping the existing support team at Ahrefs with very small processes. Like, when you help a customer with their issue, the first thing I learned in technical support in Canadian companies is that you must ask, “Is there anything else I can help you with today?”
Because you shouldn’t cancel on customers unless you make sure that you’ve done everything for them. This is a small thing but it really helps.
I tried to implement these kinds of processes. I need to verify if they are doing it consistently or not but it’s a good thing. And because it was a hosting company, I learned about domains, I learned about DNS, I learned about hosting [and] I learned about how people are building their website.
Some are using WordPress, some are using internal website builders that hosting companies offer. It was quite fun to understand how the web works technically – not on a super deep level like how developers understand it.
Woj: The structure of the Internet, in a sense.
Tim: Yeah, yeah. That was the learning curve.
You’ve also produced two plugins – is there one you enjoyed producing more?
Tim:Well I enjoyed producing both plugins but the problem with them was that there was no market for any of these plugins. So, I think I came up with the idea of tweetable quotes before I even found the clicktotweet.com website. I was working at Template Monster – a marketplace of website templates which are built in-house as opposed to ThemeForest that is a marketplace where people can build their own templates and sell it to other people.
And as I was working there, they were constantly looking for ideas of things that could be added to templates to be on the verge of things. I tried to pitch them the social media template because at the time – it was like six or seven years ago I don’t really remember – but that was the time when Facebook introduced their ‘Like’ button which completely changed how the Internet works. Everyone started implementing it into their websites.
I’m a product-oriented guy. I like to come up with different ideas. I was trying to offer them a social media template that would take care of all these ‘follow me’ and ‘tweet this’ button, and stuff like this. They didn’t enjoy this idea [but] I still had it in my head. I enjoyed it. I wanted to create it. So, when I launched my personal blog, I wanted to have these kinds of Tweetable quotes on my blog.
Woj: It’s a good idea.
Tim: Yeah. And I created the plugin for myself in the first place and then I tried to sell it with Clickbank. And it was hard because there was no search demand for it – and the easiest channel to create a business is search. If people are searching for something, if they have problem, you can be there, you can rank, you can get these leads and sell it to them. But in terms click to tweet, no one was searching for it back then.
We had to use content marketing. I had to reach out to different blogs. I had to offer to write them articles about social media [and] about building traffic to your blog. And within those articles, I had to plug my own plugin.
When you say content marketing, do you mean guest posts?
Tim: Yeah, guest posting. I was trying to guest post for well-known blogs in the blogging world, especially in the content marketing niche and include my plugin in my articles so that people would go and buy it. So it was quite challenging to create demand for something that people weren’t actively looking for, but it was also a learning curve in terms of creating a product and marketing a product from scratch. So I think I learned a ton while creating my own WordPress plugins and trying to make sales.
Woj: Yeah, right. And I guess it’s a saturated market. There are a lot of similar tools out there, so that would have been even more challenging.
Tim: Right now, yes. But back when I started, I created one plugin for tweetable quotes and another plugin for content upgrades.
Woj: And [with content upgrades], is that where if you want to see the rest of the content, you leave your email address and you’ll receive an email?
Tim: It’s not [so much] about the rest of the content. But the content upgrades were more about providing something extra. It could be a PDF download, it could be a cheat sheet, or it could be a video walkthrough of the same thing that you talk about in an article – but you just try to provide it as an extra value that is directly relevant to the article. You ask people to leave their email to get [the extra content]. It was a good way of converting readers into email subscribers, and I think I was the first person to launch a WordPress plugin specifically for that.
There were other kinds of similar solutions but there wasn’t a solution that would focus on this strategy. And, again, there was almost no search demand for it. So, I had to write guest posts for other blogs and put this kind of strategy in front of their audiences and build sales for my plugin.
How did you create a domain outreach list to potentially publish your content?
Tim: Back then I had to use Google for this. So, I was just searching Google for things like social media blogs, and I probably found 25 popular social media blogs and 32 popular blogs about blogging – and those were basically my outreach prospects. I think I was using Alexa at the time to see which of these blogs were getting more traffic, and this is how I was picking the targets to write a guest post for. But these days, I would totally use Ahrefs for this 🙂
Woj: It’s a better tool. It’s already automated.
You’re now living in Singapore – how have you adjusted to the heat?
Tim: I don’t think I’ve adjusted to the heat. I still hate the heat, to be honest. And I don’t like to get exposed to the heat very often. I try to stay inside the office or at home or visit places like museums that have a roof and air conditioning. So, I don’t expose myself to the heat a lot.
Woj: And all the tech around you? Because I know you’re a big fan of tech.
Yeah, I enjoy it a lot. It’s super clean. It’s super convenient. And you don’t waste your time on things that you shouldn’t waste your time on. You can focus on stuff that matters to you. You can focus on work, on family and not even think about things like filling your tax forms.
I know, for example, in U.S. it is a huge pain to fill your tax forms. In Singapore, you just have to go online, click Next a few times, click OK, and you’re done.
Woj: It’s simple back at home as well. Will you be heading to Australia soon?
Tim: Of course! There are two conferences there and I’m going to be speaking at both of them. One is around March [Digital Marketers Australia] and one is in May [Search Marketing Summit].
Woj: Oh, that’s good to know. I used to co-run Big Digital and we stopped doing that. I might start another event up in Adelaide. It’d be cool if you’re around because we usually did it around SMS time, so it was easy for the speakers to come across. If it happens, maybe you could come to Adelaide for it.
Tim: Let me know.
Woj: For sure. So, there’s also a lot of talk about keywords salience at the moment. That’s where the number of keywords and the distance between keywords on a page is having a similar impact as links. Google is now, through machine learning, able to determine entities on pages and could consider it similar to a link…
How important do you think [keyword salience] is versus links?
Tim: Well, we don’t have any data about this, that’s for sure. It’s hard for me to tell. I still think links are fundamental to how Google works.
Woj: …and Page Rank as well.
Tim: Yeah, Page Rank for sure.
It’s the foundation of how Google became what it is today because it’s probably the best way to see which pages have more value. When the content of the page is similar, how do you know which is better? Even for humans it’s hard to figure out which article is better if they don’t have any external ranking system for what people think about them.
And in terms of co-occurrence of keywords and matching different concepts and different entities to each other, I think – and it’s just my opinion – if it is a signal right now, it is weak.
But in the future, as I see…
…one of Google’s ultimate goals is to start to read and understand content in the way a human does. But since they can do it at scale and they can read all content there is on a certain topic, they will probably be able to identify which content is more comprehensive, which content has more authority and things like that. So, going forward, I think [keyword salience] can be a thing. There’s a high chance.
But right now, even while talking to other SEO professionals, I don’t see that they are putting a lot of effort into studying this concept and to implement it.
Woj: There’s a good article on Briggsby that talks about the semantic elements on-page and ordering tags, having the correct order of the salience and different things. It’s interesting. Understanding the keywords and the salience versus physical links could be a good idea to explore.
Will Ahrefs develop some sort of entity relationship counter between pages?
Tim: Yeah, we might. Right now, I can tell you that most of our brain power is focused around building the best web crawler out there.
We’re trying to release a super advanced crawler and we’ve been working on it for more than two years. And we’re secretly hoping that once we release it, it will be crawler technology that is even better than Google’s.
So, if that happens, we might get some credibility for it. I think our guys might even come to some conferences to present our technology. I don’t know about this, but if we’re able to figure out a better web crawler than what Google has…
Woj: And how would you compare that? Based on speed?
Tim: Based on being smart. It’s not the speed. Speed is achieved by simply adding more servers. [Our crawler] was the topic of my talk today – I was saying that crawlers only save URLs and then there’s a scheduler that will identify which URLs that the crawler should crawl next. So there’s a delay between discovering a certain URL and crawling it. What we’re trying to do is we’re trying to get rid of this delay.
Woj: Minimise it.
Tim: Yeah, and we’re trying to build a crawler which will automatically identify where should it go next and not tap into four pages and quickly identify which links are more promising, which links have more value.
We’re trying to decrease the time between discovering a URL and crawling it. And [one] way we would know that it’s better than Google’s is simply by analysing log files and looking at the crawler behaviour on random websites and seeing if our crawler behaves in a smarter way than Google’s.
So, if we’ll be able to nail that, of course, [we will move onto] the next big thing, but I cannot say that we will because I’m not the one making these decisions.
Woj: You could become a search engine.
Tim: Eventually…
…we could become a search engine.
But it’s a better question to ask our CEO rather than myself because I’m not the technical guy. I’m just pushing [Ahrefs] technology to the world.
Woj: That’d be interesting though if you became a search engine after becoming a marketing tool. And then if your crawler is smarter and your schedule is better, it could be a real competitor to Google because at the moment, there’s no real competitor.
Tim: Yeah. But we need the technology to understand the content of the pages, which, right now, is something we don’t have.
Woj: And you mentioned 90 days. So that 90-day period, is that like an average for the delay?
Tim: No, it’s not an average for a delay. It’s just an arbitrary number that we picked for displaying the delayed metric.
People get emails from us with their newly discovered backlinks, and a lot of people were complaining that whenever we discovered a backlink and report it as new, it came from an older page. This happened because it took us a while between discovering the URL and crawling the content of the page under this URL to then sending the link to our customer. We thought that a good kind of timeframe is 90 days and so we think that if it took us more than 90 days from the moment that we discovered the URL to crawl it and see the link, then we’ll display it as delayed.
And if it took us less than 90 days, then we consider it was quite fast. So we consider it fresh and don’t like mark it as delayed and send it as new.
Have you read Dejan’s recent article about link inversion?
Woj: He said “if there’s multiple instances of the same document on the web, the highest authority URL becomes the canonical version. The rest are considered duplicates.” It’s an easy way to basically pinch someone’s rankings if your domain is a higher authority, if you reproduce the same content. Have you seen cases of this?
Tim: No, I haven’t seen any cases of this, but I think Dejan’s article was quite persuasive. I think he presented it with a real example. So, there’s no way for me to argue with him since it’s hard data.
Woj: Also, I think Bill Slawski looked into a patent afterwards and it wasn’t explicitly verified but you can make a correlation.
What about link ghosting? Have you heard of this where links get removed but they still count [and] they pass the authority through?
Tim: I think one of the possible reasons where the link ghosting idea or myth comes from is the actual scheduler part of crawling the web. Once the link gets removed, it might take a crawler quite a while to figure out that the link was removed. Especially [in cases] where – like I explained – the page that was hosting a link didn’t have high Page Rank, if it is not updated very often or if there are other signals that are telling Google that they shouldn’t visit this page too often.
There’s a delay between the link being removed and Google realising that the link has been removed. Then there’s also a little bit of delay between Google recalculating the value of the link that was lost and taking it into account while ranking the pages.
So maybe it could be just this delay or maybe Google might be considering that, for example, if you, at some point, had a link from the homepage of BBC, maybe that should still benefit you in the long run even if the link is not there.
Woj: So, it could be an interesting one to study because I imagine they have a historical record-set with different rows [that showed results] on particular days. And like you’re saying, if it originated from like a really high authority domain – for example a .edu, .gov or the BBC – why not still credit it if it once was there?
Tim: From our end, we wouldn’t be able to study this. A lot of people tell us, “You have so much data, you should be studying all these things.” But think about it, if a certain website got links from the high-quality websites and those links disappeared, then this website should be an authority website [in its own right] – so it might be getting a lot of other links from other good websites anyway. So, you cannot isolate the value of those links.
The only good way to study it is to launch a brand-new website, get some links from high authority websites to it, then remove all those links and look at what happens. An even better way would be to have a few of these [startup] websites to confirm statistical significance as they say. There is nothing better in SEO than an actual experiment – correlation doesn’t mean causation.
How important are internal links? It’s something I think we often skip over.
Tim: In my opinion, they are very important. And, like, again, I’m not the practitioner in SEO. I only do SEO for our website. And for us it is quite different because we have [a well-known] brand and it helps a lot. But I’ve seen a lot of cases and I’ve spoken to many, many SEO professionals who have increased the traffic to their website just through smart interlinking – by figuring out which pages don’t get a lot of internal Page Rank, optimising this and like having a huge uplift in traffic just by optimising the internal linking.
Woj: And even pruning pages and getting rid of low traffic URLs. I call it the bonsai process. Make an ugly bush into a nice Bonsai.
Tim: Yeah, true, true. We did that for the Ahrefs blog. We’re constantly deleting the content that doesn’t perform well and updating the content that does perform well.
How closely does the Ahrefs team review Google patents when considering calculations of metrics?
Tim: In terms of calculating our URL rating, we did study their Page Rank patents but then we try to not overcomplicate things because the higher the complication, the higher the cost of computation. Then we’d have to charge our customers more, which they don’t want. So, we have to keep the balance between making our metrics simple and easy to calculate and at the same time, they should be effective and effectively show the value of the pages.
How big is the data science team at Ahrefs out of curiosity?
Tim: The data science team is not big, but the thing is most of our backend developers are well-versed with mathematics and data science. But we don’t really have a big team that could work on AI or machine learning or stuff like this. We only have two dedicated data scientists. But like I said, other members of the backend team who are working on the crawler [and] who are working on the databases that we have, they are very good with mathematics and all those concepts. So, yeah, each person kind of can contribute to making a good product.
I noticed when you mentioned your recent definition of metrics – there’s no spam score currently. Is that something that you’re going to include or is there a reason why spam scores haven’t been included?
Tim: We’re figuring out the best way to do this and we’re not rushing it, partially because of what I read in Rand [Fishkin]‘s book. Did you read his book, “Lost and Founder?“
Woj: Yeah, yeah.
Lost & Founder #NYC book launch @NeueHouse feat. @randfish & @iPullRank pic.twitter.com/7KNkZdXtbW
— Woj Kwasi (@WojKwasi) May 10, 2018
Tim: So, he was saying that [Moz] created a spam score and if I’m not wrong, he said that they didn’t invest as much effort as they should have invested into it. And they released it in a pretty raw state and the community wasn’t too excited about this.
Woj: It had a negative impact.
Tim: Yeah, and it backfired on them. We don’t want to repeat their mistakes and don’t want to release something that might backfire at us. We’re still testing options and seeing like if we can get a good metric that people will be happy with.
In another article you studied the correlation between domain rating and organic search traffic with 200,000 plus domains and saw that they do correlate. So was that something that you anticipated?
Tim: Of course, of course. If a domain has a ton of backlinks, it means this domain is popular. What does popularity mean? It means that a lot of people are visiting it. You don’t even know… it’s the chicken and egg problem. Does a website have a lot of traffic because it has a lot of backlinks or does it have a lot of backlinks because it has a lot of traffic? But those two are clearly connected. You cannot have a website with tons of traffic and not get a single backlink.
Woj: Yeah.
What’s your favourite link building or acquisition strategy?
Tim: For us, it is creating something worthy of being mentioned and reaching out to people who can mention it. This is what works for our content marketing team for the blog that we have at Ahrefs. Whenever we create an article with some interesting research, we’ll try to reach out to some SEO professionals, show them our data and some of them will later mention it or even include it in their slides and their presentations.
You mentioned that Ahrefs is working on a sophisticated crawler. Are there any other exciting things you’ve got planned?
Tim: Yeah, a lot of exciting things. One of the exciting things is that, like I said, there will be a [displayed] date of when we first discovered the URL. We’ve started to implement this in our tool called Content Explorer. A lot of people update their articles [and] they update the published date on their pages. Within Content Explorer, we are about to have two dates – one which is when we first discovered the URL with content and the second date is the one that is shown on that URL. You’ll be able to identify in Content Explorer the articles that were published long ago but are consistently updated.
In other words, you will be able to identify evergreen content quite easily. We will also have a graph – so whatever keyword you put into keyword explorer to find articles that mention that keyword, the graph will show how many new pages were published and how many old pages were updated over time.
Woj: It’s very cool.
Tim: Yeah, it will be quite interesting and no one else does it because no one else has a tool like Content Explorer. And one other cool thing about Content Explorer is we stopped deleting pages when they go missing or when they go 404.
A year from now – as we fill our database with dead pages – you will be able to search Content Explorer for content from non-existing pages.
Woj: So we’ll be able to filter by 404?
Tim: So, yeah, you’ll be able to search the 404 pages that mention link building in their content, and sort by the number of referring domains. Then you can find articles that talk about something – they don’t exist, but they have a lot of backlinks then you can use the broken backlink strategy, which is cool.
And in terms of keyword research, we’re about to release data for YouTube, like the actual YouTube searches.
Woj: I saw that.
Tim: Yeah. Amazon searches, Bing searches. It will be awesome! I don’t think there’s any other tool today that would show you YouTube searches other than the tools that are scraping YouTube auto suggest. So right now, auto suggest is the only source of keyword data for YouTube but we are about to get it from clickstream, which are the actual searches that people make in YouTube.
Woj: Is that a service that’s linked direct to YouTube, the raw data from YouTube, or like an in-between?
Tim: No, it’s clickstream data. It’s when people install different plugins within their browser or different software, and then software [through their] terms of service is asking to get their browsing data, anonymise it and resell it to other companies.
Woj: It’s a percentage of searches… not a representative of all searches?
Tim: Of course. But we’ll try to use some algorithms to extrapolate the kind of the percentage to the whole population and give you the estimates. At the end of the day, all search volumes are estimates and it’s better to have an estimate. [It might] not be particularly accurate but when you compare two keywords between each other, you can see that this keyword is three times more popular than this one. So, it helps a lot.
What does the future look like for you?
Tim: Well, I’m dedicated at staying at Ahrefs because I respect our developers a lot and I respect our founder [Dmitry Gerasimenko] a lot. He’s super technical and he’s capable of making awesome, awesome technology. And I had experience of being on my own and trying to build my WordPress plugins, my online tools, and I felt the challenge of finding a co-founder who will be technical, and who will take care of making sure that we have the latest technology, the best servers, and all that stuff.
And with Dmitry, I feel that synergy. [I feel] that he’s capable of doing almost anything and I’m pretty good at coming up with different features, different new tools. I feel like I’m better with Dmitry and Ahrefs, and even if Ahrefs transforms into anything else, I still want to have Dmitry. I still want to be next to Dmitry because he’s so good at technology.
Woj: Cool. It’s good to have a workplace where you really respect the people around you and it doesn’t feel like work. It just feels like the right thing.
Tim: Exactly.
Where can people connect with you online?
Tim: I have a dedicated page at ahrefs.com/tim – and on there I list all my social media profiles as well as some of my best articles [and] best research studies. So, if people want to learn more about me and read some of the content and research studies that I’ve done, that’s the place to find all the best stuff.
Woj: Where would you say you are most active on social media?
Tim: I’d say it’s 50-50 between Twitter and Facebook.
Woj: Cool. Hopefully some people will read this and follow you and enjoy your content like I have.
Tim: Yeah, that was awesome. Thank you.
If you enjoyed this interview and live for innovation, internal link, and want to keep up with Ahrefs super crawler news be sure to follow Tim on Twitter – and check out some of our other interviews as well:
Comments (0)