Swiftype: The Biggest Search Engine You’ve Never Heard Of – with Quin Hoxie

Posted on Aug 5, 2013 - 9:00 AM PST

You’re about to meet the founder of the biggest search engine you never heard of.

And the whole time you’ll hear his story, I bet you’ll ask yourself, “why would anyone pay for his search engine when there’s search giant offering doing good search for free?”

Quin Hoxie is the founder of Swiftype, which is the easiest way to add great search to your website or mobile app.

Watch the FULL program


About Quin Hoxie

Quin Hoxie is the founder of Swiftype, which is the easiest way to add great search to your website or mobile app.

Raw transcript


Mixergy’s audio transcription is done by Speechpad

Andrew: Listen up. I hate to have commercials interrupt this interview so I’m going to tell you about three sponsors quickly now and then we’re going to go right into the program.

Starting with Walker Corporate Law. If you need a lawyer who understands the startup world and the tech community, I want you to go to walkercorporatelaw.com.

Next, I want to tell you about Shopify. When your friend asks you, “How can I sell something online.” I want you to send them to Shopify and explain to them that Shopify stores are easy to set up, they increase sales and they’ll make your friends products look great. Shopify.

Finally I want to tell you about Grasshopper. Do you want a phone number that people can call and then press one for sales, two for tech support, et cetera and have all of the calls be routed to the right person’s cell phone? Well, get your number from grasshopper.com.

All right, let’s get started.

Hey there freedom fighter. My name is Andrew Warner, and I am the founder of Mixergy.com home of the ambitious upstart. Now there aren’t too many more ambitious projects than the one you’re about to hear about today.

You’re about to meet the founder of the biggest search engine that you never heard of and the whole time we talk somewhere in the back of your head you’re going to be saying, “How is anyone going to be using, let alone, paying for this search engine when there’s a giant offering free search that’s good. How is this guy even going to compete and why is Andrew so excited about him that he’s going to have him here on Mixergy?”

Well you’re going to hear the answers to both of those in more detail but first, the short answer is it’s a good search in a way that the giants aren’t doing right now.

Quin Hoxie is the founder of Swiftype which is the easy way to add great search to your website or mobile app. Quin, welcome.

Quin: Thanks for having me.

Andrew: Alright. How many searches are you guys doing right now?

Quin: Right now, monthly we’re at about 120 million searches.

Andrew: 120 million? Where’s Google?

Quin: Google, that’s tough to say as far as their global market. They’re definitely bigger than that. They’re well into the billions at this point and with their site search market, there isn’t a lot of info out there about how many they’re doing there.

Andrew: Okay. And the difference is that, well, not the difference but what you’re focusing on is what?

Quin: We feel like there’s a huge kind of neglected piece of the market which internal search on websites or in mobile apps so rather than a global search where you’re trying to find where something exists in the location that you want to be at and you want to just find a piece of content, find a product something like that.

It could be a store, it could be a publisher anything like that we want to help you find it within that website.

Andrew: Why can’t people just go to Google? Google has search that works within the site and just install the Google search product.

Quin: The issue with Google I think has been what their focus is so obviously Google is a global search product and their local searches is filtering their global search down to your website.

If you’ve ever done a site colon, mixergy.com search on Google, that’s kind of what your getting on your website and the motivations for a local search engine internal to the website versus a global one are very different.

Externally, Google’s having to fight spam and people trying to gain with SEO and things like that whereas with your internal search, you know what’s important. You own the content. You don’t need to gain anything, you shouldn’t have to be playing tricks to try and get things to work properly.

You should have full control and you get full control over other parts of your website. If you use WordPress it’s a content management system, you have full control over all aspects but with search everyone’s kind of just left in this black box.

If you install Google you have minimal control and it’s nice because it’s quick and easy to get it installed and you get good results like you expect with Google and they’re fast. But there’s a lot of good information in search.

It’s this box where people are coming to your website and typing whatever they’re intentions are. To be able to capitalize on that, not only returning them good results but also being able to teach the site owner what are people looking for or what are they not finding that you should have. That’s something, a need we’re really trying to fill.

Andrew: I see, right, because we want to be able to see what we as site owners want to be able to see what people search for often on our site, and based on that create content or find products to sell them.

Quin: Absolutely.

Andrew: Okay. And we also want to be able to say, this page doesn’t rank well on Google for the key term that people use, but it’s important to our users. And it needs to be the top result. And we can manipulate the search results when…

Quin: …Exactly…

Andrew: …we’re in Swiftype.

Quin: And there’s no reason you shouldn’t be able to do that. Again, it’s your search engine. If somebody searches for a particular company name or something like that on Mixergy you should be able to tell it exactly what the results should be for that.

Andrew: Right. Hey, can you tilt your camera up a little bit. I feel like we’re cutting off the top of your head, and I should’ve…

Quin: …Yeah, sure…

Andrew: …worked on that before we started. There it is. Do you freak out, though, about Google, about Amazon, about Bing?

Quin: You know, we definitely see that Google has some competitive offerings out there. There aren’t a lot of companies that would be as scary to go up against as Google. But we don’t see ourselves as being in direct competition with them at this time. They have a few offerings as far as search specific to e-commerce, search specific to just internal search of websites, but in a lot of ways what we’re trying to build is something that hasn’t quite been done before. I think it’s due to a lot of how technology’s evolved and how content creation on the web has evolved. It’s a new need that we’re trying to fill for people.

Amazon, obviously, has fantastic search on their website, and they have a great team of people there working on it. That’s been their motivation. But I think as far as who our competition is it actually ends up being, I’d say our biggest competition at this point, people wanting to build search internally.

Andrew: Internally.

Quin: The engineer on the other end who’s telling their manager, oh I could actually do this, I want to build this, I’ve heard of some cool technologies I could use. That’s kind of been what we’ve been up against in the big ticket customer cases.

Andrew: I want to learn how you got here, how you got investors and advisers like Paul Graham to support you and believe in you, and not say, oh Google’s already locked this problem down…

Quin: …Sure…

Andrew: …just move on. But one more establishing question before we get into your story, and that’s how much money are you making with this business.

Quin: We’re making a good chunk of money right now. We just very recently released our paid accounts, and it’s been growing pretty quickly. We had a beta period where we brought in a lot of free questions and got to gather, you know, what do people actually want in this case. We had a lot of assumptions about that, but bringing in all those people and learning about it has been very valuable. We just started charging people in the last few months. Growth of revenue has been great. We haven’t released those numbers yet, but it’s been really solid for us to this point. We’re very appreciative of the people who’ve come on board.

Andrew: Okay. You and I talked on June 27, 2013 at 4:19 p.m., and like I often do quietly in private conversations I took notes. But I wrote private next to everything. The number, you won’t allow me to say the revenue. When entrepreneurs say they don’t want to reveal the number I respect that. I ask questions, but I respect that the answer’s sometimes no.

Quin: Sure.

Andrew: What about the number of sites that are running Swiftype search engines. I’ve got that number. Do you feel comfortable sharing it?

Quin: Sure, yeah, that’s a number we’re very proud of. That number today is actually very different than the number I told you about a month ago. That number is just approaching the 70,000 mark.

Andrew: So, 70,000 different sites have decided that they want to use your software.

Quin: Yeah. Those are 70,000 separate search engines that people have created on our platform.

Andrew: Okay, alright. And, full disclosure, Mixergy is one of them. Maybe we can talk about that later on…

Quin: …That’s right…

Andrew: …in the interview. I’m curious to see what from your point of view Mixergy is like as a customer, because I always want us to be…

Quin: …Sure…

Andrew: …supportive of startups. And that means be their best users or one of their best customers. I’m hoping we’re like that with you.

Quin: Absolutely.

Andrew: This whole thing started when you got walked into Scribd’s offices late at night. And what did you see there?

Quin: After I interviewed with them I was here over Thanksgiving weekend to talk to those guys. I wasn’t really looking for a job at the time. I really liked my job. I was living in Portland, Oregon. After my interview, it was a long day of interviews, I walked out. It was maybe 8:00 p.m., and everybody was in the exact same place they had been when I walked into the room. So, I just went back and started talking to the guys who’d interviewed me. They were drawing stuff out on a whiteboard laid out on the floor. I walked back there and started chatting with them. We were there until well after 10:00 p.m. when they asked if I would like to go get some food in The Mission and keeping talking about this stuff.

And so, we went to The Mission and kept talking about this stuff. It’s just kind of eye opening because it’s a sort of something that I did on my own. When I got home from work and just work on things on my own. But everybody was there late into the night, working on these really interesting things. It kind of clicked at that point that this was the sort of thing I wanted to be doing.

Andrew: Most people would see a team of people in an office at 10 p.m. and say, “This is not for me. I want my salary, my life, and something challenging. But… I’m not building the business for myself, why should I want to stay until 10 p.m? Why do I want to hang around a bunch of guys that want to stay until 10 p.m.? What was it about it that attracted you?

Quin: Sure and I totally respect that. But it is…definitely the kind of person…I have found myself to be is that…there’s pretty much nothing more motivating than working on an interesting problem with smart people.

Andrew: Even if you don’t own it because there was no way that [???],at that point, was already built and successful. You can’t be a co-founder! Why…be so excited to work that late on something that’s not yours?

Quin: I think that if you are given the right amount of ownership, in terms of directing a product, feature, or things like that. It can be just as rewarding in a lot of cases. I mean, the reality of this is…especially out here, engineers can go get..a great salary. You can walk through Soma [SP] and probably knock on a door and have a great salary in a matter of days. You have to prioritize other things and to me, top priority is people for sure.

Andrew: I see.

Quin: And that’s something Scribd is just remarkable for. The team of engineering that they have there are just fantastic. That’s why I had such a fantastic experience and honestly that’s what made me decide to move to the Bay Area. The work of those people that I got to hang out with over Thanksgiving!

Andrew: That’s what got me to move out here. Scribd is on-line document sharing. Before you worked at Scribd, you were a developer at About Us. Isn’t About Us just a scrapper that does SEO optimization?

Quin: So…it’s very easy to bucket it that way. But it’s actually a very pretty cool company in its own right. I went to work for About Us after school and…It was a small company and I didn’t know about them but I went up and met the team.

They were in a similar situation, they were a lot more traditional but it was a small company, strong team and just a very familial atmosphere. They actually had really interesting problems. There are a handful of websites on the Internet that end up having a huge amount of traffic but nobody knows that they’re there all the time. And so, they were kind of one of those interesting cases where they had tons of data, lots of interesting problems, and they just needed engineers to work on it. It was a pretty fun team to work on and it was a good place to start after school for sure.

Andrew: Okay. They are big on SEO?

Quin: Yeah. Scribd and About Us have definitely formed a good portion of my career to this point. I worked on crawlers at About Us, worked on search at Scribd, and SEO. In both cases, that informed us that we can do that stuff on Swifttype. Matt and I are both fortunate enough to have pretty rich backgrounds in those respects.

Andrew: Matt Riley, your co-founder?

Quin: Yeah.

Andrew: What did you learn there that you brought to Swifttype?

Quin: Basically at About Us, taught me about web crawling which is that there may be no more difficult problems on the internet than web crawling. It’s effectively unsolvable and one of the things that set Google apart, especially in their early days. The reality is that the web is just a real mess of content. There are a lot of things that are broken and kind of always be that way. You’re always going to run into weird stuff out there.

Andrew: What is the problem with web crawling? It seems like it was one of the first things that we did on-line.

Speaker Quin: Well, it’s…definitely one of those problems where it’s easy to get a working solution. Then going from a working solution to a good solution takes you about a hundred times that long. One of the things is content extraction, actually taking a page and figuring what the real content is on it. You don’t want to pull in the navigation links so if somebody searches “contact”, every simple page shows up, that sort of thing.

Andrew: I see.

Quin: It’s an interesting problem and it’s one that you can make incremental improvements on. It ends up being an actually a pretty fun problem to work on because it’s very testable. You have very directed challenges. You see that navigation in the content and you’re like, “How can I solve that and not break it for every other site?” That sort of thing.

Andrew: All right. I get it. We use it. At Mixergy, obviously, when we do a search for a company, AboutUs.org pages come up all the time.

Quin: Sure.

Andrew: In the sense that it’s who is data that I don’t have to add captcha to see, it’s really helpful.

Quin: Sure.

Andrew: And sometimes there’s extra data too that’s useful.

Quin: Yeah. It’s one of those things when I first saw it, I was, like, oh, I remember this site. I’ve seen them in my Google results. One of the things that AboutUs really showed me, and Scribd as well, is that sometimes get a little bit of a bad rap because people that they’re not doing anything terribly innovative. The reality is that most of the companies out there are. They’ve actually got smart people who are building some really cool things behind them. So it definitely kind of turned my perspective on that for sure.

Andrew: You probably just saw me grab a note pad and pen because it just occurred to me to ask later on in the interview where it’s more appropriate. This is the first time that I’ve ever seen your face. We’ve worked together. We’ve chatted via email. I’ve never seen your face. I’ve don’t see you on the blogs that I read.

Quin: [laughs].

Andrew: I don’t see you at events. You’re nobody. I’m sorry to say it that way. But you’re nobody and still you got tens of thousands of sites to trust you and to know that you even freakin’ exist and to put . . .

Quin: [laughs]

Andrew: . . . you on the site, instead of just saying, oh, wordpress already [??]. I’ve got to ask you about that, and I hope when we get to the section you won’t just say word of mouth, because I really need to understand how you get so many people to deeply imbed you into their site.

Quin: Absolutely.

Andrew: Good. I’m glad that you’re saying yes because I’ve got to find that out.

Quin: Mm-hmm.

Andrew: Well, let’s continue now with the story. So you’re working at Scribd. Things are going well. And then something comes along that moves you towards this problem and towards starting your own business. What was that?

Quin: So Scribd had kind of been fighting with getting a good search solution for basically the entire time I was there. They had used Google for a while. They had redone their Google implementation. They had tried out Yahoo. They had evaluated some larger enterprise options. And then they had started building on open source and the first couple of tries at that didn’t work very well. I was working on my co-founder, Matt’s, team at the time, and eventually it came to his team. Around the time it did, Matt got moved into the head of product role, and I ended up taking over his team. So all of a sudden, search was basically . . .

Andrew: I see.

Quin: . . . on my radar as the main thing I needed to solve. And I was kind of excited about it. I think search is a cool problem, because a lot of it is based on this really amazing research that has been done over the years. It’s come a really long way, and then technology has kind of started to catch up with that research. But the thing that really has fallen behind is that search is actually a product problem. You need a quality product on top of this technology for it to be good for the users. And if you don’t have that, then the research is kind of meaningless in terms of how people are using it.

Andrew: I’m sorry. First of all, I’ve got to get out of my own head and just say that . . .

Quin: Sure.

Andrew: I feel bad that I called you a nobody. I’m thinking, persons in the audience must think I am a jerk for saying that.

Quin: [laughs]

Andrew: I mean in the sense that we don’t know you. Not, obviously, that you’re a nobody in search. You’re one of the top people in this space. But how does KISSmetrics know about you? How do we find out? We’ll get into that later.

Quin: Sure.

Andrew: I just had to get it out of my head so that I don’t feel guilty and obsess about it instead of paying attention to what you’re saying.

Quin: [laughs,] Oh, no. That doesn’t bother me.

Andrew: When you were at Scribd and you see a problem. Help me understand it in a more concrete way with an example of what Yahoo didn’t do with its search for Scribd, what Google didn’t do with its search. Can you give me a specific example that would help me understand the problem that you saw?

Quin: Yeah, sure. So one simple example for this is that there are all these really innovative search systems that are really good at taking one piece of text, and a ton of pieces of text, and then matching them together and saying, these ones match the little piece of text, the query that you sent us.

Andrew: Mm-hmm.

Quin: But one of the things they don’t do is they don’t have any mechanism to actually get the text in there. So everyone has all their text in someplace, in a WordPress database, something like that. And just getting to the point where you can even ask this really amazing technology to give you results is a whole hurdle that most people can’t overcome.

Andrew: Google does that if you implement Google site search.

Quin: They kind of do. If you open up Google site search, again, they’re using their global index for it, and you don’t have any control really over when your content’s getting indexed. So we had a number of customers who have switched to us who switched solely based on the fact that they would wait months, and certain things wouldn’t ever be indexed in Google. And they didn’t have any way of touching it.

Andrew: I see. And if they send Scribd lots of content, that’s a huge problem.

Quin: That was actually one of the biggest issues with Scribd was that there’s so much content, like tens of millions of pieces of content on Scribd and they didn’t have any way of really having granular control over things getting actually indexed in the search engine.

Andrew: I see. OK, now I’ve got a sense of the problem that’s there. Now it’s on your shoulders. How did you attempt to solve it?

Quin: The first step was that I had a really good team, which definitely helped. What we wanted to do is basically step back and figure out what had gone wrong before. I think that the problem was initially blown up into this grandiose piece that was almost too difficult to complete.

We stepped back and implemented it in small pieces to get something working every step of the way. What we ended up with was really good, but it took us a long time. Everyone was very happy with it. Once it came out and all the results were better, everybody was happy.

What ended up moving me towards saying, “Wow, there’s actually a lot more to be done here,” was after we had better results just better relevance in general. We had people from the content and marketing teams coming back and saying, “Hey, everybody loves search now, but we just want to partnership with this publisher, and when I do this search, their document is in the fourth position. Why isn’t it in the first position?”

For me as an engineer I was well, the search doesn’t know that we just formed a partnership with that person. The reality is that the engineers don’t want to be tweaking those things. I was just doing one off solutions for them. That should be in the hands of the people who care about it. You should be able to give tools to the content people and the marketing people to manipulate the results in the ways that make sense for them and for whatever the state of the company is.

That’s what motivated me and Matt to say, “What could be done here? Why doesn’t this product exist?”

Andrew: It’s similar to the issue that we had that we needed to solve. It happened just a few weeks ago. Someone said, “I can’t find your How to Interview course. I’m searching like crazy.” All I come up with is these I guess it’s interviews where I mentioned interview your heroes and my program on how to interview. He complained. Usually, I’d say, “Well, just scroll to the bottom or here’s the link.”

Because we use you guys, I was able to go into our control panel, or maybe Bob Hyler did, the guys whose idea was for us to work with you. He just moved it right to the top, and everything was golden. Now future people are able to find the course that they want, and not just the most popular interview where I happened to mention it. It’s incredibly helpful.

You’re working here. Is there any issue about you taking what you just built for Scribd, what you learned at Scribd, and creating your own company? Isn’t there an ethical issue there?

Quin: The solution that we built today is 100 percent different technologically than what we build at Scribd. That’s partially due to how technology is involved since then. It’s also due to the solution we were building was generic, whereas for Scribd it was for a very specific purpose. As far as that goes no, there’s no issue there. It’s different stuff, and the founders; we have gotten their blessing on obviously moving on to this.

Andrew You told Jeremy Weiss in the pre-interview that your market research involved going to Quora. What were you doing on Quora?

Quin: It was one of the things we just saw this organically, like wow, do we know the things on Quora or Stock Overflow or things like that? We’d always see people started to get accustomed to these really rich search experiences that certain sites had. Amazon, Facebook, Yelp. They’re some of the few sites that you would go to and know that you could search and get good results.

Very few other sites exist like that out there. On Quora there are all kinds of questions for people. Hey, how do I build search like Facebook’s? How do I build search where I start to type something and just in a matter of milliseconds, all the results are displayed that I want instantly?

What’s amazing about Quora is that you have two engineers who worked on that at Facebook come in and say, “Hey, actually, we tried to use some stuff that was open source, and we ended up having to build it all internally, because nothing out there could do that.” That’s the sort of thing where we’re looking at it, and we’re why doesn’t that exist? There’s no reason except that it’s hard.

It ended up being something we worked on at Scribd, and we’re just people want this. For us initially, Matt and I are both engineers. We had a good connection there as far as what engineers would want. Then it was just a matter of turning it into a product.

Andrew: The way you discovered the pain was by working at Scribd and seeing the problem first hand. Scribd has incredible connections in the valley and if none of those connections led to a solution, then you know that this is a serious problem. You then, if I’m understanding you right, validated it by looking online to see other people complaining. Then you realize, okay this isn’t just unique to huge sites like Scribd. It’s something that many others have.

Quin: Right. Absolutely.

Andrew: You worked on this problem at Scribd until 10:00 p.m. and get up the next morning and be at the office at 4:00 a.m.?

Quin: Not at 4:00 a.m. Definitely not. We got in a little bit late to work at Scribd. We lived in the same building and we would work until 10:00, 11:00, sometimes even later at Scribd. Then we would go home and we basically didn’t have anything else to do so we would just continue working.

Andrew: Until 4:00 a.m.

Quin: Matt just lived a floor above me. Yeah, we would work until 4:00 a.m. I would get to work at around 10:00

Andrew: I see. Dude you’re like the ideal developer employee. You’re taking your work home with you and you’re up until 4:00 a.m.

Quin: It’s hard not to sometimes.

Andrew: Makes sense. That’s the other thing that I noticed about San Francisco. The nightlife is pretty crappy here. It’s not like you’re missing out on something so much fun that you got to get out of the house to go see.

Quin: Sure.

Andrew: I see where this is going. Let me ask one other question about where you are up until this point. One of the reasons why people work for Y Combinator backed companies, is that they get in with the crowd. They get to meet other entrepreneurs that Y Combinator back. They get to be part of this click. They get to maybe even meet some of the partners at Y Combinator like Paul Graham. Do you get any of that as an engineer for Scribd?

Quin: A little bit, for sure. I had actually, Matt and I both I believe Matt did too, applied to YC way back when I was 17, 18 years old, something like that, and didn’t get in. It had always been kind of like this, I had always followed Paul Graham’s essays, it had always been kind of this amazing thing that seemed so distant to me. So I always followed Scribd for that reason. I thought they were an interesting company, partially based on that. Then getting there, I think part of it for me was that it made it a lot more real. I grew up in Arizona. I went to school in Arizona and then moved to Portland.

The whole YC circle and even just the Bay area in general was kind of like this fantastical thing to me. Until I got here and working at Scribd, it just kind of put everything in perspective. Meeting the founders there I was like, “These are really smart guys who’ve done amazing things, but they’re also just guys who are dedicated. They’re working hard.” They’ve got good ideas, but it kind of made it real for me. It made it something that isn’t completely out of reach which is a very new thing when that happened.

Andrew: That’s one of the things that I like about what Y Combinator’s doing and frankly even some of the sites that people love to hate like TechCrunch. The do make entrepreneurs into celebrities. The do make them into aspirational heroes. When you’re finding out about Alexis Ohanian who also read Paul Graham’s essays when he was in school and how he took a train to go meet Paul Graham and got investment from him.

Then they built their idea and turned into Reddit and so on. We’re reading them. We’re inspiring to be like them and I don’t think that that’s normal. I do just the other day remember reading some article. I think it was on the Daily Mail or some mainstream newspaper’s website. Where on the right margin the most popular article was Snooki is now so skinny that even skinny jeans look big on her. That’s who most people are paying attention to.

And I like that we are paying attention to, and I include Mixergy in that. We make aspirational figures out of entrepreneurs. Then when other people read about them and want to be part of the world, the other cool thing is it’s easy to gain entry. It’s not like you’re at the door like a shaolin want to be monk who has to sit at the door for months. You’re there. You’re here. You’re doing the work at its meritocracy. In this meritocracy, you didn’t have, actually was that first idea that you brought to Y Combinator so good that you now think, “Damn it. Those people missed this great opportunity.”

Quin: Absolutely no.

Andrew: What was that idea? You think they’re right to have said no.

Quin: It was some website builder sort of tool. It was a small idea for sure.

Andrew: I see. And this is a much bigger idea.

Quin: Yes.

Andrew: How do you go from working for this great company that you’re willing to stay up until 4:00 in the morning on your free time and help out to now deciding that it’s time to create your own business?

Quin: In a lot of ways I think it was a long time coming. In the way Matt and I worked and the way we worked together and the way we just kind of viewed things, it was the sort of thing where everything was a new idea. It was always like, “How can we do this better?” and “What can solve this problem?” When you’re staying up until 4:00 in the morning working on stuff, inevitably things come up.

Part of that, making it a reality for me or making it more within reach, was that there was never going to be a good time to do it. I was always going to have a good job. I was very fortunate in that respect. So I just needed to kind of do it. Matt and I ended up agreeing on that and decided that we’d get to a good stopping point with things at Scribd between projects and decided to go all in, effectively.

Andrew: You told Jeremy in the pre-interview that the first thing that you did was you built a prototype. How long did it take you to build this prototype?

Quin: Not that long, the way Matt and I work. I think we got most of it done in about a weekend.

Andrew: Really?

Quin: And then probably touched it up the next weekend. That sort of thing.

Andrew: What was in the prototype? What could it do?

Quin: Basically the way we wanted to present it to people and the way it seemed to have the most impact was just letting people type in a URL. All they had to do was type it in and we went and we grabbed all the content and we would show it to them as it was coming in in real time. Then they could start searching within 30 seconds and install it on their site. The install was just copying and pasting a little piece of Java script and put it on your site.

Andrew: What’s the part where you’re grabbing it and showing it to them? You’re showing them as you’re indexing their site?

Quin: Yes. We start pulling the content and we show them what we’re extracting from it. So they see all the pages fading in right in front of them to show people that something’s actually happening in the background.

Andrew: Then did you allow any tweaks of the search results within the first version?

Quin: The tweaking of the results came pretty soon after that, but I think the initial prototype didn’t have that. We actually built that before our YC interview. We got a hotel room down in Mountain View for two days before and just stayed up the entire time and built out the rest of a proper demo because we were supper nervous for the interview. We had all these people that we had idolized. There’s Paul Buchheit and we use gmail every day and Sam Altman and all these guys who we really idolized. We’re going to be showing them this product that we built. We wanted it to be perfect. So we stayed up for two days and built this prototype that actually turned out pretty nicely. We had all the result tweaking and everything like that.

Andrew: What was in there? Again, I’m going back to my notes here from your conversation with Jeremy and you said that when you were doing your research ahead of time, before you had even built the product, you saw that people wanted auto-complete. They wanted spell correction. They wanted to have multiple types of results. They wanted to exclude pages. How did you know what to put into that first version and what to leave out?

Quin: We prioritized things and also decided time was so constrained. Jumping back to just before that trip down to Mountain View, we’d actually decided in an effort to push ourselves into it, to leave our jobs at Scribd before the interview. Which was kind of terrifying at the time. In retrospect it all seems fine, like a good idea, but it was scary at the time. Once we did that we only had a few days before the interview. That’s when we went down to Mountain View and just said, “We have this much time. Let’s have a ranked list, basically, and get through as many of them as we can.”

The way Matt and I work is we sit with headphones on six feet apart and come up for air every hour or so and say, “What did you build?” “Alright, I built this.” And then put it together. That’s what we ended up doing and we got through quite a bit. It had the things where you could just put in a URL. We had different types of data. So if you had user profiles and videos or books, all those things could be shown and ranked separately. Then we also had the re-ranking of results. Where you could go and edit a result and you could drag it around and reposition it. Things like that.

Andrew: It just keeps bringing up all these memories of why we moved to you guys. I don’t want this to sound like I’m promoting you. Frankly, I don’t care whether people sign up for you. I never get any commission. I’m not an adviser or investor or anything, but I do believe that if I turn people on to you that they’re going to think more highly of me for that suggestion. One of the issues that I now remembered is I did this promotion with a publisher where I was going to interview five of their authors in five days. They promoted it on their site. They linked to this page that I promoted it on my site.

Then I did those interviews and posted a different post for each of the interviews. When I used Google or WordPress Search on my site, every time people searched for the authors that I interviewed, the promotion page would come up which was useless to them afterwards. Yes, I could have linked it up, but that’s one of so many other issues that I don’t want to have to go and find and link up and recognize that this is something that’s causing a problem. That’s just one of so many issues. That’s why, for me, excluding pages was so helpful.

Quin: Sure.

Andrew: Did you ever reach out to other entrepreneurs that Y Combinator backed and get them to tell you what to expect from the process? How to speak to Paul Graham? What Paul Bobwhite [sp] likes to see when he’s talking to a potential investor and potential investment?

Quin: Absolutely. So we talked to some people who we knew. We talked a lot to the Parsk [sp] guys because one of the founders was a founder at Scribd and another guy we worked with at Scribd. We talked to those guys a fair amount. Then we also just kind of reached out blindly to some past YC companies. We saw some parallels as far as what their markets were. Obstacles they would have. One of the groups was Mailgun, who is a YC company a few batches before us. We reached out to them. I just said, “Hey. You guys don’t know us, but if you have a moment I would really appreciate some time to ask you some questions, get some feedback.” I think they replied in about two minutes and said, “Yeah, come by our office sometime.”

So I went over there and met them, Ev and Taylor. They sat with us for like an hour and a half and just gave us feedback. They were like, “I do see a lot of similarities. Here are some things to consider.” This was all before our interview and they really fine tuned our approach to it which was pretty remarkable. It is one of those things where unless you live out here and you’re kind of entrenched in this, it’s such an amazing benefit of being here. There are a number of other people, too. Sam Odio from Freshplum helped us immensely during YC. He didn’t know anything about us or who we were. Of course the Scribd founders were really helpful as well.

Andrew: Do you think some of those doors were opened for you because you were working at Scribd and you can say, “Hey, I work at Scribd.” Scribd feels like a brother to all of these entrepreneurs. Do you think that helped?

Quin: In some cases, yes. But in cases like Mailgun, I don’t know if those guys knew at that time at all that we had any connections. In fact, we were trying to downplay it at that time because we hadn’t left yet. We were so paranoid. We didn’t want anybody to know who we were.

Andrew: I see.

Quin: At that time we were pretty under the radar and they just were totally willing to have a conversation with us. Now I know how busy those guys must have been at that time and they were still willing to just help somebody out because they had been there before.

Andrew: What’s the best piece of advice that you can give someone who is pitching to the Y Combinator team about joining Y Combinator based on your experience?

Quin: I think that would probably be it actually. There are tons of people, especially nowadays. There are so many people who have been through this process before. Almost all of them were probably in your position. You have a lot figured out and you have a lot that you haven’t figured out and you can probably get help with that. I think there has to be a lot of empathy there because people are really busy. Make your schedule work for people and you can get help. That’s the amazing piece of the community out here. That people really are accessible.

Andrew: I think I’ve noticed that. Not just here in San Francisco, but also all the successful incubators, all the successful accelerators have alumni who are happy to talk about them. The accelerators that feel like a mess or people feel depressed or sound to be depressed when they talk about them, their co-founders probably aren’t excited to help out. If you’re in Y Combinator, you’re kind of excited that there’s a new member of the team somehow. I don’t know how they find the time, but they seem to.

Quin: Yeah.

Andrew: So you get in there. You already know how to build your product. You know the problem. What do you learn from Y Combinator that you didn’t know by yourself?

Quin: I would say that a big piece of it was, like you said, we are pretty good at engineering and product engineering and that was a big thing that we had to do for Swiftype. There are all these things that we weren’t very good at or we didn’t have experience in. Fundraising for example. Fundraising is one of the absolute, forgive the pun, goldmines that you can get from [??]. There’s so much information there and so much knowledge from the alumni over the years that you can basically have any bizarre question, be in any kind of complex situation with your set up and somebody is there to answer it for you.

Andrew: What do you mean? What kind of complex issues did you guys have?

Quin: For us, I wouldn’t say that we had terribly complex issues. You know when you’re structuring terms and things like that. I think one of the best pieces of advice we got was to not alter terms for every investor. I know that you talk to people and they’re like, “Well, we’re going to add a little bit more value than the other guys you have, so why don’t you give us a discount?” Or, “Why don’t you give us a lower cap?” And that sort of thing. We were totally willing to do that for a while and then somebody was just like, “Don’t do that. You’re going to regret doing it later.” Now I can tell you that we are glad that we did not do that to this day.

Andrew: Can you give me a name of someone you would’ve wanted to give more to?

Quin: We wanted to give more to?

Andrew: Yeah.

Quin: Sam Altman is an example.

Andrew: I see. Okay. Did Sam ask for it?

Quin: No, he did not.

Andrew: Okay. So why would you want to give more to Sam? What is Sam bringing to the table that other investors are not?

Quin: Sam has a very personal way of dealing with us. A lot of investors, I think they deal with you in a very professional sense. Where they’re looking at you as someone who’s running this business, which is totally correct and understandable. Whereas Sam has been through the entrepreneurship part. He’s been the founder. So Sam will kind of qualify any advice he gives you with, “This is what makes sense for me as an investor to tell you. And there’s also this piece where you need to consider this for you.”

Andrew: For example?

Quin: As far as grand direction of a business, you could take a small evaluation and look for a small exit. Or you could take a large evaluation and a bunch of money, but here are the things that come along with that. That’s the sort of thing that you get from someone like Sam that you may not get from a larger investor.

Andrew: Still with all the support you confess to Jeremy Wyse that raising your seed round was very challenging. Why was it challenging?

Quin: It was. I think the main thing was that it’s just very time consuming. It’s basically time consuming until you prevent it from being time consuming. Matt and I, which is a little bit non-traditional but we think it was the right way to do it, we basically did all of our meetings with both of us present. We have very similar skill sets and socially are very similar. That’s just the way that it made sense for us to do it. But every time we were in a meeting with an investor nothing was happening on the product. We did a lot of our seed round during Y Combinator.

There’s a huge amount of time during then that every time we were in a meeting things are just at a standstill. Or if something goes wrong with the server, we have to say, “Oh, I need to use the restroom real quick.” Step out of a meeting sort of thing. That’s the sort of thing that made it a challenge was that we were in this kind of momentum phase where we’re just trying to push through all these low hanging fruit features, kind of treat the design and get it all just right. Many, many hours of each day were out of the office because we were in meetings and we were not building it.

Andrew: How much money did you raise?

Quin: We raised about 1.7.

Andrew: I said earlier that I’m amazed by how you’re able to get customers. Your first customers came, I imagine, from Y Combinator, right?

Quin: Yes.

Andrew: Other entrepreneurs?

Quin: Yes.

Andrew: One of the first ones you told me in private was Emmit from Twitch?

Quin: That’s right.

Andrew: Emmit, who did a phenomenal Mixergy interview. Anyone watching this who has any interest in creating a company with customers, Emmit, I heard for a long time was the guy to talk to. Then I did an interview with him and I understand why. Phenomenal interview. Not because of me, because of Emmit. Do a search on my site, you’ll use Quinn’s technology, for Emmit or Twitch TV and you’ll find one of the best interviews on the site. What did you learn from talking to Emmit? I just realized that I’m hesitating because I’m not bringing it back to product. Let’s do this product question. Then I’ve got to find out how you got outside of Y Combinator . . .

Quin: Sure:

Andrew: . . . [SS] . . . tens of thousands of customers. What did you learn about the product from building it for Emmett?

Quin: They informed a lot of how we built out our developer tools and the actual developer experience.

Andrew: . . . [SS] . . . developers in-house?

Quin: Yeah. And so we had actually kind of geared towards the semi- technical or non-technical people with just typing in a URL. When we started working with Twitch, he kind of gave us some background. They were basically saying, hey, we have this existing implementation. We’ve kind of outgrown it and we need a new option. So maybe I come to you guys. So we kind of stepped in, started working with them. We didn’t have a very detailed API or anything like that. We developed the API and our developer tools, like our official libraries and things like that, in concert with the Twitch.tv developers. It really shaped a lot of how that stuff has come together.

Andrew: I see.

Quin: And so it was actually just a phenomenal experience. And their flexibility, you know, working with a young company like us. Obviously, we’re hustling for them. We are staying up nights to build, but over- promising in getting them something the next day sort of thing. But their patience with us in being able to kind of shape the design of the API and the tools and things like that was just phenomenal.

Andrew: Okay. So where do you get your first set of users, outside of Y Combinator?

Quin: A lot of that was just kind of reaching out and talking to people directly. I think that’s how it goes. Paul just recently wrote that essay about doing things that don’t scale, and I think that is very true for us. Especially in the early days, it was a lot of talking to people, asking them, hey, what do you think of your search? It was even things like going on to someone’s GetSatisfaction page and seeing that their users are saying, hey, your search is broken or we don’t like your search. Then being to approach them and be like, hey, looks like some of your users aren’t very happy with your search. We think we could solve that for you.

One of the benefits we had was that we could, by just typing in their URL and building a search engine, we could kind of bootstrap that for them in a very unique way where we could say, hey, I made you an account. And this actually works. You could say, hey, I made you an account. Here’s your user name and password. If you log into Swiftype, you already have a full search engine for your website. Just go try it out. That was actually really valuable for us, and it ended up being a valuable kind of sales feature of the product. But that’s definitely how we helped a lot of people early on was, they’re just like, wait, so I can just install this right now and it’ll work?

Andrew: So you just created it for them. And you said, here, put this on your site and it’ll work for you.

Quin: Yeah.

Andrew: This is what a lot of people would think doesn’t scale, and would stay away from it because how I am going to reach more than 100 people in the next month? But Paul Graham in that essay said, do things that don’t scale. And that’s what you were doing, but I don’t know that people could tell, or be excited about the search just as it is on your site. Right? Isn’t it until you start to manipulate the search results that you realize how helpful it is?

Quin: Yeah. There’s a few things that we see as far as feedback works for people where it kind of clicks for them. So one of those is, inevitably, they go into the dashboard. They do a couple of searches. And then they say, oh, these results look okay. Or I don’t want this page on here. Then we can just tell them, well, just remove that page, or add a page, or move that page around. You definitely see it click for people at that point. It’s always kind of a cool moment to see, oh, I can just change these? And you change it . . . [SS] . . .

Andrew: So how did you do it so that a stranger that doesn’t really know you, doesn’t really trust you, but somehow had a search engine created on your site, how did you make it so that woman understands immediately, aha, I can edit this, and that’s what the benefit is of using Swiftype.

Quin: Thankfully, it was pretty intuitive for people. People kind of got it pretty well once they were in there. We would actually just kind of push people directly to an example. A lot of times, we’d let it run for a day or two and say, okay, your most popular query is this. Why don’t you go into your dashboard. Here’s the page. Then take the result that you want to be the top result, like whatever you think is perfect for that, and drag it up to the top.

Andrew: If someone was complaining about my site, for example, on GetSatisfaction, you’d say, hey, Andrew. I see people are complaining. I created an account for you. The account name is Mixergy. The password is freedom. And all you have to do is go to this page and see what a search result would look like. Is that what you’d do for me?

Quin: Yeah. We’d kind of like pass them through to their most popular query, what might have been the most popular query.

Andrew: Okay. Then how do I know right away that I can adjust or see anything of meaning. What’s the first thing that you want me to see when…

Quin: …So we show you little hints…

Andrew: …Okay…

Quin: …And we say, here’s something you could drag. Why don’t you try dragging your result to the position you want it to be in. Then try another query, and move those results around. What we also tell people to do is try it, move the results around, and then go back to your website and do the search. Because even though they just dragged the result it always amazes people that they go to their website, do the search, and it’s exactly what…

Andrew: …Right there…

Quin: …they just did. Which is very cool. It’s the sort of thing where I think it’s because it’s been such a black box in the past that the fact that they really have that control, to control exactly what their users see for their most popular queries, is like a revelation for people.

Andrew: Did you also tell Jeremy, if I’m understanding this right, that you would go and put it on people’s sites.

Quin: Yeah. Actually, that’s a good point. We built this kind of live preview mode that is one of the things that we would send to people. We had the issue where if we sent somebody an account a lot of times they wouldn’t want to log in. So, we’d send them a link that basically embedded our search onto their website through just some kind…

Andrew: …So, it looked like it was their site?

Quin: Right.

Andrew: But it had this link on it. I see. So, they could have the…

Quin: …Yeah…

Andrew: …full experience of using the product on their site. I see, okay.

Quin: And so they could see the search results kind of side by side. They would just see how our results looked. They looked nice and they were relevant. Then we’d say, hey why don’t you go try and change some of the look…

Andrew: …I see…

Quin: …that sort of thing.

Andrew: Okay. What else did you do when you were still in the getting thousands of users, maybe hundreds of users, part of your growth?

Quin: Another thing we did was challenging people with things that they’d like to be able to do with search. One of the things that also kind of clicked for people, and was a real light bulb moment for them when they tried Swiftype that they would then talk about, was showing them what users were searching for and weren’t satisfied with – like they didn’t have results for, things like that. A lot of these people would run it for a few days. And we’d send them an e-mail and say, here are all of the things people have searched for and they got no results whatsoever. A lot of times, almost every case actually, people would have some really popular query that they didn’t have content for. And for anyone who cares about their content that’s like, oh my gosh…

Andrew: …I better go and create this page for them…

Quin: …why don’t I have this content. Exactly.

Andrew: I see. But that still assumes that they have Swiftype on their site.

Quin: Right.

Andrew: How else did you get people to actually implement it? Did you do any paid advertising?

Quin: We didn’t do any paid advertising early. We haven’t really explored that a lot to this point. One of the things that we did that really, really helped our distribution was kind of creating integrations. WordPress is a good example. We created a WordPress plugin. We didn’t know a whole lot about the WordPress community at the time. We’ve learned a lot now. It creates this really seamless interaction for people, and the growth based on that has been really great. People go in and they see that they can just one click install this plugin. It creates a full search experience. It’s good for us, and it’s good for them. It’s good for us because we actually control it end to end at that point.

Andrew: I see.

Quin: So we control how it looks, how they interact between WordPress and Swiftype, and what the structure of the content is. That was an amazing growth avenue for us, because the WordPress community talks a lot, too. So we had all these people who’d come and they’d review the plugin. They’d promote the plugin for us by writing about it. We didn’t just get small things like somebody would just post a question on a new story on a site that says, hey you finally fixed your search, how’d you do that. They’d be like, oh it’s this great thing called Swiftype. It definitely grew a lot through avenues like that where we made the integration pretty frictionless.

Andrew: Did Bob Hiler help with the Swiftype plugin?

Quin: He was really helpful early on. We had probably one of the early versions out that he tried. He did a lot of really interesting stuff with it which basically allowed us to put in little instrumentation points to where people could do more complicated things, more complicated than the defaults. So we actually built those out based on what Bob wanted to do, then Bob built some really cool stuff that he actually open sourced.

And now a lot of other people are using to enhance their search, which is really cool. Bob Hyler is my mentor, a guy who has been helping me tremendously. Any time I work with someone, and the WT Valet guys, who manage my site, will tell you he just starts to take them on as a project, and he feels that it’s not enough to just implement, they have to survive after we implement, and he has to improve their stuff. Often he’ll open source what he builds.

Andrew: Do you get a lot of that? Do you get people who jump in like that?

Quin: We do. We built all of our developer tools and everything like that open source. We did that just because that kind of what our background is and it’s just how we would do it. The stuff that we’ve gotten help with is, which is remarkable, the developers for Twitch, they were making contributions. It just works really well on both sides. It gives people a little more confidence, I think, because they know they can go in and change things. We’ll have people who will kind of want to do something that’s not generically applicable, but they will fork our projects, and they’ll kind of make some changes themselves and leave it up just in case anybody else wants it.

The help we’ve gotten with that has been pretty remarkable. It’s very heartening to see people wanting to contribute back things they’ve worked on and for our benefit and everyone elses.

Andrew: I’m looking through my email right now. I just did a search for Bob Hyler in Swiftype, and there was a period there where he was almost an employee of your company. He was almost one of the developers.

Quin: Yeah. He worked with us a lot. He always jokes like, “Oh, I’m not a coder. I don’t write code”, but boy he got a lot done. He did some pretty impressive stuff, too.

Andrew: I know. He’s like a machine. He can just crank stuff out, and he keeps saying the same thing to me, “I’m not a coder. You have to understand. Don’t give me coding stuff. I can’t do it.”

Quin: Don’t tell him I told you that. Don’t tell him I told you he can write code.

Andrew: What?

Quin: Don’t tell him I told you he can write code.

Andrew: I research a lot. I will go on line. Oh, I see here the original email from him was basically removing Legit. We removed Legit which we used to use and use Swiftype. If I were to replace Legit, I would use Swiftype. That’s from his email from July 9, 2012.

Quin: Wow. It’s been a year. That’s amazing.

Andrew: Yeah, so there would be one searchbox, and if they looked for 37 signals, they would actually get what they wanted. First interview, the name, with Jason Freed, for example if they were looking, second interview name, with Jason Freed, etc., right? We wanted also to be able to organize the interview results properly.

Quin: Yep.

Andrew: Yeah. And then after all that, some interview that mentions 37 signals and another one and so on. He goes, yeah, you just explaining it to me. All right.

Quin: So I think Bob is actually an interesting case as far as another thing we did early on, was we kind of started getting some organic sign ups. We did like a long john tech crunch and started getting a god number of organic sign ups, but a lot of what happened was people would sign up, they’d kind of poke around, but we hadn’t optimized our funnels or anything like that at that point, so people would kind of get in and poke around and just be like, “Oh, that’s interesting”, and then bounce from the website.

I believe in the case of Bob I was just going through and parsing out people’s names from their email addresses they signed up with, searching them, and if they looked like they were doing something interesting, I would just send them an email and say, “Hey, my name’s Quin. I’m one of the founders. It looks like you’re doing something interesting. I’m about to help you out with it.” And I drew up Bob and saw that he worked with Mixergy, and I was like, “Oh, I’ve got to talk to this guy,” which is a case that came up a lot of just direct reaching out to people who had signed up and saying, “Hey you’re already interested, how can we help you get this installed?”

Andrew: So what else did you do to bring more people in?

Quin: We def . . .

Andrew: Go over crunch help . . . Sorry, I just interrupted as you were answering. What else?

Quin: The (?) was really good for us, and we got some more growth through that. There were a lot of comments and things like that on Hack News and TechCrunch, and they’ll reach out to people directly and do the same sort of thing like, you know, here’s an account. Here are the ways we think you could do some interesting things with search particular to your case. That was definitely like the direct, personal, reaching out, having founder in the signature so they know you’re real, and it was very helpful for us early on.

Then, after we kind of spurred that, word of mouth really kind of took over. We were reaching out strategically. We were reaching out to influential people. We wanted Mixergy to be a customer, not just because we wanted to use the search ourselves, but because saying Mixergy is running our search is very good for us.

Andrew: I see that’s good for you if people discover you, but except for the except for the fact that Bob wrote a post on our site saying we switched to Swiftype and people check it out and give us feedback and so forth, no one would know it. Unlike Legit [sp], which when we were using them their name was underneath the search box. I don’t think your name was anywhere on our site.

Quin: On a lot of our free installations, especially during the beta period, we actually would put search powered by Swiftype. And we do get quite a few leads just from that where people do a search, it’s a good experience, and search powered by Swiftype, if any of those people have a website, jump onto our site and give it try themselves. So that was good for us. I think that part of the reason that it worked well is that initially we are spreading a lot in kind of developer communities.

One developer’s site would say search powered by Swiftype and their readership is largely developers themselves. The same kind of thing kind of happened with publishers. We definitely would get leads from places like the post did for you guys, but other sites as well were content creators and bring them back in like that.

Andrew: You’re talking about is as if it was in the past. Do you guys not do that anymore?

Quin: We still do. It’s a decent source of traffic. It’s definitely not huge for us.

Andrew: What is huge for you then?

Quin: Right now the biggest thing is probably, two things. People writing about us themselves and doing promotion just through people writing about it on their blogs and things like that. Asana just wrote a really amazing post saying, “Here’s how Swiftype and Optimizely have transformed how our marketing team works.” Things like that are just fantastic because their readership are people that we want using our product so it transfers a lot of their customers or their fans to us.

The other thing is that we’ve worked on partnership opportunities as well. Ideally down the road we’ll have partnerships to where a website builder like Wix will be able to offer an option to have a one-click install of Swiftype. We’ve actually worked with Cloudflare, one of the early partners for that. Cloudflare has us powering some of their things where they do automatic search of your website if the user can’t find a page and things like that. Those partnerships have been another good avenue to bring people in.

Andrew: What’s the financial relationship in a situation like that? You pay them?

Quin: Right now it’s actually mutually free. We’re providing them a good product and they’re helping us with our growth.

Andrew: With the name. Okay. Can I tell you something? I did an interview with the founder of BangWithFriends and I thought people are going to be outraged because of the headline. Look at the name of the company. He was telling stories about how people, basically bang with friends. No one was outraged about that. There was a lot of outrage that there was no money in the company. If there’s no revenue, how can it be a real company? You could see it in the comments. Where do you get your money?

Quin: Right now most of our people who pay us are paying us based on their volume. How big are they? How many searches are they doing? Then we have some features that kind of step people through the pricing tiers as well. What will eventually be happening is that any eCommerce site or things like that that installs Swiftype, we’re going to improve their conversions.

The eventual goal is we are going to be improving conversions, raising the bottom line for these people, we’ll be charging them kind of based on the direct value to them. It’s the sort of thing that is hard to do early on. It’s hard to do when you’re a small company and you’re kind of just feeling around, but once we get to the point where we actually have people kind of measuring things through us, we have a really rich analytics suite right now, and eventually want people to be able to track conversions and have those influence their Swiftype results. Then we’ll turn that around into a sales call effectively and be able to say, “Look what we’ve provided you as far as lift. Here’s what we think is a fair monthly fee.”

Andrew: Without disclosing it ahead of time?

Quin: It will always be part of the sales conversation. It’s the sort of thing that works both ways. Obviously that helps us as far as how we can discuss pricing, but it also informs us because we know that they are getting value out of it. If they’re not actually getting value out of it, it’s not actually helping their conversions, then we probably shouldn’t be charging them a huge fee. It works for both sides.

Andrew: I see. If I’m understanding you right, the idea is eCommerce site implement Swiftype search. Swiftype tells them result number three gets you this many more sales than result number one. You should probably move number three to number one to encourage more people to click it. Then you can come back and say, “Because we gave you that information, you increased sales by X. Wouldn’t a small percentage of X be a good price for you to pay us?”

Quin: Basically. But the real key is that we actually move that third result to the first result automatically.

Andrew: Oh, I see.

Quin: The grand piece of it all is that we let those people tell us what’s important to them. That could be any number of things and we factor them all in. One interesting example is take an unbranded search on Amazon like LCD TV’s. If someone searches LCD TV’s they have hundreds and hundreds of LCD TV’s and they all have those words in them. So which ones do you show first? Why don’t you show the ones for somebody you have a partnership agreement with? Or why don’t you show the ones where you get greatest gross margins? That’s the sort of thing that we let people build in to optimize for whatever is best for them.

Andrew: I’m writing a list here to see if I can sum up for myself the different ways that you got customers. Tell me what I’m missing here. First group of people, friends and then people who you saw complaining online or being complained about. Next group of users came to you from plugins. So you were creating a WordPress plugin for example and then all the people who are in the WordPress community started talking about it and talking it up and voting it up and so on.

Another way that you got users was by having the powered by Swiftype on search results. Partnerships. Potentially helpful, not yet huge for you, but you talked about Cloudflare and how when their users need to search, you talked about the partnership. I won’t even attempt to summarize it right now. Quoting loudmouth like me because we talk and we blog so we’re good people to be on your service.

Quin: Right, we try to track down influences. We never went up to anybody and said, “Hey, you need to talk about this.” Or, “Please talk about this.” Or anything like that, but it is the sort of thing where we would track down influences. We basically keep a bunch of tabs with Twitter searches open and look for influential people who are complaining about search. They could be complaining about search on one of their websites or they could be complaining about search on a website that they frequent. Then we go to that other website and we say, “Hey, you have this person who’s saying you’re not doing a good job. Why don’t you work with us?” And then we reply to them.

Andrew: How do you do that? You’re the developer. You have to be head down focused. If a tweet goes off, you’re jumping on it? Seriously, how do you balance the fact that you’re a developer who still has to code with the fact that you’re a promoter who frankly when someone bitches about my search on twitter, you have to jump in there?

Quin: I think that that’s been one of the real lessons that I’ve had to learn through this. I write a lot less code today than I did in the first few weeks of Swiftype, that’s for sure. A lot of it is just having those interactions, but we found them to be really the key to this. We have to have those interactions with people because that is what has caused us to grow. We have countless posts and tweets and stuff out there who just say like, “The founder from Swiftype just reached out to me to help me with my search just because I complained.” Or something like that. All those cases end up being worth it. They really do. It has been an acquired skill I think for myself and for Matt. In that we have to set aside code and it can be frustrating. It can definitely be frustrating.

Andrew: I bet. It feels like code is more meaningful and it’s easier to rock it and to master it, but…

Quin: Yeah, and it’s a lot easier to quantify in a lot of cases. There are times when I’ll look up after about a week of work and say, “Well, I feel like I haven’t got anything done.” That just means that I’ve been talking to people all week on the phone, Twitter, email, anything like that. I think that that actually ends up being what counts. You hear stories about this. You hear stories about founder persistence and tracking down customers and making sure they’re happy and things like that. It is really something that has paid off for us.

Andrew: Does the NSA get all my search results if I use Swiftype?

Quin: No, they do not.

Andrew: Not yet. No requests?

Quin: No, nothing like that.

Andrew: Why don’t you do news stories the way DuckDuckGo does about how this is the alternative to Google because of, see, I’m not even positioning it right. But, does it burn you when you see [?]DuckDuckGo get tons of attention for less searches being done on its service than you because they hooked into the right story?

Quin: No. Definitely not. I mean, I think that diversity in the search market is actually a good thing for us right now. At different levels, it’s kind of fragmented. But, at the top level, I mean, when you say search, everybody thinks of Google. And that’s going to change. The idea is that we’ll be a part of that. I think what [xx] has done and what [xx] has done is actually good overall, because like, any even slice of competition that there can be, is ideal. If there isn’t competition, then like, you know, it’s not good for anybody.

For us, you know, part of the goal early on was to kind of lay low, and be quiet. I think that that was beneficial to allow us to grow, because a huge part of the value we had is that we have this whole network of sites. With all that whole network of sites, it allows us to improve the product really dramatically. So every one of them benefits from what we’ve learned from each individual. And, so, for us to be able to build up that network before we have, you know, people breathing down on us or checking in on us has actually been a positive thing for us. . . [crosstalk][xx] It is, yeah. . .

Andrew: It is. How does it feel?

Quin: It’s good, it’s good, because we are proud of this stuff. It is something, and I think we need talk about this more, because, we do want to get more exposure for the service; we want to get more people using it. You know, I’ve never had a user tell us, “I actually don’t like this; we don’t want to use this.” It’s more the case that when people see it, they actually really like it. We need to get more people to see it. We’re just having office hours with Kevin Hale from [?]Wufu, who also did an excellent interview on your show. . .

Andrew: Thanks.

Quin: . . . His feedback was basically, “Why doesn’t everybody know you exist?” I think that’s kind of a problem we’re looking to solve.

Andrew: That’s my feedback also. Why don’t more people know you exist? Alright. So, part of the reason is you told Jeremy you need to be more social. Why are you not more social?

Quin: I think it just has to do with [?] knots in my personalities. I think one of the things that you said like, [xx] heads down, focus on the work. We kind of tend towards that as shown by how we ended up working at [xx]. Really, like, that is kind of our personalities. It’s something that we’re really working towards changing, because you notice [xx] or like, a lot of times in the past, if we had the option of sitting and banging out a feature in the middle of the night versus going out to some meet-up, we would choose building the feature. It’s the sort of thing where we need to shift that. It’s something we’ve been working on to kind of say, no, we’re going to go meet these people and we’re going to tell them about Swiftype. We’re going to kind of do the in-person version of what we did early on over calls and e-mails and say, hey, you should be using this and here’s why.

Andrew: Is part of the problem that you’re a vegetarian that doesn’t drink, and no one knows how to talk to a vegetarian who doesn’t drink at a party?

Quin: Gosh, I try not to let people know when we’re at parties. I can get through usually a couple of hours solid without. . .

Andrew: Without people knowing about it?

Quin: . . .Yeah, yeah. People would be like, I don’t think you’ve had a drink this whole time. I tell people if they ask, but. . .

Andrew: That’s another reason to be here. We had a brunch over at my place and we had Mimosas because that seems to be very big here.

Quin: Yeah.

Andrew: I think half the room did not drink. They were on no drinking. . .

Quin: Oh, is that right?

Andrew: . . . Kicks. And then, we e-mailed people who were coming to brunch to find what they ate. People were vegan, so we need to have eggs that are not made out of real eggs. The founder of [?]Kinotopia does not eat wheat. I think I can say that publicly. So we needed. . .

Quin: Oh.

Andrew: . . .Cornbread with no wheat. I believe we did that, [x]. So it’s not that uncommon; so, you’re in a good space for that.

Quin: Yeah, I know. It’s definitely, there’s so much diversity out here. Everyone’s pretty open to or expecting there to be cases like that, for sure.

Andrew: Frankly, the meat-eaters in this city who drink scotch feel a little bit ostracized by the smokers who will not touch meat and will not touch drink.

[laughter]

Andrew: Alright. All that aside, congratulations on the growth. Now that the interview is almost over, you know how safe a place this is, can you now say what kind of revenue you guys are generating? It’s still early days, we understand.

Quin: We aren’t sharing that yet. But, I assume that down the line we will be ready to share it. We’re happy about our [?] other numbers right now, and just kind of want to get those out there in the discussion before we start sharing more.

Andrew: That makes sense. The website is Swiftype, I don’t love your name. I got to say that.

Quin: You don’t love the name? [crosstalk]

Andrew: It’s Swiftype?

Quin: Yeah. It’s Swiftype.

Andrew: So it’s S-W-I-F-T-Y-P-E?

Quin: Correct.

Andrew: When you say that to people, do they know how to spell it? I’m worried my audience is going to have to go over, and they’re going to get it wrong.

Quin: So, having worked at [?] Scribd, which is like, the most difficult brand name to explain with people; it’s actually has gone over pretty well. Some people put two t’s in it, and a lot people put two n’s on my name, but it all works out in the end. . .

Andrew: Do you guys own Swiftype with two t’s?

[crosstalk]

Quin: . . .We do. So it redirects back, so it’s usually not an issue.

Andrew: It hasn’t been an issue for Kevin too; he has Wufu. . .

Quin: That’s right, that’s right.

Andrew: . . . How he came up with Wufu is like, Wu-Tang clan and Fu- something, I forget.

Quin: Yeah.

Andrew: He’s doing well with it. Alright, it’s Swiftype. If you’re a premium member, right at the top, the first thing you’ll probably see is the searchbox, because we’re so proud of it. If you’re not a premium member, you can search on the right margin of our site and see what Quin’s software looks like. And, frankly, you can just go to Swiftype and apparently you can see what it looks like on your site just by visiting their site. Right? They can do that?

Quin: That’s right.

Andrew: We’ll all keep track of this company. It is the biggest search engine you guys never heard of, and it is so fun to watch you guys grow from a distance, because you really are taking an ambitious project and you’re just growing and growing and growing it. I’m glad that it’s not my own little secret anymore.

Quin: That’s right.

Andrew: Thank you all for being a part of it. Bye, guys.

Quin: Thanks so much for having me.

Sponsors I mentioned

Walker Corporate Law – Scott Edward Walker is the lawyer entrepreneurs turn to when they want to raise money or sell their companies, but if you’re just getting started, his firm will help you launch properly. Watch this video to learn about him.

Grasshopper – Don’t make the mistake of comparing Grasshopper with other phone services. Check out their features and you’ll see why Grasshopper isn’t just a phone number, it’s the virtual phone system that entrepreneurs (like me) love.

Shopify – Remember the interview I did about how the founder of DODOCase sold about $1 mil worth of iPad cases in a few months? He used Shopify. It’s dead simple and very effective. To get a longer free trial, use this code: Mixergy

  • http://mixergy.com Andrew Warner

    This post got 8,719 views yesterday, largely because it was on Hacker News.

  • http://www.jeremyweisz.com/ Jeremy Weisz

    ..And because it was a great interview with an exceptional founder (Quin and team) and amazing product.

  • James

    I’ve been annoyed with the search function on my blog for some time now. I get emails complaining about how it doesn’t give users the results that are relevant. I never suspected that there was a service that let me customize it this way. Thanks for doing this interview!

  • Arie at Mixergy

    That is so great!

  • ChrisStelzer

    My site is fairly small right now and I would love to use this… However, I’m having trouble deciding if it’s worth it. Eventually I will be using this, what a genius product. Thanks Andrew for the fantastic interview.

  • http://www.JiansNet.com/ Jian

    Thanks Andrew and Quin for this wonderful interview. I have been into search technology for a long time. I’ve also made my own search services for websites. But I found it was very tough to crack the market since search is not something that obvious. Also, I don’t have VC backing like Quin did with Y Combinator.

    Couple of things I learned from Quin where I didn’t do are:

    1) allowing users to tweak search results

    2) allowing users to remove unwanted search results

    3) giving users the ability to add search themselves to their site, sort of like self-serve feature, rather than bugging the search provider to prepare the search for them.

    I think I might add these features to my search service and see if people will like it.

    On the search service based business in general, I’ve followed the search industry for a LONG time. Way back in the 90s or 2000s, there are couple of site internal search providers, such as freefind.com, jrank.com, etc. So this is not something of a new idea.

    But the fact that none of these search services grew big tells something, right? I think it was because search is more of a feature than a business. For example, sites could have bad search feature, but overtime, they could improve and then, they don’t need 3rd party search services.

    Another story, when the Chinese search engine baidu.com started, they were powering other websites’ search box, such as the popular media site in China, sina.com. Baidu had VC funding, and at the time, they made strategic shift to model after Google and abandoned their business model of providing search services for other sites.

    So overall, after many years of looking at the search industry, it just seems to me that it is very difficult to grow a business out of providing search for other sites.

    Quin mentioned providing search for eCommerce sites as a potential revenue source. That’s a good thinking. But, again, I think the problem would be, how you are going to compete with existing 3rd party search providers, such as sli-systems.com, who have been entrenched in providing eCommerce search?

    So those are the challenges I think in providing search services. It seems to me that doing a site with powerful search feature (example, the other Y Combinator backed vertical search company octopart.com) is a better way to go, rather than doing 3rd party search service, which is behind the scene.

    While I admire Quin for coming up with the user friendly search features where Google wouldn’t provide, but I am still doubtful if this is a good business, compared to just do a vertical search website.

    I am just speaking of the above from my own experience. When I started JiansNet.com, I was using it to experiment with search technology, but I find it challenging to get users without putting up content, so overtime, the site becomes a content based site rather than search based. With content, I can still tweak the search algorithm etc., but it is so much easier for inbound marketing.

    Anyway, just throwing my thoughts around, I am still passionate about search as Quin mentioned, yes, a lot of websites just don’t have good search feature. Also, search is something that’s immensely useful in daily lives. But, in terms of a good business model, it seems it is hard to do with just search alone…

  • http://mixergy.com Andrew Warner

    We use it for our internal site, which only a dozen people on the Mixergy team see.
    Why does size matter?

  • ChrisStelzer

    Andrew I guess what I meant to say was that my revenue stream, at this point in time, is very small. $17/month for me, sadly, is a big expense. I’ve been meaning to ask you how you consider the cost:benefit ratio of adding new features to your site. For example, the transcripts of your interviews or swiftype or using a video host like Wistia?

  • ChrisStelzer

    I forgot to mention that I’m bootstrapping, using no outside sources of money other than what I earn from my website.

  • http://mixergy.com Andrew Warner

    What I’m trying to say is that you don’t have to pay anything. Look at the pricing page and you’ll see that they have a free version.

    As a bootstrapper, when you look at a pricing page, always check out the fine print under a site’s pricing options.

    For us at Mixergy, I add things when people complain about a pain they’re having with the site. Search was such a huge pain that I heard about it every single day. I tried a bunch of different options and I would have paid hundreds of dollars a month to alleviate that pain for my audience.

    Transcripts were such a huge issue that it bled into my interviews. Neil Patel urged me to do it when he was on camera & I had a live audience watching. They pushed me to listen to that need. Turns out there are many people who don’t like long form audio or video. They like reading. For example, my mentor, Bob Hiler, might have never seen a single interview I did, but he reads wicked fast so he prefers reading the transcripts.

  • ChrisStelzer

    Andrew thanks for taking the time to reply. I appreciate your feedback. I never read the fine print, that is a good tip. I’m not really getting any feedback that not having a transcript is painful, then again I don’t have a huge audience either, about 300 on a good day when I publish an interview. So I guess I’d have to be making a decent amount of money before I could pay for a transcription service.

    Do you wait until customers come to you with their pain points or try to pre-emptively address issues that you know other business owners are experiencing?

  • http://mixergy.com Andrew Warner

    Both.

  • Jessica Edwards

    I like learning new things about SEO. This is a good article and I will be bookmarking the page for further research :)

  • Alasdair Morgan

    I think you might find Billy Murphy @ foreverjobless.com/ an interesting person to interview. Check out his blog. Intelligent posts. Pretty sure he’d go for it too because he just started a new site to teach people about e-commerce.

  • jazztothebone

    This is a fantastic plugin. Downloaded and installed it and I am so excited to be able to see what people are searching for! This will change the entire direction of my site. I just couldn’t figure out Webmaster tools. Thank you so much for creating this plugin, and thank you to Mixergy for promoting it! Wonderful!

    A little Off topic: I just realized today that Scribd no longer has a free account, and people are paying to download the things I put up there.

    Shannon

  • ChrisStelzer

    Great, thanks Andrew.

  • http://www.HireYourVirtualAssistant.com Owen McGab Enaohwo

    I also think another way for the founders of Swiftype to generate revenue is in the Search Retargeting space and sharing the revenue with publishers.

  • Pingback: The Gold List: 10 Must-See Interviews of 2013 | Business Tips

New Here? Start With These