Senthil: Hey guys, welcome to a special edition of SEO On Air. Today we have the big daddy of SEO, Rand Fishkin. Rand, welcome.
Rand Fishkin: Thanks for having me.
Senthil: Great. So, during this discussion, we would like to discuss more about SparkToro. About the product, about how we are using it as an SEO agency, the benefits we have seen. And we’d also like to ask you a few more questions, which I hope nobody else has asked you before. So, hopefully, this spans out good.
Rand: Sounds great. Let’s do it.
Senthil: Awesome. So I’m a bit of an old-timer. I’ve been doing SEO since 2007, so I’ve been following you a lot. Moz is something which everybody knows by now, but I still like to call it the SEOMoz because that’s in our blood. That’s what saved us during the Panda and Penguin updates. You co-founded Moz in 2004 and then the bestselling book, The Lost and Founder, and then SparkToro. It’s been quite a journey for you? So what, what keeps you awake at night?
Rand: I’m definitely very, very worried about this pandemic right now. I think that’s hitting a lot of people pretty hard. We have some sort of friends of friends and family who have passed away here in the US. It’s terrifying to know, to think, how many people have died recently. So that is definitely keeping me awake.
Also, on the business front, I’m worried because a lot of people’s budgets have gone down. The US economy has taken a huge hit and it’s going to be a long time before it recovers. It’s not necessarily a bad thing to build a new startup in a down economy. Moz launched its Open Site Explorer the first time in 2008. That was when Lehman Brothers collapsed in the financial crisis.
So that did end up working out okay for us, and Casey and I hope SparkToro will be similar.
We believe that a lot of folks will be looking to change their behavior as marketers and maybe move away from spending a ton of money with Facebook and Google and try to go towards how can I build these marketing relationships myself and potentially save money and be more creative? And SparkToro is certainly in that field. So, we hope we can help with that. I would say I’m not sleeping terribly. I have my nervousness, but I’m a lot less stressed than I was in my last few years of Moz.
Senthil: So after the book launch and all the other things that are going on, what was that aha moment that made you think that SparkToro is the one that we should be going up? Is there any specific incident that made you think there is this problem we should solve because your community didn’t have a strong tool?
Rand: Yeah. With SparkToro, it was really around some startups that I was helping in the latter part of my career, especially ones where, and I’m sure you folks encounter this too, where someone has a product or a service and no one’s really searching for it. So instead of trying to rank in Google, you have to go reach the people who are the right target customers and target audience in places where they already pay attention to your product, your marketing and your new message. I was actually on the board of directors of a company called Haiku Deck here in Seattle. They ended up selling, but that company was experiencing that pain point where I had to pick up a hundred keywords that they could rank number one for in Google, or it wouldn’t have really helped their business.
Search was not the way for them. They needed to get in front of the right people who wanted their product and bring them into the fold. And this challenge of finding any marketing channel that can reach the right audience was really painful and hard. So, I did some research that involved how to solve this problem, and the big answer was Market Research, which is essentially surveying tens or hundreds of thousands of people. But those studies are done by market research firms. However, to put together custom studies like that, these firms literally charge tens of thousands, if not hundreds of thousands of dollars for one report. And that survey gives information about the folks, how do they pay attention to your market, what they read and listen to.
And self-reported survey data is bad. It’s not good. If you survey someone and ask them what podcasts they listen to, it’s not nearly as good as looking at their Spotify profile and seeing what they subscribe to or looking at their YouTube profile and see what channels they subscribe to.
Self-report data versus real data- it’s just not great. And so, Casey and I realized, “Hey, we can go crawl this data ourselves.” We could crawl it manually. The same way Google does with Twitter and LinkedIn and Facebook, Reddit, YouTube and Instagram and then aggregate those profiles and make them searchable.
So you could just search for any profile and hopefully help solve this problem. We didn’t know if it would work, but that aha moment was sort of just at the end of my Moz tenure. Like the last year, I was there. And I started the company the morning after I left Moz.
Senthil: Oh, okay. That’s great. I’m going to show you a live demo right now. Like how exactly it’s going to save time using SparkToro. We’re going to explain it like how exactly as an SEO, we are perceiving SparkToro and trying to put and how it’s saving us money. But before I jump on to that, just one quick question. So, a couple of weeks back, I was speaking with Neil Patel. I was of the impression that there should be about 70, 80 developers behind the tool UberSuggest. And he shocked me by saying that there are only eight people behind the tool. It’s a very lean team and he’s pulling in terms of features. How many people are powering SparkToro?
Rand: Just one. So Casey built the tool himself. He built the backend and the frontend. We contracted with a UI person and a brand designer. But Casey has built the whole thing himself. And obviously, that is why it took a little bit longer. It took two years of development to build it, but I actually think even if the team were eight people or 10 people, it probably would have been a year.
Maybe we would have saved some time. But Casey and I knew that we wanted to go very, very lean for this. And we had an idea of how to build things in a lean way and how to iterate. So this was not unintentional. We did it this way, knowing what we were up to, and obviously I’ve had the experience of Moz, right. There I had a team of developers that were initially two people and then 20 people and then a hundred people. My sense is that you can, with a very, very small team, build a lot of high-quality stuff. You can see that SparkToro is a very focused tool. So it does not have hundreds of features. But it does have a dozen and they’re very focused on solving the problem and getting the data.
Senthil: Exactly. So folks who are listening to this podcast, don’t just throw in all your bug reports to Rand. Please do send, but be careful about the feature requests that you’re going to send out.
Now I’m going to show you exactly how SEO we are utilizing the tool characters if I’m, if we are wrong and how we are using it because, you know, this much better than us. So, when we are doing outreach, typically let’s take, for example, if one of our clients from Australia, they are a furniture company, they are selling furniture products online, and you have to do outreach and figure out, which blogger would like to talk about them.
Traditional approaches that we’ll be just searching on Google, for the keyword “furniture guest post Australia.” And we’ll just try to figure out if there’s any good website that’s popping up. And if there are like good websites, then we’ll probably do a bit of research, check the DA, check the Spam Score and the backlink profile. And also just, you know, go through the websites of the audience, what type of content they are adding on the site. And then we shortlist this site and like this, we repeat this exercise to at least 40 to 50, because we know that we can’t just rely on just one website because the client might, it’d be requiring at least five to 10 outreach on a month.
And if I’m going to do that, then it’s going to be a repeat exercise. It’s going to take us at least about five to six hours to shortlist these websites and then do the outreach. And then if the person responds, if it’s going to take a lot more time. Now, the thing that we really found about SparkToro is that we don’t know have to jump around here and there to multiple tools, trying to figure out what the DA is, what is the audience size and everything. I’m just getting everything about this website in SparkToro.
And most importantly, I’m also looking at the other websites that the same audience are visiting. And what can also do is to create a list, add all these to it. And it just doesn’t stop here because next, what we are also going to do is, um, we also are going to check the social profiles, people who can do the shout out to these posts that we are doing, and we’re going to create the list.
And then finding the YouTube and Podcasts that are actively watched and listened to literally takes just 30 to 45 seconds max. My complete media plan is ready for us right before us. That is something which otherwise is going to take, one senior SEO to do the research and then a junior SEO for doing the outreach and then a social media person to research on the social media profiles, run it through the various tools. Three people combined together about 15 hours overall to create a list of 25 to 30 websites.
And that is going to be literally like 15 hours that calculates to 500 to $600. And it is not just about the money, but it’s also about the time. You know, the school saves us literally 14 hours and 15 seconds. And I have the entire media plan in front of me. This is, yeah. I think one of the nice things about SparkToro. It is very streamlined and fast, but also to your point. If your goal is to find one or two websites or social media sources that attract the audience that you want you can find it really quick.
The compare audiences feature within SparkToro is actually even better. It just takes the game to the next level. So if I’m going to look at people who talk about furniture, but also listen to one particular website. I’, even more, refining my target websites that I want to reach out to.
And, uh, you know, last time when I was speaking with Neil, he mentioned that the way that he sees backlinks is that the relevancy is the number one thing when you’re building a link out. And this is like super relevant sites and social profiles that we are getting in a blink of our eyes.
This is like gold stuff. I think there’s a lot of stuff in the store for agencies if they use SparkToro.
Rand: Well, I will say, one thing we’ve definitely seen during the pandemic is a lot of the agencies and consultants, a lot of the individual practitioners that we talked to, they’ve had their credit cards pulled. This means no more corporate cards, no more tool budget, those kinds of things.
And so what we did with SparkToro is we actually increased the free plan. So originally it was, I think, five searches for free. Now we’ve made it 10 for free, and we show much more of the results each time you query. We don’t charge twice for a query if the audience size is smaller than 150 profiles. So, we’re just trying to make the free plan as generous as we can for a time when we know people need to get value from free tool usage, as opposed to having the budget that they might’ve had three months ago, six months ago.
A lot of folks have signed up for free, I think we’ve had 13 to 14,000 people sign up for the free account version so far and a lot of them have used somewhere between three and eight and that’s great. It feels good for us to know that the SparkToro is helping a lot of people, even if they don’t have any money for it.
Senthil: That’s great. That’s a great help for the community.
Rand: Yeah. And I think this is something a lot of marketers should be thinking about with their clients and their projects too. Right now, it is a great time to build up future demand, grow your email list, grow your organic traffic. Don’t worry so much about conversion rate. Don’t worry so much about payment. Right now is a great time to build the brand and free value, and earn traffic.
Dileep: So, I personally found SparkToro has a lot of data when it comes to people who do influencer marketing and also manual blogger outreach. There are so many agencies out there that are spending a lot of money on macro-influencers who the target audience are not probably following. However, they are spending money on these people and the kind of the revenue earnings that they’re getting is really less. But with SparkToro, we are able to identify micro-influencers whom you can reach out to, and probably they will be happy to promote your product probably for free. What is your take on this?
Rand: Yeah. I mean, Casey and I intentionally built SparkToro. It does help with what you would call marketing to sources of influence. At least in the United States, there’s this weird thing that’s happened called influencer marketing. Five or maybe 10 years ago, Influencer marketing sort of meant all the websites and podcasts and blogs and social sources that your audience pays attention to. And now it really just means, Instagram and YouTube.
I like to tease and say, these are half-naked people who pose with your product for $500, that’s sort of an influencer. And we’re not really in that world. SparkToro will not show you, Hey, this person has 10 million followers on Instagram. And when she posts, she gets 10,000 comments, that kind of thing. In fact, SparkToro has very little Instagram data. It is much more centric on websites, Twitter, LinkedIn, YouTube, and a little bit of Reddit.
There’s been github profiles, medi that kind of stuff. So much broader. And I think that the part that you talked about, what’s really substantive for me is when you go and pay an influencer or you go and pitch a publication, or you go and do marketing in a place, and it has a big audience, but it’s not your audience. And that makes a huge difference. A big audience is not worth all that much to you unless you’re selling or trying to brand something that every person needs.
But if you are a furniture shop, if you are an interior designer, if you are selling a book to entrepreneurs, if you are writing science fiction novels for people who are passionate about that world – whatever your niche is, you want to reach people who care about that niche. And so SparkToro is really at that, it has that one column, that percent of my audience that pays attention to this source and that is just so much more meaningful.
That’s what you’re trying to do when you do those really expensive market research reports. What percent of the people, who we think are our customers, pay attention to this media source versus that media source? And just being able to get that at your fingertips is so valuable. Now, I think if SparkToro were reduced to like the simplest version of the product, that’s all it would show. You’d search for the audience and all you’d see is the name of the publication or person percent of the audience. That’s sort of like the core.
Dileep: Okay. Adding to that, I also found three free tools that you have provided. I couldn’t find it on the homepage, but when I did a dig in, I was able to find it. But those are really good tools because I know there are a lot of agencies out there in the U.S and other parts of the world that kind of sells a lot of followers. Selling followers is now pretty much trending. So, if I am somebody who feels that I have very little followers, now I can buy a few. So, you have cracked that using that fake follower tool. I think you are cracking down on black hat ways of a social profile building.
Rand: Yeah. We’ve had a lot of requests from folks asking if we would expand the fake followers tool, especially to Instagram. I think Instagram is where a ton of the fake accounts are being created manipulatively. That’s something that I think we want to do long term, but short term, we’re very focused on the paid tool at this point. So probably, I’m guessing, we will add more Instagram data to SparkToro itself to the audience intelligence tool, then we’ll probably do some promotions with that with fake followers.
A big part of why we do fake followers, why we built that initially was to be able to identify real activity versus fake activity. Okay. We don’t really want to crawl an index and use fake profiles in our database.
We want to make sure that you’re getting real people. So I think fake followers were essentially like a free tool for us as well. We had to build it for our own purposes for the paid tool. So let’s just make the data free for other people to use too.
Dileep: So, you mean the Spark score is also based on this particular data?
Rand: Yeah. So Sparks Score includes the idea of how much real engagement you get, and that data comes from a part of how we analyze the fake followers. The other thing is we use fake followers to help see what to crawl and index and what not to. So when you look at the profiles that are in the 70 million profiles, which primarily they’re mostly in English as you’ve seen right now, but eventually, we want to start scaling to the rest of the world.
But that database of 70 million profiles is very focused on profiles we know to be real as opposed to ones that we think might be fake. And so fake followers was essentially a way for us to index the right kinds of real accounts of social media.
Senthil: We have a rapid-fire round that is totally three things. One is the rapid-fire round next is the personal round, and then a super personal round. So the first thing, the rapid flight round that is going to be five questions. You’re going to answer with an Yes or No.
Dileep: Does dwell time help in improving organic rankings?
Senthil: The more links, better ranking. Is that true?
Rand: Generally, yes. Okay. Not always, but there’s a high correlation.
Dileep: To rank on top, you need 2,500 words of content?
Rand Fishkin: No, no.
Senthil: Okay. So that’s two yes, and one no. The fourth one – the more social engagement, does that boost ranking? That is, if we pay a hundred dollars, boost our posts on Facebook, does that mean it’s going to help with the ranking?
Rand: Ranking in Google?
Senthil: Yep. Ranking on Google?
Dileep: Okay. So mobile speed can improve ranking?
Rand: A teeny, teeny, tiny bit. Tiniest.
Dileep: Okay. Because Rand, especially with that well-timed question, I have seen you, probably almost on the verge of a confrontation with the people in Google when it comes to you telling them that yes, dwell time and time spent on a page is an ever-ranking signal. But Google has been rejecting them all the time.
Rand: Well, Google has a very complex way of analyzing user and usage data signals. They look at search patterns and they look at people who click a site and come back to the search results and choose a different result. And whether they solve their query on one page or another, they probably have extremely complicated formulas for this. And so, when they hear someone like me or you say dwell time, they’re like, no, of course, we don’t use that.
For our purposes, it’s sort of like close enough. And certainly it’s the case that if your website does a better job of solving the searcher’s problem. And so people choose you more and when they choose you, they don’t go back to the search results. Even if that didn’t help rankings, you should still do it. It still helps your website get visitors to turn into customers and turn into happy repeat visitors. So you should do it anyway, sort of like page speed. If page speed wasn’t a ranking factor, would we do it any less? Heck no, because we know that when things load slow, we click the back button. We don’t come back. So in a way it’s sort of a yes. And also it doesn’t matter, do it anyway.
Dileep: Yeah. So anyway, it’s better for your website.
Rand: What are you trying to do with web marketing? What’s the point of ranking in Google? The point of ranking in Google isn’t to rank in Google, it’s to get people to come to your website, solve their problem, buy from you, become fans of yours, get into your brand, get on your email list.
And those things are all improved by doing things like making your site faster, improving your pages per visit, making searchers more successful with solving their problems. So your goals and Google’s goals are aligned. Whether Google wants to admit exactly how they measure what we sort of would call “searcher satisfaction rate,” I don’t really care. I don’t care if they say no, we don’t want to use dwell time. They use something that measures searcher satisfaction that almost certainly includes something that’s correlated on our side with dwell time and bounce rate.
So if you’re bringing your bounce rate down and you’re driving your dwell time up and more people are visiting more pages on your site when they visit you and they’re solving their problem better, you’re winning. You’re winning regardless of what Google says they use or don’t use.
Dileep: Yes, yes, definitely.
Senthil: Great. So now we’ll move to the personal round, Rand. Now, you have to tell us exactly what was going on in your mind the day before SparkToro. And on the other side, there is this conundrum going on, with the COVID and all the other things. I personally can understand what should have been going on in your mind, but can you tell us what was going on in your mind the day before the tool was launched?
Rand: I was feeling, generally about SparkToro, I was feeling confident about the launch because we’ve had a long period of beta testing and a couple of months of early access testing. And because Casey is just so incredibly reliable, The things that he’s built, like they really hold up as compared to some of my previous experiences.
I will say I was feeling this sort of existential frustration. Like, gosh, I know that if we weren’t launching in the pandemic, I know that if Covid weren’t happening, so many things would be better. So many people would be alive who have died. So many people who have to stay home, who have lost their job. So many things. But also, just sort of selfishly and personally, like, gosh, if this hadn’t happened. I knew from our February cohort of early access what it looked like when we sort of launched SparkToro to a very small group.
We sent 2000 of our 20,000 emails before the pandemic hit, or before it hit the US really, and that group performed at like a 5% conversion rate. And by the time we sent our last early access group, same email list, same general people, it was down to 0.5% because, so many people were just like, whoa, I can’t spend anything. I don’t have time to check anything out. I can’t pay attention to anything but this sort of driver in my life.
We were sending those emails and as we were sending them, we were seeing so many bounce backs that didn’t say out of office, they said so and so no longer works here. I was getting those and I’m just seeing that increase. Usually, it’s like a 10th of a percent of an email list has that up to about 15% of our email list, saying “this person no longer works here.” So that was like kind of eating my heart, as we’re going to launch.
Dileep: Now the second personal question, Rand. So, you have written a fantastic book. One reason why I can connect with you is because I have read the book, almost each and every page, I have just read it in depth. So I can relate to you more and the kind of personal experiences you have shared in the book. It has been a great knowledge for me because I don’t think any CEO will reveal so much information of how he started, how he struggled.
You have, dot by dot, explained everything. So let me ask you this question, writing a book or launching a product, which one do you think is more interesting and also more time-consuming? Which took you more effort, which one was more stressful?
Rand: Okay. What’s crazy is Lost and Founder SparkToro both took almost exactly the same amount of time to build and get to launch. I will say, I think the stress and pressure around Lost and Founder is a little bit more around the book because once it’s done, you can’t ever change it. I mean, you can go back and subsequent printings and sort of fix an error or something, but that initial print will be out there. That book exists as it is.
The audio version that I recorded in a studio exists as it is. And so if you have a terrible take on something or you make a point really poorly or history looks unkindly on the advice that you give, that’s really hard to take back.
I don’t know if you guys have read, The Hard Thing About Hard Things by Ben Horowitz, that’s sort of a very popular book in a lot of circles.
But like for me, I don’t know how well that book holds up. It has this sort of bravado and hustle culture attitude that (to me) doesn’t play as well, as it did when it first came out. That’s not to say you can’t still learn things from the book, but I don’t know if it’s going to be that way. I hope Lost and Founder holds up for another 10, 15 years. That would be wonderful for me. SparkToro, on the other hand or any software product you build, if it’s not working, you can change it.
We can change it for every person who visits it tomorrow by fixing one. Someone emailed me this morning, “Oh, if I add a podcast to my list from the overview page and it goes into the wrong section” Great. Casey fixed it. So tomorrow that problem won’t exist for anyone. So that pressure is different.
Senthil: So yeah, that’s it, Rand. It’s been a pleasure speaking with you. It’s been like one of the best podcasts that we have done. We enjoyed speaking with you.
Rand: I really enjoyed it too. You guys are great hosts, and I look forward to hopefully being back in the years to come.