Hands down, Google is still one of the best platforms to drive growth for your business.
It’s a competitive platform and one that is continually changing, so how can you stay one step ahead when it comes to doing SEO that will generate significant returns in the form of organic traffic?
Luckily we got to sit down and ask Ethan Smith for his advice on how to grow user acquisition through search. Ethan has been an advisor for websites like Yummly, MasterClass, TicketMaster, ThumbTack and many more.
Get ready to take some notes; there is a lot of great info coming your way 🙂
Happy Growing !!
Kieran: Hey everyone, on this episode of the GrowthTLDR we’re going to talk with Ethan Smith.
Ethan is a growth advisor to some of the most successful websites in the world including, Yummly, Thumbtack, Ticketmaster, MasterClass, and Honey.
Thanks for being on the show Ethan, welcome to The Growth TL;DR. Thank you for being on our show.
Ethan: Thank you for having me. Very excited.
Kieran: Let’s talk about doing scale up SEO. Ethan, for these big sites you work on, do you have a go-to SEO playbook for large websites?
Ethan: I have a set of principle that I try to use. But step one is to make sure that things other than SEO are in a good place. If you decide to grow purely based on SEO, you’re probably not going to do well. You would do well ten years ago, but not today. So, making sure that you have good product-market fit, you have good unit economics, you have good content, all of those things will allow you to have a base from which to grow. Otherwise, you’re just acquiring users and burning them. So picking the right company, selecting the right market, the product at the right stages are all super important.
The second this is there’s a class of things that have worked well over the past few years, and then there’s an entirely different set that will work well over the next coming years. So, the stuff that worked well in the past was a lot of algorithmic, large site, hygiene related things, where basically most sites accidentally do things that cause them issues, and then you can clean those things up.
So the types of things that those are, are typically you have way too many pages that are empty, so you have a search page for example, and all of your parameters are indexed, or you have a dev instance that’s a duplication of the main site that’s indexed.
There are all these things where people did not need to have 100 million pages indexed and accidentally did and then they have issues. So cleaning all that up and putting things in place to prevent those from happening is one thing.
The second thing is internal links. Internal links are not a new concept, but the type of logic that works is not obvious. So, typically internal links are either based on relatedness or trending. So, if you have a site with a thousand articles and you show recent posts, then it’s going to be the same recent posts on every page, and the other 990 posts don’t get links, you tell Google that they’re not essential and they don’t get traffic.
Trending, the same type of issue. Relatedness, usually it also skews towards, you know, an 80/20 or 90/10 of URLs, so you have this very skewed internal link architecture. Flattening that, we’ve seen consistently, had a huge impact. And this is even on sites with pretty good internal links where if you flatten that and find some pages that are not getting links and adding links then those tend to do well.
Kieran: For internal links, is that just ensuring that 90% of your content isn’t buried deep somewhere and there’s no architectural support to get internal links to that content? Are you an advocate for all links being distributed equally or do you try to weight them towards the most important content you want to rank?
Ethan: I would say both, but there’s a balance between links are for users and links are for bots. Two different goals. For bots, it’s more important to have flatness in your internal link architecture. For users, it’s more important to be relevant. I’ve seen some evidence that Google tracks which links are clicked on and engagement on links. I guess that they’ll do that more and more moving forward.
The other thing is that it’s not about hierarchy or links from the homepage. It’s about reducing the number of hops from a crawl point. So a crawl point could be the homepage for sure, but it could be your login page, you about page, your team page, there’s all kinds of different crawl points that have a lot of authority via links, shares, things like that, and it’s about reducing the number of hops for any important page from those crawl points.
Scott: So, Ethan, one thing you mentioned earlier is that you could grow ten years ago solely from SEO, but not today, could you talk us through that a little bit more? Why is that?
Ethan: Well, in short, the reason why is because that’s what Google has decided but Google has gotten better and better. They have never wanted sites that get all their traffic via SEO because you’re probably not creating real value. So they don’t like that. And so, over the years, they’ve tried more and more algorithmic solutions for that, and they’ve gotten a lot better about that.
But, you know, in short, Google wants brand names, recognizable brands, brands that people like and have an affinity for, they have a bunch of signals to track that. One of those is, do you have traction from any other channel? So they’ve built their algorithm specifically to monitor for things such as that, and so that’s why it’s important to be able to grow the other channels.
Kieran: Ethan just really quick. Are you saying that Google looks at challenge diversification like that’s a sign that this is a real authoritative site and that this is a site that we should surface up content from more regularly because they have better traffic from a diverse set of channels?
Ethan: I don’t know for sure all the inputs. I know the spirit of what they want. What they want are big brands and brands that are not SEO plays. I don’t know how they’ve instrumented their algorithm to measure that. I have seen loose correlations with things such as shares, links, you know, all the standard things, whether not you have other channels with traction. I have seen correlations there. I can’t tell you for sure the exact instrumentation of the algorithm, but if you generally go for increasing your brand and getting traction on other channels, that tends to work well.
Scott: Speaking of the brand. Your brand and your product. Do you have any experience where SEO has influenced the product roadmap at companies that you worked at? Like if you could share any specific examples of product decisions that were made to boost traffic. And just in general, like how can search or organic traffic influence what you build as a product?
Ethan: I would generalize that to if you want to grow, your growth strategy should be integrated with the product strategy and not, you know, tacked on the side and in conflict. If it is, then it’s probably not going to be as successful. So, whatever your key growth levers are, should be baked into the core product.
The types of things that are important for SEO that is related to the core product are things such as your content strategy, your engagement strategy, your authority strategy. And then, are there constraints on content, are you building Symantec search or do you have used cars where people need to sign up? But all of those things are important. So which of those specific levers are essential for a particular site varies.
But you’ll see things like on Trip Advisor and Yelp, a massive part of their strategy is to grow reviews. In increasing reviews, they get a ton of traffic because that’s a signal and they built their product that way. Other examples, Thumbtack for sure, has developed their organic acquisition strategy so that you’re acquiring on a particular intent.
One of their big things is not just going broad for all everything local, but with high intent who wanna hire pros that are worth a decent margin. So all of their acquisition and search intent is based on acquiring that type of user and that type of intent, building landing pages to target that and then peppering them.
Kieran: How do you do keyword research for such large sites? How do you map keywords across websites with millions of pages?
Ethan: Definitely. I would up level from keyword to topic and then to search intent. So, keywords, I think, are too granular and too specific. Typically, you have a single topic, like so for MasterClass, Salmon is a topic, and there are multiples search queries that target that. So, there’s salmon recipe, how to cook salmon, cooking salmon, so there’s a one to many relationships between a topic, and a cluster of search queries.
Then there’s the intent. And it’s important, not just to have a page that ranks, but to have a page that matches that intent and then can convert people. The way that I typically do this is I will start with what’s the current set of content for a company and what’s the entire addressable market.
The entire addressable market for something like Ticketmaster could be something like tickets, and then it could also be lyrics. Then you map what’s the two by two for volume, addressable market, and search intent. Then progressively go through which buckets, you know, lyrics, high volume-low intent, tickets, medium volume-super high intents, like mapping out all those different search intents.
Within those search intents, there are specific topics. So for lyrics, you know, there’s a particular set of songs or artists that people want lyrics for. For tickets, a different set of topics, Taylor Swift tickets. I try to define where are we currently playing with our content, what is the addressable market, where are those gaps. Within those gaps, what do we think we can compete well on? What are the topics within that search intent and then what are the multiple keywords per topic prioritize by some volume competition opportunity?
But that’s kind of how I think about topics or search queries.
Kieran: Just two quick follow-ups. To make sure, so our listeners know, when you say addressable market… because I thought that was a really interesting way to go through from search intent to how you categorize content. Addressable market, you’re just using some sort research to look at keyword volume and then kind of aggregating it up to like to like a total addressable market number?
Ethan: Sort of. It’s a little bit more complicated than that, but it comes from external and internal data. So internal data is what you described, which is we’re a recipe site, here are all our recipes, list of ingredients, list of dishes, list of cuisines. So that’s internal data, and then you can take that and then expand that with any keyword tool. You might have some gaps.
So, maybe you didn’t know that diets were a thing because you just hadn’t built that in your taxonomy. So, how do you discover that? The way that you discover that is by identifying competitors and making sure that they’re not winning on some clusters of search intents, like nutrition facts, or things like that. So you do that based on competitor analysis, and you can do that by looking at SEM or SEO tools, find some competitors, get their keywords, and make sure that there aren’t any significant clusters that you’re missing on.
Kieran: Let’s get into the second part of this interview where we’re going to talk about content gaps, how you can get a gradual growth curve for SEO, start-up SEO, and just why is Ethan so good at search.
Kieran: Ethan, what are content gaps and how can they help us to compete in search against very large and established websites.
Ethan: Yeah. And it depends on the size of the company. So, if we’re talking about Wayfair, they can directly compete with Amazon. If we’re talking about a brand new commerce site, then they can’t directly compete with Amazon. So depending on what stage you are at as a company and how much authority you have, you should pick which level of competition to go after.
So for a brand new site, you would maybe want to go after a set of niches. The ways to do that are similar to what we discussed in the previous topic. Find all of the clusters of topics. And maybe Amazon has a few of the long tail search intents, but not some others. But basically you’re taking your set of queries or your seed terms, which is what we call them, and you’re expanding them and finding more related keywords and longer tail versions of that, and then you’re finding competitor data.
There are some interesting things you can around the grouping of topics that we worked on. But basically, you can group many small long tail phrases into single groups. Like we talked about topics versus search queries. I think that there’s a lot of interesting opportunities if you correctly group things.
If you look at the keyword level, you’ll see zero search volume, and you won’t compete for it at all. But if you can be smart about group long tail phrases into topics, then you can be more thoughtful about finding these topics that have volume, but people are falsely not competing for because they didn’t have enough data to say that they have volume.
Kieran: That’s cool.
Scott: So, we’ve talked a lot about search, which makes sense because you’ve worked with companies like Yummly and Thumbtack and MasterClass. What about some other channels, it seems like the growth would come from search because they’re such large websites. Is that true? Have you had much success diversifying away from search and getting growth from other channels or areas?
Ethan: Yeah. I would say two things; the first thing is I think it’s always best to grow on at least one other channel other than SEO to start. So, that you have some traction. The second is that you should never be reliant on any single channel, whether it be SEO or paid or social whatever it is because these things change and you don’t want to be fragile to all the sudden everything is gone, and your company is going under because it was all SEO.
So, I always recommend starting with a channel other than SEO to get some traction and then diversify. Decide which channels have the most opportunity and then go after them. They should; ideally, all integrate with each other. So SEO can help with paid and vice versa. Same with social, like the more organic users that you get via search, the more shares you can get on social, the more shares you get, the better you can rank on SEO. So they should all be nicely integrated with each other.
Kieran: And something else we talked about, and we’re going to link to it in the show, know that you did a great interview with First Round, where you spoke of a gradual growth curve for search. I think this episode is where I air a lot of my dirty laundry around SEO. But, again back in the day, I owned… because I had a lot of learning from doing things wrong. I built my link network from dropped GoDaddy names and setting up different C-Class IP addresses and interlinking them all.
Ethan: We called them a community, but yeah-
Kieran: Yeah! I called them my social network before there was Facebook. And I had these enormous spikes of traffic for all these different sites I was managing, and then it got torn apart by Google, and everything got penalized. So that’s why I love that you talk about gradual growth curve for search.
Can you talk a little bit just about that concept of what you meant by a gradual growth curve and what search tactics do you think companies should invest in to optimize for that kind of curve?
Ethan: Yeah. So, there’s nothing wrong with fast growth. If you have an amazing site with an amazing strategy, there’s nothing wrong necessarily with growing fast. The problem is that there’s a lot of spammy sites, and there are characteristics of those spammy sites that in and of themselves are not illegitimate but highly correlate with illegitimate things. So, massive growth in a week correlates heavily with spam. Doesn’t mean everyone who does that is spam, but it highly correlates.
If you think about what are the highest correlating patterns for spam, you should avoid those. One of them is growing way to fast. And we’ve seen that quite a bit where all of a sudden, a hundred million of your pages are indexed, and even if they’re good pages it tends to spook the algorithm and introduce risk. So what we typically recommend is to go a little bit more slowly. One other thing is that as you’re launching pages, they should acquire traffic if they’re not acquiring traffic then that’s also not good. So if 90% of your pages are empty or have bad engagement, that’s a problem.
So, if you take a more iterative peace meal approach where you’re going through and launching individual page types, one at a time, chunks of pages. If you have ten million products, launching a thousand at a time. Seeing which ones work, removing the ones that don’t, keeping the ones that do. Iteratively going through that then you can avoid spooking the algorithm without huge anomalies. And you can make sure that you’re maintaining a healthy index of pages that perform well.
Scott: Another thing you mentioned in the First Round article is highlighting how some of the least sexy parts of SEO, so to speak, are sometimes the most impactful. So in particular, you were talking about how to crawl your website and track those metrics. Just regarding working on good hygiene for those type of websites, what kind of difference has that made in the overall growth of search traffic?
Ethan: There’s a long list of things that you should not do. You can introduce all the checks to find those. Because, if you don’t then probably things will silently fail. So adding more and more rules to check for bad things, is a good thing. Part of that is tracking crawl, indexation, things like that to see do you have ten million empty pages or do you have a bunch of pages with no links or do you have, you know, things that fail on this rule that we’ve encoded.
These are all things that cause traffic drops or stalls, so it’s really a question of how serious is the problem and when you fix it, then you get back to your baseline. I have seen hygiene related issues where removing bad pages have, you know, had a significant impact on search traffic.
Kieran: Is there anything that works great for increasing traffic to large websites that’s surprising to you?
Ethan: I’m still surprised at things like removal of pages is a thing. You would think Google’s smart so if we shouldn’t have a bunch of empty pages then they’ll not index those, and it won’t be an issue, and it is still an issue. So, probably I would say that things like that and the fact that they still work in the magnitude, same with internal links, that always surprises me.
Kieran: We’ve talked a lot about doing SEO for large websites where you probably have a bunch of resources to do a lot of things, like developers, designers, etc.
If I’m a smaller company with limited resources, what should they begin with when planning to increase their search traffic.
Ethan: Make sure that you have traction on another channel. Find somebody else who’s doing it well and copy them.
So, there are a few different strategies in SEO. There’s big site SEO with 10 million pages; there’s article SEO, there’s enterprise SEO, completely different strategies. But my recommendation for any new startup is to find some benchmarks that are doing well, either directly in your space or related space, and try to figure out what’s working and then copy them. And depending on the site, who to copy varies, but that’s my recommendation.
Kieran: Yeah, we’ll link to an article where you give an example, I think it was from Trip Advisor, where you provide an example of how you can do that. Do you think small companies should work with a recommended consultant? You think that can get them a lot of leverage in the kind of early days or do you think they need to have someone full time, in-house.
Ethan: It’s always better to have someone in-house, but it’s tough to find good SEO people so try to find someone for as long as it takes. And in the meantime, get somebody else to help you. Consultants vary regarding quality, but if you can’t find an excellent in-house person, certainly it’s better to have a consultant than just freewheeling it.
Scott: When you just said it’s tough to find good SEO people, what are those things that are missing? You know, when you’re looking for an SEO person, what is that rare set of skills, or maybe particular couple skills that are out there that is hard to find? Like why is it so difficult to find good SEO people?
Ethan: So, who to hire depends on your category. For an enterprise site, a good SEO manager is going to be very different from an SEO manager at Thumbtack. For something like enterprise, it would be about creating ten product marketing pages and building authority versus someone like Ticketmaster or Thumbtack where its more about launching thousands and thousands of pages and how to manage those.
So, I would first narrow in on exactly what your strategy is because then you can be a lot more targeted in who you find. Again, for something like an enterprise, it might be easier to find someone to do that because it’s about content authority. But, defining your strategy is essential.
And then a lot of things that work well for SEO, it’s not necessarily hard to find someone to do each task, but someone who can look at all of those tasks and combine them into a cohesive strategy is a little bit more unusual because something like engagement and authority is very different. Those skills are very different from content, which is very different from technical SEO. So, making sure that all of these silos are nicely integrated into a cohesive strategy is the thing that is a little bit hard to find.
Kieran: How do you think search professionals are going to need to adapt to mobile and voice in the future?
Ethan: Yeah. I think mobile and voice is extremely interesting and also scary because they tend towards winner take all where you’re either position one, or you’re not getting any traffic. And if you have an answer on voice and you are number one that doesn’t mean that you’re converting at all, you might stay on Google Home and never leave.
And I don’t know what will happen. The things I think are essential are what are the characteristics for who is that single answer and then how do you get people to convert. And it seems as though Google is more and more pushing towards this question and answer style framing of results for searches so figuring out how to position your content such that it answers a question, such that it’s in the right format, that it’s selected as that winner an then figuring out to get someone who heard that answer on their phone to actually come and buy something or convert on your product. Those are the things that have not been figured out yet, and I’m very interested in figuring out, and I think are vital to performing well on voice and mobile.
Kieran: Did you invest much to format your content for voice when you worked at Yummly? I can imagine that when people are searching for recipes, they’re probably looking for a single recipe vs. a lot of different versions.
Ethan: Sort of, I would say that most things that work in voice are not specific to voice per se. It’s more about the thing that I just mentioned, which is what is the question that the user’s trying to answer, how do I position my content to be an answer to that question, how do I format it so that Google thinks it’s the best answer to that question, and then how do I get them to convert. But none of those are unique to voice specifically, and we don’t need to do our own voice technology or like a special voice strategy. It’s more just a general strategy of Google’s moving towards winner take all, how do we make sure that we’re that winner.
Kieran: So it’s like content invested in the format of your content.
Scott: So, it’s no secret that you’ve had a pretty impressive and successful career so far to date. Like you’re a growth advisor for companies like MasterClass and Thumbtack and Yummly and Ticketmaster and Honey and like a bunch of other ones. What are some of the skills you’d say you’ve mastered that have helped you get to where you’re at today?
Ethan: There’s a few. I think probably the number one is a level of irreverence and not caring about what is defined as best practices or what is trending, like voice search and apple watch, and you know. Talking of Googles launching next week and you’re, you know, get special access, and you get to be featured so let’s do that. So like just not caring about any of that stuff and purely caring about results. Which will then lead you to work on internal links and crawl and those supposedly boring things that are impactful? But just caring about results and nothing else is very freeing and it lets you focus on the things that are important.
The second, I think is some level of analysis and ability to analyze well. Many people analyze, but there are key things that often are not done correctly like confusing correlation with causation, and selection bias. And people who do this weird thing on Spotify are more engaged and making sure that you ask is that because it’s selection bias or just because that the features helpful. But being thoughtful and analyzing data the right way, I think is also super important.
And probably the last is in school you’re taught to do your own work and never to ask somebody else for the answer and if you’re working you can get a lot of answers to questions without having to figure them out yourself. So being very aggressive about talking to other people and asking them what they did, and what worked and what didn’t work, and leveraging that has also been super helpful.
Scott: That’s awesome. Yeah. When you were saying, you know, a lot of it is not just necessarily chasing whatever the latest update is from Google or whatever, and it’s more just getting down to practicing the fundamentals of what works. Such as content pruning, and scraping, and all of those things you were referencing earlier.
I just immediately had an image of UCLA’s basketball coach, John Wooden, pop into my mind, and so I’m like okay Ethan is the John Wooden of SEO where its like you just need to practice the fundamentals, it’s not necessarily throwing up ally oops and all of the shiny things that come by. You have to be good at getting the fundamentals down and being very tight on them.
Ethan: Perfect analogy. Thank you.
Kieran: The other thing I think to re-iterate the importance of is because you mentioned it there and you had it in your article where most challenges and problems that were trying to solve, someone out there has tried to solve something similar. So being pro-active about trying to find those people and have good conversations I think has helped me a lot and does help a lot. So again, we’ll link to that article where you give an example of how you did that in the show notes.
Ethan: Awesome. Thank you!
Scott: Our last question here is regarding you were talking about talking to other SEO managers and getting inspiration from them. Who are some people you look up to as you’re developing your SEO strategies, certain people, certain companies?
Ethan: It’s usually not specific people, it’s generally for whatever I’m interested in right now, who’s doing it well. So, probably whoever’s doing voice search well is… There’s going to be some new people there, they developed an interesting strategy so let’s go find that out. So it’s more about when we identify an opportunity who are the people who seem to be executing it well, and that changes.
But the companies that do SEO well are Trip Advisor, for sure is like the best of all time. Pinterest is excellent. And then usually there are people who are very good at a particular aspect like Thumbtack is good at conversion and content. Apartment List is excellent. Ticketmaster’s pretty good, as well. Honey is awesome. We are doing very cool things there. Usually, it’s more ad hoc and for something where we think there’s opportunity, who’s doing it well today and let’s figure out who built those things and asked them what happened.
Kieran: Awesome, thanks for joining us on the GrowthTLDR Ethan.
>> Google Play