Andrew: Three messages before we get started, if you’re a tech entrepreneur, don’t you have unique legal needs that the average lawyer can’t help you with? That’s why you need Scott Edward Walker of Walk Corporate Law. If you’ve read his articles on Venture Beat you know that he can help you with issues like raising money, or issuing stock options or even deciding whether to form a corporation. Scott Edward Walker is the entrepreneur’s lawyer. See him at walkercorporatelaw.com.
And do you remember when I interviewed Sarah Sutton Fell about how thousands of people pay for her job site? Look at the point she made. She said that she has a phone number on every page of her site because, and here’s a stat, 95% of the people that call end up buying. But most people though don’t call her, but seeing a real number increases their confidence in her and they buy. So try this. Go to grasshopper.com and get a phone number that will make your company sound professional. Add it to your site and see what happens. Grasshopper.com.
And remember Pat Buckley who I interviewed? He came up with an idea for an iPad case. He built a store to sell it and in a few months he generated about a million dollars in sales. Well the platform he used is Shopify. If you have an idea to sell anything set up your store on shopify.com because Shopify stores are designed to increase sales, plus Shopify makes it easy to set up a beautiful store and manage it. Shopify.com.
Here’s your program.
Hey everyone. It’s Andrew Warner. I’m the founder of Mixergy.com, home of the ambitious upstart and I’m smiling because I’ve got Ilya Lichtenstein back. We did an interview a little while ago, in fact about 5 minutes ago we ended it, where he told me he got 35,000 users within 3 months to sign up for MixRank.
Ilya: 3500.
Andrew: Why do I keep saying 35,000? 3,500 users for MixRank, now I’m suddenly making the number seem smaller than it is, but it’s a sizable number of people.
Ilya: Well, it’s all business, it’s the advertisers.
Andrew: It’s advertisers, right. We’re not talking about the average user who happened to pop in and will then pop out and then go look at a sticker site or a Farmville site, we’re talking about real business people. And you turned a profit you told me within a week of launching the business, true? Did I get that right?
Ilya: You did. We did a lot of stuff to build up to that so it wasn’t an overnight success by any means, but when we launched the startup we were more than Ramen profitable basically from day one.
Andrew: What is MixRank? What is a single sentence you can use to explain MixRank to people who don’t know what it is?
Ilya: MixRank is a spy tool for display ads that lets you spy on your competitors ads and see what’s working for them and learn from that.
Andrew: I think I told you in the pre-interview that I get frustrated when I ask entrepreneurs what they did to build their businesses and they say, well Andrew, it just happened. It sounds like they Forrest Gumped their way to success. Are you going to tell us that? What are you going to say?
Ilya: No, I’m going to tell you exactly what we did and it really was a process that took months and months. We started working on this in earnest, probably in January of this year. So from January 2011 all the way to launching in June we took some very specific steps to get to the point where we need to be, so that’s what I’ll talk about.
Andrew: Ilya, you can see in my face that when you say we took very deliberate steps I smiled. This is the kind of stuff I love to hear. I want to know specifically what entrepreneurs do to build their businesses. All right Ilya Lichtenstein, co-founder of MixRank, what’s the first step that you took that got you to that kind of success?
Ilya: The first step is I had this idea, I had this Internet marketing business before and I built some of these tools for myself to automate some of my processes, but it was very much a service business. It was just something that didn’t scale beyond me. I was building ad campaigns and all that, so I wanted to transform that not very scalable service business into a highly scalable startup.
A niche startup that can actually sell software and be something where we can get as big as we can make it without any limitations basically. The first thing I did just to start this was I went through everything that I did to build campaigns I wrote down. I wrote down every single step that I have ever taken for every ad campaign that I built and I have this process. From this free form thing I now have a very specific process that I followed that I knew other people followed as well, or at least I assumed they did, I will validate this later, and from that process . . .
Andrew: What kind of steps were on this process? What did you write down?
Ilya: First, do some research. Figure out who’s advertising where. Then think of some root keywords to start with. Then expand those keywords into Ad groups. Then write a few headlines. All this stuff.
Andrew: I see. You said, When I was buying ads, what’s the first thing that I did? Well, I looked at my competition. What’s the second thing that I did? Blah, blah, blah and you wrote it all down in a list, OK.
Ilya: Wrote it all down in a list, step-by-step and I tried to make the list something that someone going in with no experience would be able to follow and do. Very, very specific, very detailed process. Then I went through it again and I thought okay, well if someone with no experience can take these steps and go through this process, what could a machine do or what could a computer do of this process? How much of this could I automate? How much of this can actually reasonably build that would work? That’s where the product started emerging in a way. I thought whatever I could automate, even for myself, I know I would pay for that as an advertiser. I starting thinking a little bit more about the market and the industry and just like the very broad audience, not specific like, they hate the sort of the topped out approach, the advertising market is X or the X percent of it, we only capture the X percent of it.
I wanted to think 1) How many people do I know who will pay for this. 2) How many people exist in the world that would pay for this, and 3) How many people exist that wouldn’t pay for this but would pay for something similar or something bigger. That’s where I thought could this be a real business? Could this be a scalable, venture scale start-up? I had a few pieces that I thought, “Yeah, I could pay for if someone did this part for me. Without thinking of technology or products, I was just thinking would I hire a guy for $2.00 an hour to do this for me for 30 hours a month or something? If yes, then it means I’d probably pay $60 a month for a technology product that would automate it. That was the initial research and I had a list of a half-a-dozen big steps or big features that I knew at least I would pay for.
My focus was always on building something that people would pay for, you don’t have to do it that way but I wanted something where I could see the value and make sure this is valuable to people. I have that and I had started writing some very simple scripts. Not even a product just a few lines of messy code just something very, very small and I know you always want to think big, but I wanted to start thinking small. Just do this one small piece, do this one little thing. I had these scripts and I used them. Then I started like I think I just went on instant message one day and just messaged every single one of my friends that had any relation at all to this business, this industry so either they’re a start-up founder or they’re an entrepreneur of some sort or they’re an affiliate marketer so the widest range of people possible. All the way from people who do this for a living to people that have no clue [??].
Andrew: So, if you and I had been friends back then, the way we are now, and connected on Skype, you might of just sent me a quick message on Skype saying . . .
Ilya: Quick message saying, here’s this scripts, would you use this or try it out and see what it does for you.
Andrew: It’s a script it’s not a web site that you’re sending them to?
Ilya: Yeah, just one file. It’s a file, not even a web site, it’s like run this script on your server and see if . . .
Andrew: OK. For our friends on Hacker News, what was it written in?
Ilya: It was PHP. It was really simple PHP scripts, incredibly rapid proto-typing, basically.
Andrew: OK. What was the feed back when you sent that out?
Ilya: The feedback was really mixed, actually. It was surprisingly mixed because I thought obviously it was really cool. I built it and this was my first draft and feedback. I’d never gotten feedback before from such a wide range of people and most people were like, what do I do with this? What’s the point? Why does this exist? Why did you send this to me? There were a few who sort of got it, knew the industry who for some reason, whatever there process was, was similar enough to my process where they could see very clearly that this thing that I’m spending an hour on or a few hours on a week now is done by this technology, this code.
So they said, “Yeah, I would use this, cool.” The thing, they said, was most costume people said they would use this, you know, kind of like, future tense, or maybe. A few of them said, “Yes. I want to use this. Send me more. Send me more stuff like this. I want to use this right now.”
That’s when I kind of started thinking, ‘Well, hey, maybe I’m on to something here.’
Andrew: Before you continue with this story, I’m curious. Were you worried that some people would just take the script and use it and never have to pay you again?
Ilya: And many of them didn’t, I’m sure.
Andrew: Now, you’re trying to build a product that they’ll end up paying for and you’re basically giving it away for free, without ever being able to charge for this [??] script.
Ilya: Well, this was something I hacked together in, like, a week. So it was really very minimal.
Andrew: I see.
Ilya: And it was cumbersome enough to use [inaudible] The question I thought you were going to ask is something very different. The question I thought you were going to ask was, ‘Are you worried that people were going to see the scripts and see that they could use it and steal your idea?’ And decides–
Andrew: What about that?
Ilya: — to do this. For some reason, I never really had that fear. My thinking was, ‘Well, if someone is so passionate about this idea and is so, so interested in it, they’re going to drop every single thing they’re doing, quit whatever job they have and start working on this, then, I would love for them to join me. Be my co-founder, and help me build this.’
Andrew: I see. OK.
Ilya: There are not a lot of people that are like that. So, I’ve never been a fan of being in stealth mode or hiding your idea because it really, really isn’t about the idea. It’s about the execution, as anyone will tell you.
Andrew: And is this, partially, you doing some early marketing in addition to product research?
Ilya: I was just, I guess the cool term for it now is customer development.
Andrew: Mm-hmm.
Ilya: It was just the very early stages. The biggest fear for me was always that I would spend months and months building this big elaborate thing and then realize that nobody wants it. And so, as you know, the whole wheat start-up philosophy, that’s their focus and that’s what they try to promote. It’s like launch quickly, get feedback, iterate and all that.
This was not even something that’s launchable. This is just some code that I wrote. But, I wanted feedback immediately from people in the industry. My thinking was that, like, if these people who are living and breathing this business or doing this every day, if even they don’t want to use it, then I have to do something else.
I have to change it because if even the most passionate people about this wouldn’t want to use it, then I’m in trouble. So fortunately, that wasn’t the case. A few people kind of saw the value once used. This was, like, really simple script that just, I think it found traffic sources from media [??]
So just do a few searches and find sites that it thinks would send you good traffic or whatever. It’s like really simplistic.
Andrew: OK.
Ilya: Set out like an ugly URL–
Andrew: What’s the most useful piece of feedback you got back?
Ilya: The most useful, there were a few, kind of, formats of feedback. One of them was, ‘What is this? I don’t know how to use this.’ One of them was, ‘I would use this if you had this feature, if you made the UI better or something.’ The most valuable was probably, ‘How can I help you make this better?’ or ‘I love this. I want this to become better because this is a product that I’m going to use and other people are going to use.’
It wasn’t, I guess it was just the beginning of a conversation. They wanted to engage with the products and they wanted to help me think through it and all that. Not just, this thing sucks. I want all this list of 100 features. But, just kind of like that, almost starting a conversation.
Andrew: All right. I was interrupting your flow. I’d like you to continue, but before you do, can you turn the blinds down. The ones that are over your shoulder?
Ilya: Yes. Is there…
Andrew: I see that you and I have been talking now for so… Oh, there you go, the lighting.
Ilya: I know, all the lighting.
Andrew: You and I have been talking for so long…
Ilya: I can’t unfortunately. I live in the Mission which is the sunniest part of San Francisco. If I turn them that way? Is that good?
Andrew: Let’s see.
Ilya: Am I just going to be bathed in light always?
Andrew: [chuckles]
Ilya: I don’t know. I don’t know how much I could do, unfortunately. It’s okay, they can still see me.
Andrew: Yes, good. In fact, I don’t even care if we have a beautiful shot either one of us. I just care about the ideas.
Ilya: Right. Yes.
Andrew: I don’t know why people watch the videos and if you are watching the videos, help me understand why you watch a video instead of just, listening the background. All right. So, go on. You were starting to say what you did next.
Ilya: Because they want to see me smile when I say something interesting and useful and…
Andrew: That’s good for me and you because we can kind of sense each other but I don’t know why the audience would want to do that, say, instead of read the transcript or listen to the audio. They do. I’m glad they do. Go ahead.
Ilya: Yes. Well, now you’ve made me self conscious about the lighting. That’s OK. So, yes, I had these rough scripts that I had kind of hacked together, and I knew I wanted to start up. There are places to start up. There are generally a few places that you can do a start up in New York. You can do it in Boulder or something. Really the place is San Francisco and the Silicon Valley and the Bay area.
I just finished college; I had been in school in Wisconsin at the time. I finished that, and I’m like, all right, I want to do a start up. So, I’m just going to come out here to the Bay area. I didn’t know anyone. I’ve never been here. I read about it on Tech Crunch. It seemed cool. I had some savings from my previous business. So, OK, I’m going to come out here and just meet people and look for a co-founder because that’s the next step before I wrote more code.
I had this very rough prototype. I knew I needed a co-founder, someone who was really technical and was really deep into technology and could complement my skills. A lot of people look for a co-founder that has the same skills as them because that’s easy and they could talk very easily, but I think it’s real important to have someone that has complimentary skills and one that would be a good fit.
And so, I started going to every type meet-up. For three months I’d go to every single, like every day I was forcing myself to go to tech events and talking to people and meeting people, and just talking to them and talking through this stuff. I met my co-founder, Scott, and we started working together, and we started building this rough prototype.
Before we had written a line of code, to sort of get back to how we got the users and got the traffic, before any other code was written, I started blogging about the topic, blogging about marketing and marketing tips and all that. That was incredibly helpful, actually, for me because that resulted in a lot of people emailing questions and feedback. Like, I don’t understand how to do this or I’m struggling executing this part, sort of stuff that I’d been . . .
So, I was in the Internet marketing industry for five years, and stuff that I took for granted and that the core of people that I did some initial talking to took for granted, that they assumed was not the case for pretty much everyone else that I talked to that was just a start up founder or someone that wasn’t doing this for a living. That’s a much bigger audience.
I started writing blog posts, and at first it got no traffic whatsoever. But I kept at it, and slowly it picked up. It picked up on Hacker News and Twitter and Social Bookmarking and whatever, and at first it was like, ten subscribers; then, there’s 11 subscribers; then, there’s 100 subscribers, and then there’s a couple thousand subscribers.
Andrew: You’re just doing this because you want to build up.
Ilya: Part of it is SEO and traffic. So, that was my goal, where we have to start getting traffic. The other thing that I had, before a line of code was written and before I knew the product was, we had a landing page. Now, everyone has [??] landing pages and all that, but we didn’t use that. I think that was before that was big. We just had an email box. We had some ridiculous domain. It was called inside.io. It wasn’t even a dotcom. Those weren’t available. The domain said I had a blog, and it said, if you want to know.
So, the very important thing about the blog is that it was on that domain, the same domain as the landing page so people would go to inside.io/blog, and then they would click from the blog, or they would see like, what is this thing? Why is this guy blogging about it, and they would end up on the home page which had this email box.
I remember struggling with figuring out what text to put on the landing page. This was one page, literally a one sentence description of what they’re putting their email for, other than like, we’ll let you know what this thing is launched.
Andrew: Why one sentence? Why didn’t you spend a little bit longer explaining this product?
Ilya: Because I didn’t even know what the product was at all. I think back then it was very different. The vision for the product was very different. It was very small. If I actually said what it actually was, this product helps you find traffic sources for scaling media buys, not a lot of people would enter their email because they’re like, I don’t do media buys right now, or I don’t need to scale or find traffic sources.
And so, it had to be something that was sort of tantalizing and intriguing, but at the same time it had to be something that was sort of descriptive. I wrote so many different headlines and tested different logos and stuff and different graphics. I think at the end of the day it was like, inside.io helps you get traffic, like that.
Andrew: Inside.io helps you get traffic. That’s what the headline was.
Ilya: That’s it.
Andrew: Were you AB testing this to see which headline worked best?
Ilya: No. It wasn’t. It should have been. Honestly, we did not get enough volume to do any sort of AB testing.
Andrew: OK.
Ilya: It was just taking forever.
Andrew: It was just you, writing your rough drafts in public and looking for the headline [??]
Ilya: Yes. I would change it every few days and anecdotally see how any emails I got or whatever. It wasn’t really very much, two a day.
Andrew: What was next in the progress?
Ilya: Next, we started working on this product’s, what I call where we actually start building a rough prototype. At the same time, the blog kind of started taking off. What I was alluding to before was that people really started emailing a lot of feedback and a lot of questions.
They would say, “Well, I don’t get this” or “Here’s what I don’t know how to do.” And that was so, so helpful because our goal by that point was to make it as easy as possible to get traffic. Like, that’s the big picture, however you do it.
That means that anything that people didn’t get or even didn’t want to do or didn’t feel like they were qualified to do, that’s where the features came from.
For developing the products, it really was all about how can I, some people are sending me emails, like, “How do I do this?” I wanted to figure out what do I build so I can respond to this email with “Use my site, that’s how.”
That was my goal. Started getting more traffic on the blog. I did this thing I spoke about really briefly in the last interview we did. I posted on Hacker News and I said, I’ve been doing marketing for awhile. Email me and I’ll help you build a campaign or I’ll help you get traffic for free, manually.
I thought I would get a few people just to give us feedback. We got over 150 responses. Different businesses emailing us, me. I responded to every single one and that–
Andrew: How could you help them, 150 people?
Ilya: I thought if I was this person, if I was this person right now that was emailing me, if I was in their shoes, what would I do? What would I do, what would be my next step? What would I do to double my traffic in the next month? If I had to, if there was a gun to my head and someone said, ‘You have to double your traffic in the next month or we’ll kill you.”
What would you do? What would I do? So with that mind set, it was very interesting, for very different business and start-ups because it was all different approaches. I tried to give someone one thing, like, if I had to double traffic, here’s what I would do.
Whether that’s buy ads or create content or do some kind of social media stuff or whatever, whatever it was. As a result of that, because it wasn’t a very scalable method, after–
Andrew: But you were sending them an email with advice. You were just saying, I’m a guy who’s bought traffic profitably in the past. I know marketing online really well, I can give you a quick email response that will tell you what to do. And it might just be go buy traffic and it might be, consider building a blog or whatever.
Ilya: It was fairly long, like a paragraph or two. But, yes, like, one actionable thing they could do. That’s all I wanted. That was not very scalable, but created so much goodwill with the community. When I send them an email now, they know I’ve got all these emails now, like I got 150 emails. Sorry, I got back to you few days later or whatever.
They know and they know I took the time out of my schedule to them and to talk to them.
Andrew: Why were you doing this? Beyond being nice to them and maybe helping your reputation in the space, why do this? What was the benefit?
Ilya: Yes. Beyond the karma and all that, it was just to get exposure and to get these conversations. It was really like sales. It was the first step of the sales cycle because what I got them to do is, these 150 people are emailing me with their problems. The problems they have right now with marketing.
What is not working, what they’re struggling with. It’s the best sales you could hope for. I don’t know that much about sales but I’ve been reading about it. It seems like any good sales conversation starts with that, starts with a problem.
Andrew: Oh, I see. So anyone who told you about their problem while you were building out the site could potentially buy the product when you build it out.
Ilya: Yes. Exactly. It also could get feedback. So it was all about feedback in the beginning.
Andrew: OK.
Ilya: I’m really, I’ve always been data driven and wanted to confirm all of my assumptions and all that. That helps so much. What I remember happening is, now that I took the time to respond to them, when I say, “We have a beta that we launched, please give me your feedback. It’s free, please try it out” with that list of 150 people.
That was our, like, the very beginning of our email list, was those 150 emails. I think the conversion rate on that was over 80%, it was incredible. It was the highest conversion I had ever seen on an email campaign.
Those people had some kind of relationship for me. They knew me. There are all these tricks to writing mass emails to make them seem very personal, as you know. It really was something where I did at some point send them a personal email and they knew me. That right there, just by entering that one post on Hacker News, that’s 150 people on the data list, that are interviewed, the most passionate, and active users that are exactly the target market. So that was a start, the other piece was people start reading the blog and sending it around.
So we had about 700 or 800 people, on the list, before we had written a lot of code, we had a very rough prototype, and some of them kind of knew what it was, and others just wanted to try it. That’s one thing that we knew about all of them, that they all had this problem and wanted to get more traffic, that’s why they were reading the blog and emailing me. So, around that time, we had this rough prototype and this email list, and it was about March of this year and we applied to [??] and that was a really interesting process. It was very intense, we had this crazy frenzy to define our product and write this application.
I think one of the questions on the application is how do you know that people want what you’re building, and we had such a good answer for that. Because we said, well look, people are sending us emails by the hundreds, telling us that they want what we want to give them, so that helped. It was a frenzy to fill out our [YC] application, we ended up doing it all in the middle of the night. My girlfriend at the time was visiting, and I was like sorry but you have to go sleep on the couch because we have to work on our YC application. We ended up just like cranking this thing out like in a mad all-nighter and basically just getting every answer came back to that point. How do you know people want what you’re building, because that’s their biggest fear as investors.
Andrew: I see. Every question that they asked you answered.
Ilya: Well, we answered somehow with that, it’s like how much money will you make. Well we can make this much, we know, because these people are on our email list. How did you come up with this idea, well we came up with it because all of these people emailed us.
Andrew: There’s something that you mentioned, that I didn’t ask a follow-up question on. You said that you got 150 emails based on that hacker news post offering to help, but then you also said that you ended up with 600 people on your emailing list. How did you do that?
Ilya: So that basically there were a few things. One, is that the one thing that the way you learn, and I’m sure this is like building any email list. You have to offer value up front, let’s provide something of value. So I would create these things, for example, we were talking about Facebook apps earlier. I had this script that I wrote that collected all of these images that I thought would be a good fit for Facebook ads, something like 15,000 images that you should use on your Facebook ads. Answer your email and I will email you a link to download this stuff.
The reason that worked really well, is because they’re not answering their email for some pie in the sky Navy or whatever. They want that thing that I’m going to send them right now, and so they’ll open their email for that.
Andrew: By the way, I’ve asked others about that experiment, and they said, Ilya did really well with it, but a lot of those images were copyrighted. He was just pulling and scraping up whatever images were available.
Ilya: I did, yeah. That was sort of skirting the line, I suppose. So I just crawled, all of these Facebooking images, basically from ads.
Andrew: Paul Graham, I think he told this to [??]. He said that one of the key factors that he looks for is naughtiness. Start-ups have to do slightly devious things, Paul Graham said in Tech Crunch. You can tell if people have a gleam in their eye, you don’t want the people who would be obedient employees, we’re not looking for people who did what they were told in life. So there’s your one opportunity to do something a little…
Ilya: Yeah I think that falls into the slightly devious category.
Andrew: All right, just slightly. So you did this, you collected email addresses from people who could potentially give you feedback, a lot of them were giving you feedback, you applied a Y Combinator you get accepted to Y Combinator. What’s next?
Ilya: So the next thing is, we get accepted to Y Combinator, and we’re so excited. We go to office towers, with Paul and tell him about what we’re doing and all that. He says OK, my advice to you is launch your product right now. We’re like Paul it’s not really ready, there’s nothing really there, it’s a really crappy prototype. So the next week, we come to office towers again and we talk to him and he said my advice to you is to launch. We hear that, like, we’re working on it, and this went on for about a month, and YC is a three month program. Finally, my thinking was well I’m not getting any value out of these office hours because I keep telling you the same thing. Just for that. Just so the YC partners talk to me and give me useful advice we’re just going to launch.
When we launched our products I would . . . This is a free beta by the way that we launched. I was terrified when we launched our product because I thought it was so bad. Our competitors that have been around for a long time had so much more features and so much more data.
We crawled our display ads. When we launched our product we were not tracking banner ads at all. That’s the thing. Display advertising to most people is banner ads. We just crawled AdSense and TecSense ads because that was easier to build.
I thought we would just get torn apart in the comments and everywhere else. They would say this guy is so much better and our competitors by the way, our real competitors are huge multibillion dollar companies that have hundreds of thousands of employees that have building a product for 10 years. That’s generally who we would go up against.
I thought people would destroy it and we’d have a bad launch and it would go poorly. We launched it, we launched everything and all the features we were lacking nobody cared. Nobody cared at all that we lacked those feature.
All they cared about is we had one feature that solved some small problem that they had. They didn’t care at all that it didn’t solve all of their other problems. The YC advice was very, very good. It’s just launch quickly with whatever because nobody will care if you are lacking something.
Andrew: All right. What happened next? How else did you get customers?
Ilya: We had no customers at that point.
Andrew: Yeah. At this point it’s still free beta. Anyone can use it and not pay you a penny.
Ilya: Totally free. Yep. We got some press that was helpful and then I also started getting into the marketing stuff which was somewhat of a risk. What I know is buying traffic where I get paid right away on the back end.
For this we had no ideas if any of those people would convert or pay us. We started thinking. . . We’re in YC we had demo day two months away. We needed to show a track by demo day and a lot of growth.
We were. . . their goal is to have you grow about 10% a week until demo day and show the hockey stick graph. We were growing about 30% a week. The way we did that is by really thinking through who would be the most likely to sign up for this right now.
Our only objective is let’s sign up. Get someone to sign up, we can sell them later. We can convert them. Set the scope very small because that’s a solvable problem.
Think who has that red hot burning desire that problem that is now consuming all of their time that we can solve. That was at least starving off the professional marketers, the affiliates, the SCM consultants, all those guys.
I thought OK. All those guys we’re. . . this is very sort of vague and not specifically my thinking. I was thinking where do they go online? What is their day like? Just thinking through and the biggest advice I got and I don’t even remember I might have even got this on Mixergy but maybe somewhere else.
Someone, and I’ve heard this a few times, someone said, “OK. Who is your target customer? Specifically who? What’s his name? What does he do? Where does he go? Where does he work? What does he do after he gets home from work?”
They had this very specific profile of this one hypothetical guy that either had a small niche commerce business and he sold products online, couldn’t afford to hire someone full-time to do marketing for him and couldn’t afford spending a lot of money testing. He just wanted to scale up his ecommerce business.
That profile was the profile of a lot of people that emailed me basically. I have OK. What does this guy search for? If I were him, I didn’t know a lot about marketing I would search for how to use AdWords, AdWords tutorials or how to buy ads online. Stuff like that. Search for that. Search for sites that are about that.
I looked for people that had email lists. Looked for people that . . . so that’s one of the things that I love in a media [??] is one has an email list so I can mail to people that were writing posts about this.
Our goal is we have to get traffic quickly. We couldn’t really do some campaign search where it takes a lot of time to ramp up and optimize and all that. It had to be going for sites and getting their whole audience right at once.
The best thing for that is an email list or a blog where they could write a blog post that gets syndicated by RSS and all that. We went to this blog all these blogs was about advertising tips, paper clip stuff, SCO whatever and we thought, and we still didn’t know if people would pay for this, or whatever, and so we thought, well we had to be risk adverse, so my background’s in affiliate marketing, so we said, “Well, we’re going to pay you as an affiliate, even though we don’t really have an affiliate program, we don’t have any proven track record, but we’re going to pay you as if we did, and we’ll just say, just name a number that you would be OK with receiving total from this ad campaign.”
So he says, “Well, I have 10,000 visitors on my blog, and my click drive on ads is 1%. I earn $300 a month off this.’ I was thinking, “OK. Well, based on the data that I had already, this is converting at around maybe 20-30%,” so based on that, that means that I could pay you this kind of CPA payout, and make you happy. If you send traffic, it’ll convert at 20%, or whatever, you’ll get paid per conversion in CPA as an affiliate, but you’ll still at least, we think, make more. Sometimes they didn’t. Sometimes it failed completely, it just didn’t convert on the traffic, and they missed out.
They wrote a blog post, and it didn’t make them any money, but a lot of times it killed it. It did amazing for them, and it made them so much more money, because there were no advertisers that were middle men. Nobody was taking a cut. The reason that we did that, and did the affiliate thing is that in any advertising endeavor, you really have to focus on aligning the incentives. You have to get the incentives right, and when we give the incentive that it’s tied to the performance of their traffic, it’s not like an advertiser puts up a banner, and pays them a flat rate. They are incentivized to sell you, to sell for you. We had dozens of people that we had never met before, that knew nothing about us, didn’t know us, that we were paying maybe a few hundred dollars each, that would sell this product for us. It said on the top, ‘Disclosure: This is a sponsored review.’ We would insist on that. We didn’t have many of them, but after that disclosure, it was a very positive review, and I’d like to think that it’s because they like the product, but also it’s because they were selling it for us, because they were incentivized…
Andrew: Where did you find all these bloggers, who are going to write these articles for you?
Ilya: You should [??] your Mixergy premium, because we have a course about that, but basically just thinking through really carefully, ‘Where…’ [??]
Andrew: Oh. I thought there was a certain… [??]
Ilya: … walk by?
Andrew: So, you found them the way that you taught us in the course?
Ilya: Yeah.
Andrew: I thought maybe you… I see.
Ilya: So that was part of it, but the first step was just thinking through, like, ‘Where do the people that are most likely to convert, and that are most likely to be a good customer for us, where do they go online? Where do they spend their time?’ and looking at industry sites. We very quickly realized that we couldn’t afford to advertise on sites for agencies, or for pro advertisers, all the big SEO sites. We couldn’t do it. They wouldn’t do an affiliate deal. They had too much at stake, too much traffic, all that. We had to go to the smaller guys, and be scrappy and hustle for every one of those, and make it up on volume.
Instead of going to one big guy, we’d go to ten small guys that maybe have the same traffic combined, but that would do this deal with us, and sign up and not go [??] process, so by rule of thumb, always just never deal with anyone who has an ad sales team. Those that we talked to were just like, a small guy. He’s been running this blog for a few years. He has an audience. Another trick that I learned from reading something by Tim Ferris, that was that he tries to target very diverse audiences, very different audiences.
Andrew: Can you repeat that? We lost the connection there for a moment. What did you learn from Tim Ferriss?
Ilya: The way he markets his stuff is that he tries to target very diverse and very different audiences. He might target people who are really into the environment, and people who are really into working from home, and all these diverse groups that don’t really intersect. That’s what we try to do very specifically. Within this marketing niche, but nonetheless, targeting… We had this matrix of who our users would be, or we thought would be. If you think of a Cartesian[sp] plot, like x axis, y axis. One axis was expertise, or skill, or experience. The other axis was spend, or how much did they spend on advertising. We wanted to go for that one quadrant eventually that has high expertise and high spend, but we wanted to test everything.
For example, we tested things with people that have low expertise and high spend, so people like lawyers, or doctors, real estate agents, that are in a competitive niche, that spend a lot of money on advertising, but that don’t have the expertise. That would be valuable, or maybe low expertise and low spend, so maybe like the niche e-commerce guy that sells baskets or something that he weaves, or someone that sells some crafts that might not be the most useful for feedback for the bigger picture, but might give us a real good feed back on how to build a product that is so easy to use and so valuable, someone who has no experience and no knowledge of this. So for each quadrants we have bunch of these groups, so lawyers, doctors, eCommerce guys, what ever.
Then we thought, where do they go online? Like what do they do? And who do they trust? Who’s the authority figure?
So if I was targeting entrepreneurs, and actually this is how we met, in that, the first initial thing I was thinking, “Well, entrepreneurs like Andrew is some of those entrepreneurs trust that they learn from, so I think I want to do a media buy on Mixergy, starting out”, and all sorts of stuff like that, and I would sit all day and sent all these emails, and my response rate was about 20%, 80% of them just never responded or they didn’t want to do a deal or they said refer to our ad sales team. But of that 20%, we started doing these affiliate deals just like dozens of these really small CPA deals, and that worked incredibly well. That worked really well, because we were tracking everything.
And so the thing we optimized for, since we couldn’t optimize revenues yet, we didn’t have any revenue. The next best thing we had to optimize was engagements, so we have all these internal dashboards where we track things, like how many searches has someone done on our site, how many searches have they done that was a bad experience, meaning that they didn’t get results, like we have no data on this. Or they bounced very quickly off, that they didn’t engage with the report that was received for them or analytics. So we, that was another thing, like we want people on the one hand who are experienced enough to use our site, and sort of know what to look for and all that, but on the other hand, we wanted to be very conscious to have some number to optimize for. That’s the number I want to optimize for as far as like ratio of good experiences to bad experiences on the site.
And that led to a lot of features, like things we’ve launched, like keyword search, we realized that people didn’t know who their competitors were, so they couldn’t search a domain, so they were typing in keywords, before there were keyword search in there. Like their user experience, they want some search box and type in keywords and we build the keyword search feature based on that.
So optimizing for that, optimizing for engagement and usage of the site and tracking it based on all the track stuff I did as an affiliate, tracking it by traffic source and saying, ‘Well, this guy is sending us traffic that’s converting worse, but the people that are converting are engaging with the site, they’re using it, they’re coming back every few days, we want to find more of that source.
Andrew: You build all that yourself?
Ilya: So we used KISSmetrics also, which is fantastic and [?] is amazing, he helped us set up with that. Using KISSmetrics, Google Analytics, also some internal dashboards that we use, and just tracking everything always and scaling. And I remember very vividly when we had like the first one of these blog post come up. And one guy emailed us, he sent the email to support, he’s like, ‘This is awesome, I love this, I’m so glad that I saw this blog and I found your site, and like I can use this so much’. That was more rewarding for me at least than probably like all the revenue that I’ve ever made as an affiliate. It’s cool, and it’s a way to get a feedback which is very quick and maintaining that growth, because we could go out and probably launch one of these post reviews or email listings or what ever, every few days, and just keep it up. And so every week we were growing 20, 30, 40 percent a week, just based on that. Because we got some new audience and get exposed to that whole new audience.
Andrew: But when you exposed to whole new audience, what happened to the old audience? Are you still buying ads from them?
Ilya: So we, the way we did with the affiliate, they just put up a blog post, and that blog post is there forever. And we only pay them for signup, CPA. Now we do revenue share, but we just do it for signup.
And I don’t know if it is the exact rule of thumb, but you hear that it takes people average of 7 exposures to an ad, before they convert, take action. So that means that they see it on the blog, for eCommerce merchants and then they see it again on a blog that’s like ad words tips. This guy endorse them, and this guy endorse them and that builds and a lot of the time then, I wish our tracking is better so very significant percent of our traffic and analytics is just direct, like direct traffic.
So I guess that could be word of mouth, but it could be like something, like someone like, “Oh,” they’re reminded that, “Hey, I saw MixRank”, Yeah, I remember, I was going to sign up for that and check it out. Repeated exposures, I think, help a lot. If you could do that very cheaply, where you’re not spending a lot of money, that’s really good. The other thing that we did that worked pretty well was re-targeting. I feel bad, because anyone who has ever been to our site in the past few months has now seen our ads everywhere.
Andrew: I’ve got to do re-targeting. I keep hearing how effective it is. I know it works on me. I go on one website and suddenly I see it everywhere. It feels like a much more substantial site. Of course, if I see their ads everywhere, I’m much more likely to click on them.
Ilya: Yeah. We were very scrappy with all of that. We didn’t want to spend a lot of money. What we did is, we went on 99designs and we posted a project for a banner. 99designs works really, really well for banners specifically, because you get all these different takes on your products. It’s not like a logo. A logo, you need a consistent identity. Banners or creatives or any advertising, you get all these different perspectives, all these different takes on it. You can test all of them, pretty much. What we did is, we had a winning bidder that we paid. Then I went to the top five that we liked. I said, you already paid this. Let me pay you some X-amount to license this image from you, even though you weren’t the winning bidder.
As a result, we had five or six ad variations that we can test. The ones that I thought would do best were not the ones that did best, which I’m learning is always the case. You have to test and get data. The re-targeting worked really well. I think that’s why it’s so hard to track. I know Google is testing some sophisticated tech to try to track the impact of re-targeting. A lot of it is not clicks. It’s people who remember this logo keeps popping up, I want to check it out. The other side effect of re-targeting is that when you set it up on your site, you set up a pixel that says stop showing ads to people who sign up. I think maybe some have signed up just to stop seeing a hundred banners everywhere.
Andrew: I didn’t know that. I’ve got to do re-targeting. I keep hearing how effective it is.
Ilya: Yeah. It was so cheap for us. It was a few hundred bucks a week at most and we’re getting a lot of exposure.
Andrew: If anything, from what I hear, I had the re-targeter on here, and he said one of the big issues he’s getting is, he can’t give people enough inventory. The problem is that there are only so may people who go to mixrank.com and re-targeting works by just targeting the people who go to MixRank. So you are limited by the number of people on your site.
Ilya: We could have done a lot more with that, by the way. If I was really sophisticated at it, I would segment it by traffic source and fire a different pixel for every one. I would say, people coming from TechCrunch, it will say MixRank as seen on TechCrunch or as seen on Mixergy or whatever. That’s something I would test actually.
Andrew: That’s interesting. That would be interesting. I wonder if you can say, I don’t want to just target people who went to MixRank. I would like to, if I appear on Mixergy, ask Andrew to put a pixel on a Mixergy interview. Let anyone who saw me on Mixergy, see me everywhere else. That would be interesting.
Ilya: Yeah. We did that some for investors we were raising around for the past few months. We were targeting investors specifically. A few other people have done this too, where they set up a special page that they only send to their investors and that has that pixel. That reminds you, give us money. That worked too.
Andrew: What else do we have? I’ve got a list of notes here on what you did that worked. Is there anything else that you didn’t include in here? Any specific steps that you took?
Ilya: I think the next step is getting to customers and getting to profit. We launched after YC demo day about two months ago, we launched the paid products. We emailed our list. We said, sign up, it was just $97 at that point, a month. Nothing. Like crickets. Nothing happened. We got no traffic. Oh crap. Is our credit card processing working? Is our site up? Checking all these things. Everything was working. We got no traffic. One reason for that, was that we emailed, I think it was on Saturday night, I don’t even remember, we were just working all the time at that point, but some night in the evening. This is evening Pacific time. We literally emailed our list in the middle of the night. We were refreshing our [??] to see if anyone signed up. I remember this incredible sinking feeling where I’ve been working 18 hours a day on this thing for six months, and I have all these people who have signed up who were interested. I thought, they went to the site and it wasn’t good enough for them. We failed them. We failed our audience, we failed our users that signed up for a thing that said, we promise you that we’ll get you traffic and we didn’t. So that was challenging and I didn’t realize the obvious answers, it was in the middle of the night, nobody’s checking, buying marketing software. We got some traffic the next day, we got some revenue. The biggest change that we made that exploded our revenue and that go us beyond [??] profitable is offering a trial. You see $97 a month, you go to the start-up and you pay $97 and you get access.
What we did and we optimized their flow significantly since, what we did is the current funnel, the current flow is that people go to the site, the sign-up for or they view a report and then it says after you’ve viewed like five reports it says, please confirm your email or register an account so you can view more and you can see the top ten results for free. Then you click next to scroll through results to get to the next page that says you need a pro account for this. We’re going to give you — and it goes to a page that has a video and on that page it says, this is what you can do with a pro account. Here’s me in this five minute video using the pro account to build a huge campaign.
Now you can try it for a week for a $1. That funnel worked incredibly well. Our conversion rates, I don’t even remember what the exact number is, but I think it tripled our conversion rate doing those two things. Doing a video and doing a trial. It was through a free trial, a completely free trial so it’s a $1, a $1 for a week trial. A lot of people are afraid, they’re afraid to collect the credit card, they’re afraid to get people to commit and so they say, they trumpeted this, trial no credit card required, just sign-up really easy and that is good in a sense because it reduces friction. You always want to reduce friction when building any sign-up funnel and so that’s good, it’s less fields, it’s less friction, more people sign-up. So we have that with the free accounts. We get them to sign-up, we get their email.
The problem is those people that just sing-up for free accounts, 1) that reduces the value and the brand value of your product. They’re like this guy’s just giving this away because he can’t sell it that’s why he’s giving this away, and 2) it means they’re a lot less invested in learning your product and figuring out your sites, figuring out the value and engaging with it and with products like ours it becomes a lot more valuable the more you use it because you get more data and you look things up and you come back and monitor things over time and how they trend and all that. They have to be invested. So we know one other thing that we did to reduce friction is that we made it super, super easy to unsubscribe and cancel your trial so that you don’t have to email support, you don’t have to call anyone. You go to my account, there’s a giant red button that says unsubscribe and you just click that.
We said that on our landing page, there’s a picture of this unsubscribe button that says this is how easy it is to cancel your trial, click this giant red button on your account and that helps but the big thing with that is that I know at the very, very least they’re going to come back to my site in a week to cancel they’re trial so already they’re invested enough to come back and give it another go and we haven’t done this yet, but during this week we can send them tips and send them help.
One thing that other very successful start-ups do, and that we’re going to be starting doing, is basically anyone who signs-up, I send an email, automated and it’ll say, “How can we help?” What can I do to make this experience better for you? It’s just like an auto-responder thing, but it’s from my personal email address, so if they respond I’ll engage with them personally. They know, we say that on our landing page to so they know, they get that and they get that level of support and attention. I remember watching your interview with WuFoo and that was their thing, really lighting fast support and all that. That’s how they succeeded.
Andrew: Yeah, they were live, that was a live interview and I had someone in the audience send a tech support request while the founders were on to see how fast they’re response was.
Ilya: We’re not that good, but I do that. Another thing I was going to get, so now it’s all, the optimization function has shifted. It’s not about optimizing for engagement now, though we still track all that, it’s about optimizing for retention, basically. Getting them to stick around. Step 1, put in your credit card, pay us $1, buy something. Now you’re paying us something and that means that you think you’re getting something of value. It’s kind of like a foot in the door technique of sales where you get people to make a small commitment upfront. They are more likely to make a bigger commitment later. We do that.
Anyone who cancels their trial, we send an email. We haven’t even automated this yet. This would be like personally sending emails. Hundreds of people who cancel their trial. When you click cancel it says, give us a reason, why are you canceling? When they do I say, what can we do to make this more valuable?
A lot of the time people say, I can’t afford it. Or this isn’t worth the money. I love that. That is the best feedback I can get. Because that means that, one, it’s worth “x” to them. It’s worth something. Somewhere between 0 and 67. It’s valuable to them.
And, two, I can’t afford it is a myth. If I were selling you a $100 dollar bill for $50 you would be able to afford it. That means I can ask that very direct question, how can we make this more valuable for you? What can we do to make this product worth it? Worth the money.
Andrew: I see. This is a drop down menu where one of the options is, I can’t afford it.
Ilya: It’s not even that it’s a free form text box.
Andrew: How do you get to follow up with them if they say they can’t afford it?
Ilya: People say, I can’t afford it and then I say, what can we do? What’s a feature we can build?
Andrew: Do you manually go in and email them this?
Ilya: Yes. Manually. Every single one. We should definitely automate it. But, manually, hundreds of these people. I read all of these and then ask them. Every time some one cancels. Any time. I send them that email right away.
I say, hey this is the founder emailing you, this is a personal email. This isn’t a generic spam thing. What can we do? What can we do to make this product better? What can we do to make it worth value for you? Because when you think of what they have to do and the decisions that have to go on in their head.
One, they have to quick on a link, spend time reading your landing page copy, learning about your products. Then they have to give out their email address to you. They have to trust that you wouldn’t spam them, or you wouldn’t abuse their email.
Then they have to trust that you wouldn’t steal their credit card details, and they pay you money for something they think is worth paying money for. Something they can’t get anywhere else right now. They are getting it from you. You are the one that they are willing to pay money for.
And that is kind of amazing too. Thinking about it. That with a start up you can just create value from nothing. People said they couldn’t find this any where else and that’s why I’m going to pay you for this.
Andrew: First of all, let me ask you this question. First, you get them to register by giving you their email address for the free trial. Then you say, if you want any more you need a paid membership.
Ilya: So any time they want to scroll through more results or anything like that.
Andrew: But first you get their email and then you ask? Here’s what I got on my list. Here is what Ilya did, and his co-founder, and his whole team, to make MixRank get so many paid members.
First, you got people to tell you, I’ll address the audience instead of you. First, he got people to tell him what their problems were related to marketing. He got 150 emails when he posted an offer to give free help on Hacker News.
He also gave something away in return for getting email addresses. He didn’t just say sign up for beta launch information or sign up for my mailing list. He said, I will give you all these images that you can use in your ads. All you have to do is give me your email address and tell me where to send it.
He launched quickly and we talked about some of the issues that he had with Paul Graham pushing him for a month to do it and he finally launched. He said, buy ads from small sites. Sites that don’t even have ad teams. And actually in the Mixergy course that you taught, Ilya, you showed people how to find those sites and how to solicit them.
What else did he do? He said, buy ads from diverse audiences. You don’t just want to look for start-ups or specific kind of marketer. You want to say, well who is buying ads on line? Maybe lawyers are. Maybe basket sales people, used as one of your examples. What else?
Buy on a pay per result basis so that you align the interest. The person who is selling you ads with your own interest. Even though you could of paid on the CPM you made them get money on a CPA. That is probably why I didn’t sell you an ad. I wanted a guaranteed CPM.
Buy retargeting ads. I like to recommend ReTargeter because the founder was on here. Talked about his business and seems like a great guy. Is that the company you used to buy retargeting ads?
Ilya: We used several. I know Arjun is one of them. He’s a great guy. We used AdRoll. Google also has a retargeting product. We used several. I tried as many Ad miles as possible. All of these are self serve and easy to get started with.
Andrew: If we are saying get people to pay, get them to pay as quickly as possible. Ilya gets people to first give them their email address, and if they want more they need to pay just one buck, one dollar to prove you are a real customer.
Ilya: Not zero dollars.
Andrew: Not zero dollars. The website is mixrank.com. Let me ask you one last question, but before I do, I want to give a plug to mixergy.com/premium. If you like this kind of interview, you’re going to love the courses, because the courses are all tactical. In fact, as I mentioned, Ilya brought his computer screen up and will let you watch as he found sites to advertise on, as he wrote the email that he sends to sites asking for ads, and as he went through the whole process of figuring out where the right ads are for him to buy.
In fact, any issue that you have, such as finding traffic for your website, and you determine you need to find ads, you can read blog post that will provide you with link information or you can buy a book on Amazon, although they are pretty much out of date as soon as they come on Amazon. Or you can say, you know what, I want to go to the expert, and I want to get the expert, the guy who does it everyday, to show me how to do it. That’s what Mixergy is about. If you have an issue knowing how to buy ads, go to mixergy.com/premium. Ilya will show you how he does it, and then he will give you all of the scripts he uses to copy his process, including the exact email he sends out to buy ads from people who would not ordinarily buy ads. Mixergy.com/premium, if you’re a premium member, go on and you’ll get it for free, its part of your paid membership. If you’re not, go to mixergy.com/premium, sign up and you will get all these courses.
Ilya, one thing that happens when I give people a big list of things, like I read off here, that they could do, they feel overwhelmed. What is one thing that they could do, that would get them results if they are trying to find more customers?
Ilya: One thing that I didn’t mention, that has been incredibly [??] for us, which was a really good tip from Ethan, was to mock up a bunch of fake links to features that don’t really work, and see where people click. When they click on the feature, it will say enter your email and we’ll contact you when we launch this feature. If you mock something up with or without a landing page, or with a product you’re thinking of building, and try to send traffic to it. Build something small, get it up quickly to test the traffic. You don’t have to have a full landing page or a product to start sending the emails to find out who is willing to send or sell you traffic. All you need is an idea and an audience you want to target. Just start small, test lots of things and go from there.
Andrew: I have to keep remembering to do that. Just put up links on the site, see how many people click on them, and based on the number of clicks we get, product or not. This is a great place to leave it, and I suggest that people of course go and check out mixrank.com, to see what Ilya is buying. Thank you for doing this interview, and thank you all for watching.
