A year ago they were nearly bankrupt. Last month they did over $1 million in revenue. But this isn’t a rags-to-riches story that’s supposed to inspire you and give you a happy feeling inside. That’s more of Oprah’s territory.
No, what you’ll get from me in this interview is an obsession with specific decisions and tactics that Jesse Pujji’s Ampush Media took to turn the company around. You’ll hear how figuring out Google ads and discovering the hidden magic behind Facebook opened up a new world for them.
Jesse Pujji, Ampush Media
Jesse Pujji is the Co-Founder of Ampush Media, a performance marketing company that gets it’s traffic from biddable media platforms and converts that traffic into leads.
Three messages before we get started. First, are you a user of 37signals web apps? Then you’ll want to know about SignalKit.com. It expands the utility of all those business tools you’re already using. SignalKit.com. Very plain and simple. They make your 37signals tools more powerful. Do you ever wish you could see your Basecamp to-do’s across multiple accounts, all on one page? Or would you like to get better notifications of your Campfire chat rooms? Or maybe you use Campfire as a way to add live chat support to your website? Check out some of the products being created by SignalKit, a suite of tools designed specifically for 37signals users like you. Let me give you the webpage again. SignalKit.com.
Next, after he sponsored hundreds of Mixergy interviews, if I asked you, “Who is Scott Edward Walker,” what would you say? The answer, of course, is the lawyer that specializes in helping startup founders like you. I’ve known Scott for years and have privately recommended him whenever a founder asks me for a lawyer. Scott Edward Walker of Walker Corporate Law.
Finally, if you need a number that gives you unlimited extensions, who do you turn to? Grasshopper.com. If you have, or if you need, to have your voicemail transcribed, who do you turn to? Grasshopper.com. If you need a toll-free number, who do you turn to? Grasshopper.com. It’s a virtual phone system that entrepreneurs love.
Here’s your program.
Andrew: Hey everyone. My name is Andrew Warner, I’m the founder of Mixergy.com, home of the ambitious upstart and the place where I interview like a bulldog to find out how much information I can get from successful entrepreneurs about how they built their businesses and bring it to my audience. I say like a bulldog because today’s guest and I have spent, I’d say, two hours just going over this story and making sure I get as much information as possible in this interview.
All right. So, how does a company that nearly went bankrupt suddenly generate over $1 million in monthly sales? Well, joining me right now is Jesse Pujji. He is the cofounder of Ampush Media, a performance marketing company that gets its traffic from biddable media platforms and converts that traffic into leads. This is the story of how he almost went bankrupt and how he came out on top.
Jesse, welcome to Mixergy.
Jesse: Thank you so much, Andrew. I really appreciate it. Glad to be here.
Andrew: Does doing all the prep work make this interview easier for you, or does it add an extra layer of nervousness because now we have to get everything that we talked about into the interview?
Jesse: I think it’s helpful. It’s good to just kind of figure out what works, what doesn’t work, what’s interesting to people, what’s not. Because you are the expert here and it’s the first time I’ve ever done this.
Andrew: Good, I’m glad to hear that. A lot of times people will say, “Andrew, these interviews are great but what are you going to do next?” And I go, “What am I going to do next?!? I’m taking this stuff so freaking seriously that I can’t think about anything but doing a good job with these interviews and you want to know what I’m going to do after the interviews?”
I never understand people who walk into a room and say what room am I going to go into next? How about just be in the room for a little bit, conquer the room, conquer the nation, and then talk about what’s next?
Jesse: [inaudible] They’re always thinking two years out.
Andrew: Right. I take this stuff seriously. I want to make sure to do a good job. So, let’s start off with this. How long ago did you nearly go bankrupt?
Jesse: Actually, about one year to the day we were essentially looking at each other going to probably really didn’t know what we were doing that much in this business, so we should start to think about what we’re going to do going forward. So, it’s actually funny. We didn’t mean for that to happen. But about a year ago, so this time last year, we had yet to generate a profitable revenue. We had costs. We were very close to thinking about closing up shop or thinking about changing around our ideas and doing things a little bit differently.
I would say failing, because we had sort of a plan to go for two or three full years of trying something entrepreneurial, even if we failed once. So we weren’t going to fail, but we’re going to at least move on to something different and think about a different idea.
Andrew: How much money did you guys lose, all in, by the time you hit your low?
Jesse: By this time last year, I think we’d probably put, of our own personal money, well over $100,000.
Andrew: Well over $100,000, you guys got from where?
Jesse: From Wall Street.
Andrew: That was the job that you all took right after school, right?
Jesse: So, right after school, me and my two co-founders, Chris and Nick, we all took jobs on Wall Street or in professional services, and we luckily caught kind of the tail end of the good times there. And so, we were able to take some decent money off the table and put in our bank accounts and kind of prepare for the day when we leave to do something entrepreneurial.
So, when we left, the first thing we did is, all right, we have these bonuses, let’s put them in together and start to do something.
Andrew: All right. Let’s take this story very slowly because I want to make sure to get all the details that I can. I don’t want this to be a . . . I want, at the end of this interview, for people to say, “Huh, Andrew interviewed someone who nearly went bankrupt and he came out on top, it just goes to show that if you get into financial trouble, you can always recover”
No, that’s BS, that’s the kind of stuff that I think other interviewers would do. What I want to pull out of this is what specifically caused you to go nearly bankrupt and lose $100,000 and what specifically did you do to turn things around?
Better still, since my audience doesn’t care about Andrew and his aspirations and, frankly, they just met you, so they probably aren’t that interested in your story yet. They care about themselves. I want to know what can they take from the story and apply every day and then they’re going to thank Andrew and Jesse, and then they’ll really care about us.
So, let’s handle this very slowly and start off with a deconstruction of what you guys do right now so that there is a big question mark about it. I said in the interview intro that you guys get biddable media, or that you get your traffic from biddable media platforms. What exactly does that mean, in layman’s terms?
Jesse: It’s any place where you can buy an ad and you can bid on it via an exchange kind of platform. So, to make it really simple, Google Search is one of our main sources. Google Ads, since they’re buying from the content network.
Google Display, Facebook, you can now bid on Facebook Marketplace ads. You can bid on Bing, display DSPs and ad exchanges, or somewhere else where you can you can bid. So anywhere where you can put a number that’s in a competitive auction in order to buy an ad.
Andrew: Okay. All right. And then the second part of that I said was you convert the traffic that you buy into leads. You and I weren’t really comfortable about using the word leads because it generates a whole picture in people’s minds. I want to give them clarity that won’t generate a stereotype. So, do you have a specific example of a company that you send traffic to and tell me what ends up happening to that traffic?
Jesse: Yeah, so we actually own, and this is probably an important clarification, we typically don’t send traffic to people. We typically send it to pages which we own, that we are running optimization on landing pages for. Then we’ll take that traffic and try to do the most optimization we can to try to turn it into a qualified inquiry would qualify prospective customer for that client.
So, the reason I don’t say leads is because leads is like this esoteric piece of data, someone’s information, that can get sold over 7 million times and there are a lot of shady guys in this space. We consider ourselves sort of a customer acquisition vehicle that goes all the way from building out the ads, all the way to actually getting that customer who’s interested, and then passing that along to our client. So, yes, it is a lead, technically, but we like to think of it a little bit differently
Andrew: But you don’t generate the sale. The credit card doesn’t go through your site. You don’t get paid based on credit card sales. You get paid based on sending the right person to the client, and what kind of clients are we talking about?
Jesse: Right. So our largest clients are university clients, both non-profits and for-profits, online and off-line universities. And then we also work with some Daily Deals folks. We work with a few other verticals, like random verticals here and there.
Andrew: Okay. It’s impressive that you can do so much in such a short period of time in this space. Specifically, how much revenue did you generate last month and what kind of profits did you generate last month?
Jesse: We’re a startup private company. We’re probably not supposed to be disclosing that stuff. I think your number of over 1 million a month is correct. The other thing that’s probably worth mentioning is that we are growing at about 30% to 40% every month. So we continue to grow at that pace and that’s kind of where we’re at.
Andrew: 30% to 35% per month is what you’re growing at?
Andrew: And, by the way, just in full disclosure, I always ask these questions even if I know the entrepreneur’s not going to reveal it because we often talk before the interview, and you in this case and other guests in other cases, tell me where your revenues are, and tell me what your profits are. We have very candid conversations, but part of that candid conversation is understanding what you shouldn’t reveal right now and what you should reveal.
So, even though I know the revenue, and I know that you intentionally walked into this interview not going to reveal the specifics of it, I wanted to ask it so that the question was out there for the audience. That’s a little too much behind the scenes here, but I want to get out there because it’s true for all my reviews.
All right, we talked about the revenue, we talked about the business. Okay. One thing that I think [inaudible] it’s just the landing pages that you create. There’s a whole technology thing you guys build. What does that technology do beyond collect the data?
Jesse: There’s sort of two parts to the technology. I’d say one is on the front end of the bidding side, and the other one is on the other side, which is on the landing page itself. So, what you have there is sort of these two things that are happening and they’re kind of working in cahoots as well.
So, on the bidding side, you’re saying, “Well, what am I bidding on [inaudible] and you’re going to adjust the bids and you’re going to adjust how much your budgeting in one direction or the other based on a number of automated algorithms. So, we kind of bill ourselves as the Wall Street guys who brought that kind of math over to the online media world.
To give some really specific anecdotes, one thing we look at is standard deviation of our media and conversion tools. So, if I have five keywords [inaudible] the second one I convert at an average of 6% [inaudible].
By standard deviation, I mean [inaudible] you look at stocks, you look at them all the time and you go this stock moves like this, this stock moves like that. So, we look at media that way. We get a better sense for how to allocate a budget across these keywords or across these forms of biddable media.
That’s one example of the way that we kind of use the math that we’ve done and come to an automated way for the biddable media we’re using. So, that’s kind of one specific example that was kind of happening on the media side, which is why we focus on biddable media platforms.
On the [inaudible] optimization platform, we constantly are testing five or six different variants, landing pages. We’ll test this headline versus this button color versus, which is a little bit more standard about what goes on there. But what we’ve really done, which is more interesting, is the backend optimization which we’ve done, which doesn’t show to the user.
We’ve developed our own unique algorithm, similar to the way that Google has developed their algorithms, to actually [inaudible] our search results. So, if you came to one of our education sites, and you’re a student looking to go back to school, say your typical person is a 35-year-old single mother, she’s looking to go back to college, she wants to know the best match for herself. She’ll give us a bunch of information such as age, location, type of school, studies she wants to do. We’ll take that, in addition to some scoring models that we have from third parties and from the schools we work with, we’ll also take how much they’re paying us, how much volume they want us to fill, how likely they are to end up matching, and we have a search results page [inaudible] under the ranking algorithm based on these different factors that show to the person.
So, just like you see ads on the right-hand side of Google, there’s a pretty complex algorithm going on back there, on our site you see the exact same thing happen, which allows us, obviously, to augment our revenue, but also give a better match to the prospective student and a better match to our clients.
Andrew: So, the first thing a person sees is an ad. Whether it’s on Google or Facebook or something else doesn’t matter. But they see that ad, they end up on your webpage. The next thing that they see there is a form, and after they fill the form, they get, essentially, search results you guys have stacked in order, based on what makes the most sense to show that specific user and your bottom line. That’s the magic there behind the business.
Jesse: If you know anything about Google AdWords, they say they’ve reconciled the difference between the user, the advertiser, and the publisher. They’ve sort of reconciled these three things. We’re doing the exact same thing. We want the best results for our users, the best things for our clients, and the best thing for our business as well.
So, sorry, there’s user, advertiser, and Google. So that’s the same thing with us. User, which is our user, advertiser, which is our client, and then Ampush, which is best for us. We reconcile those things by putting them into algorithm which figures out five or six different variables and optimizes what the results look like on our pages. o, every single person coming through the site gets a unique set of search results. There’s nothing that’s static on that page.
Andrew: Why are you showing them search results and not saying, “I got your data. I know who you are best fitted for. I will send this to the university and get you information directly from them.”
Jesse: We do that, in a sense. We don’t show them, we work with hundreds of universities clients at this point, or about 100 university clients, and we end up showing anywhere from one to five results.
Sometimes people come in and they’re looking for a pastry chef degree. We don’t have many results for that, so they’ll probably get one school or two schools. If they’re looking for an accounting online degree, there are 20 people who offer that. We’ll pick the three best or the four best for you and show those.
Andrew: You and I have talked also about having you back on to do a course to teach how you do this, to teach how others can buy ads the way that you buy them, and also maybe we’ll do another one about how you increase conversions.
Let’s get back into the story and see how you turned everything around. You and your co-founders, before you ever started this business, you had a history together. What was that history?
Jesse: We actually met at a pretty Wharton high school program. So we were these big dorks who would go, between our junior and senior of high school, and there’s a program called Leadership in the Business World and it was put on by Wharton for the summer.
We got there and the three of us were all DJs back then, actually, in our hometowns. So the next thing you know, we were like, “Oh you’re a DJ? I’m a DJ, too. Okay, cool.” We all talked about it. In that program, it was actually the biggest bait. I shouldn’t say this probably too loud, but we felt it was a pretty big bait and switch, because we went to the program, and the entire program is a month of high schoolers putting together business plans to pitch to these [inaudible] at the end. So we thought man Wharton and Penn are the most entrepreneurial place in the world, like this is great. we were so excited about it.
So we did this program and we remained friends after the program. We actually roomed together by freshman year and we get to Penn and we go, “Oh my God, this is like a Wall Street trade school. It’s not a very entrepreneurial place.” From there on in college, we actually were sort of torn between these two things.
During the summers, we would spend our summers doing trading and internships and working on Wall Street. Then, during the school year, we started about three businesses together, which actually had varying degrees of success. Some were actually pretty profitable. One was a small business that we ended up selling, which was a . . . we produced T-shirts like this one here. [inaudible]
Andrew: What did the T-shirts say? What kind of T-shirts specifically were they?
Jesse: We called it a custom apparel brokerage company. What that meant was that we would go to a fraternity or sorority, one of our friends, and say hey you’re having a date party, you need to get shirts made. Great, we can get those done for you.
The best they could find a quote for was eight dollars. So we would go, we had direct apparel distribution relationships. So we would buy white T-shirts for $1.50. We went to the screen printers and said, “Hey, you guys need more business, right? You’re a fixed cost business. We can give you business but you have to do it at a discount.” They said okay, that’s fine.
So we get this printed for $1.00 or $1.50. So we produced a shirt for $3, sell it to our friend in the sorority for $6. She would get it cheaper. We would make, on 300 shirts, $1,000. Next thing you know, the entire campus, and we actually grew to about five campuses and about 15 student employees while we were still in college.
Andrew: Wow. What type of revenues did you guys make there?
Jesse: [inaudible] I think our biggest year we did 20,000 or 30,000 shirts. [inaudible]
Andrew: And what size revenue did you get from 20,000 to 30,000 shirts?
Jesse: I’d say an average of $5 a shirt.
Andrew: Okay. All right.
Jesse: Some of the bigger [inaudible]. It was an interesting business to get into. We realized how many T-shirts are made on a college campus.
Andrew: What’s the deal, then, with the whole Wall Street thing? I understand that, maybe in the beginning, it was kind of like bait and switch, or maybe it was bait and switch. They said you’re going to learn entrepreneurship, and ta-da, you end up actually learning about Wall Street.
After that program, your life is in your own hands. If you’ve got this passion for entrepreneurship, and you’re selling T-shirts and you’re proving that you’re good at it and not just passionate about it, why not say, “Screw Wall Street. I’ll go do this right now”?
Jesse: That’s a great question and that’s something we talk to young guys about. There are two things really. We were intrigued by Wall Street. It’s a fascinating place to work, especially the kinds of jobs we were taking, were typically investment jobs, where you’re actually analyzing businesses. I spent most my time looking at technology, media, and telecom businesses. I looked at Clearwire, which was really this really interesting Wi-Fi, WiMAX business, and was one of the big guys working on a $200 million investment in that business. Or Demand Media, my old [inaudible] had led the investment in Demand Media.
I was looking at things that interested me genuinely and analyzing businesses is very entrepreneur, [inaudible] into another business and understanding it in detail it’s enjoyable thing. So we thought that was interesting. There was general intrigue, personal intrigue there.
Andrew: Okay. So I can understand why, with all that going for it, Wall Street would be interesting and maybe even a great move for an entrepreneur, for a future entrepreneur. So, you get in there . . .
Jesse: [inaudible] I guess, just to wrap that up there, I don’t know if you want me to wrap that up really quick.
Andrew: No, go for it. Yeah.
Jesse: So why Wall Street? I think there’s this question of it being interesting, just intellectually, for any kind of person. There’s a lot of interesting things they’re doing, or at least seeming that way from the outside.
The other thing was that I think the opportunity cost was very prevalent in our minds. We said you can go here and back then, when we graduated, people were making $150,000 a year right out of school. It was hard to just say hey look, this is a very interesting way to make some money and kind of bank that stuff for the future use.
The last thing, which was a very incorrect assumption, and this is what I talk to people all the time about now as a startup founder trying to recruit really smart people, I say, “Look, when I was in your shoes, to me it was only two paths. Either I do my own startup or I go to Wall Street.” And because I was being a little bit too financially motivated by that decision. I was saying I can make this much money or I can do my own startup, which I’ll learn a lot and I’ll be the big equity holder in the business. And now I go, having done this, it’s sort of the advice that we got from people before we started the business was, you should go learn something for a while. Go sit in a start up, and see how it works, and see what it takes to actually make something successful. For some reason, we had this really black-and-white perspective on it. We didn’t think that it would probably be worthwhile to go work at Facebook for two years, and imagine if we had done that.
Andrew: So you tell your future employees, or your potential employees, that they can work for you and both learn and make money?
Jesse: Yeah. We think of it as like a halfway house. That’s what we call it. We say, “Look if you want to be an entrepreneur in two or three years, you’re sort of almost there, but you probably want to be in an environment where you can learn a little bit first and then you do your own thing.”
People gave us that advice too when we were starting out. Even before we started the company, we had a very incorrect assumption that we either had to be entrepreneurs ourselves . . . we had this thing in our mind that the risk of . . . as I was saying, we sort of had this incorrect assumption and we said we either do our own startup or we don’t do a start up at all. That was kind of a flawed assumption.
Even after we took our jobs, senior year, at these big Wall Street places, we tried to start three companies our senior year prior to joining those companies. We didn’t go backpacking in Europe so that we could work on a startup and even pitched it back then. We just didn’t have the commitment.
I think that was the biggest takeaway. It’s sort of the biggest takeaway of anything is that we didn’t have the commitment just go all in. We constantly had this plan B and that made it a lot harder for us.
Andrew: You know what? Let’s put some flesh on that. I want to see it in action. When you don’t have enough commitment in an idea, how does that play out? What specifically did you do because you weren’t fully committed? Do you have an example?
Jesse: Yeah, yeah. I’ll give a great example. Instead of going backpacking, we worked on this idea, prior to starting our time on Wall Street after senior year, called Sportster.com. The idea was simple, obviously. It was hey, we have 2.0, Ajax is getting big, social networks are big, people love sports, let’s create a social network for sports. Brilliant.
We spent all this time putting together a business plan and we went pitched it, actually, to First Round Capital. They were just a small company, Josh [inaudible] was an alum. He obviously said no, and the second he said no, we were like okay let’s just go work at our jobs. We didn’t try to do it. We didn’t try to put up a website. It was just this thing where we wanted to get some external validation. We needed somebody else to tell us that it was something worthwhile to do rather than us doing it ourselves. When you don’t have conviction in an idea like that, you have the same problem.
To give another quick example, which you can edit out later if you want, between our junior and senior years, we were all working in finance, working 12 to 15 hours a day. We would come home every day and work on a high school Facebook. That was our idea. We said this is going to be obvious place, someone’s going to do it. There’s a company called MyYearbook, which is actually pretty big company now. We were head to head in users with them.
We launched in the summer of ’05. The second Facebook launched it, we kind of said, “Oh well, Facebook’s going to be the big guy in this space. We don’t really want to do it. It’s not going to be worth it. Let’s forget about it.” We had this lack of conviction in every idea that we would bring on and, as a result, we’d just give up. And I think that’s what the failure was, just giving up constantly.
Andrew: You kept giving up constantly because you always had another option. There was always something there that was a safety net. Okay, I’m going to come back to that.
Jesse: Yeah, when you could go make $150,000 at age 22, you know.
Andrew: How much did you make at age 22?
Jesse: $150,000, I mean 150 plus.
Andrew: You were making $150,000 on Wall Street?
Jesse: Yeah. Well, I was making more when I left.
Andrew: When you left. How much?
Jesse: I was at Goldman Sachs. I was working in one of their internal hedge funds. I’m sure they would appreciate me saying, but it’s well into the six figures. $300,000 plus.
Andrew: $300,000 plus you were making a year. Wow. All right. I could see, too, how that would make it really hard to leave your company. So, let me ask you that actually.
Jesse: Keep in mind that trajectory is like this on Wall Street. I was 25 and making that kind of money, give or take. There were 29-year-olds making three or four million dollars a year. I mean, the trajectory is there in terms of . . .
Andrew: So you’re on that trajectory, your future co-founders and friends are on that trajectory. Why would you leave that to go start a company?
Jesse: Because what we didn’t factor into our equation was that we hated it. When you’re 22 years old and you’re making these decisions, it was purely financial. You go, well we can do a startup, which could turn into this. It might turn into that. It’s kind of risky. Or we can go on this track where we know it will go like this over time in terms of earnings. We thought they were equal in terms of likeness and enjoy ability.
Three or four years in, we said, oh my God. I said this very openly. I would rather make 10% of my expected hedge fund earnings were 20% of my expected hedge fund earnings to do something that I’m creating value, to do something I care about, and build something that’s not there from scratch. We started realizing how much more our heart was not in what we were doing, despite the amount of money.
Andrew: What do you mean? I know a lot of people are listening this and saying 25 years old, making well over $300,000 a year, entertain yourself every day. Buy yourself a new iPhone, a new laptop, or something.
Jesse: How can you do it when you’re working 15 to 17 hours a day?
Jesse: I said, how can you do it when you’re working 15 to 17 hours a day?
Andrew: 15 to 17 hours a day doing what you hated? What did you not like about? I know that I’m getting a little off-track.
Jesse: There’s a couple of things specifically. I think investing is, in general, pretty fascinating. I was working the public market side. There’s a lot of nuance, complications to what exactly, why I didn’t like it. The way that job works is that 40% of your time is reading, 40% of your time is in Excel model, and 20% of your time is meeting with management teams.
My favorite story to tell about my time at Wall Street, when I finally realized that I needed to be an entrepreneur, was I was sitting there. I felt terrible by the way. I was a 25-year-old, sitting across the table from a CEO of a $30 billion company, in this case Johnson Controls, one of the largest auto parts manufacturers in the world.
First of all, you’re asking all these guys questions about their business, and you feel like a jerk because you’re 25, they’re 50, 60 years old, and they’ve been running the business for 25 years. And you go, Why is your margin so low? Why is this happening, why is that happening?
I’m sitting in a meeting and I go, “Hey, have you guys thought about putting humidifiers in cars? Whenever I go to put the heat on in the car, it gets really dry and I wish there was a humidifier.” The guy looks at me like I have three heads, and he literally laughed, and he said you are the first Wall Street analyst who has ever asked me a question like that in a million years.
I said, you know, that’s right. Because I don’t really care what your margins are going to be, right? Like I care to the extent that I want to do well at my job and perform well, but it wasn’t that exciting for me. That’s kind of the specifics of the job, of what you do.
But, I think, to encompass it more in one thought, it’s being on the sidelines was not enjoyable, being an analyzer and not a doer. I found my heart was not in that. I was constantly on the sidelines, looking at these guys’ businesses and evaluating them, not having any impact on their business whatsoever. I just didn’t like that. My heart wasn’t there.
Andrew: All right. I get it completely. I get that. It’s hard to imagine, considering the salary you were giving up, but I get it. So, you and your friends are in that situation. You guys decide that you’re going to get back together. What’s the original idea that makes you decide to regroup?
Jesse: The funny thing is, there was no original idea. I mean, that’s kind of the funny thing about our story is that we knew that the three of us, we trusted each other a lot, and there was a lot of mutual thought there. We knew the three of us had a little bit of means and we sort of had seen some friends of ours on the side, people who would come up with brilliant ideas and execute against them. People also said, “Hey you know what, I’m going to find a market, I’m going to find a way to get my foot in the door, and I’m going to do it.”
So we kind of said, “Let’s not worry so much about the idea. Let’s just worry about what we’re going to do, and let’s commit ourselves to it.” I talked about earlier, the commitment was so important.
To give you an example, Wall Street and consolidating, which is also something I did, you commit to these firms for two or three years, they’re called programs. You say, for two or three years, I promise I will work there and I will not leave. I’m going to sign a contract saying that.
We actually signed a social contract, between the three of us, if you will, that said no matter what happens for two years, we’re going to work on trying to be entrepreneurs. If we can dedicate three years of our live to Goldman Sachs, the least thing we can do is dedicate two years of our lives to ourselves and try to build something. We were much less concerned about the idea, at that stage, and much more concerned about let’s go make something happen.
Andrew: Okay. We’ve talked before the interview about how I’m interested in the thought process that goes into picking the right idea to go after. What are some of the criteria that you had when you were looking for a new business?
Jesse: It’s a good question. We acknowledged from the get-go, and I think this was really critical, what we were and what we weren’t. We had no experience in building a product. None of us were really technical. We had learned some stuff throughout the way, and we would continue to learn. But none of us can actually code. We had never done online advertising.
We really didn’t know anything. We joked that we didn’t know shit. We didn’t know a damn thing about what to do. So, we said that first we need to find something that’s going to let us learn and figure stuff out, so let’s try to find an industry which tends to be . . . well couple of things. We wanted something that would make money. It didn’t need to make a lot but enough that we could survive and didn’t have a gun to our head in terms of making something work.
We wanted something that was a big industry, obviously, a growing industry. Something where a math and data background, something you can crunch numbers actually would help. And then something where we thought there would be a lot of growth. Our plan was very different. It was let’s try to get into something just to learn. It doesn’t matter if it’s search engine marketing, if it’s [inaudible], but let’s just do something.
Then after we learn it for a little while, then we’ll start to be able to see the true ideas that exist in that space. Rather than what we used to do, which was sitting on chairs on Wall Street and go, a travel search engine would be a great idea and then say, “Wait Kayak exists already,” which is how we had done things to date. So we wanted to get ourselves as close as we could to the details, start to learn a lot, and then once we learned something, we could say okay, now we can apply to something that’s going to be bigger, and better, and greater.
Andrew: With regards to cash, which we’re looking for, is a business that would generate profits, kick off cash soon, and not one where you just focus on growth, growth, growth, with the idea that someday, one day you might end up with a profit. You want to feed yourself.
Jesse: Correct.[Inaudible] our own deficiencies. Sorry?
Andrew: You wanted to feed yourself on the business. You wanted the business to feed you.
Jesse: We just wanted to get something going that really didn’t have a gun to our head. We didn’t necessarily need to go to outside investors, and that was a big part of it. We said let’s just get a little bit and the, maybe we won’t keep doing that idea, maybe we’ll realize there’s another pocket of opportunity or bigger idea that we can, then, go after and that’ll be a part of it.
Andrew: Okay, all right. So I get that. I get where you ended up because of it. How about giving me one idea that you rejected that fit the criteria but just didn’t feel right or didn’t fit it properly?
Jesse: I have to go back to my Basecamp list. We had like 30 ideas on our Basecamp there. What else were we pretty close to starting? There were some other ideas where it was just a little bit more of a hustle and a little bit less of a learning curve. One of my buddies works in the mortgage business, and he said, “Hey, why don’t you guys just go hire a call center, I’ll introduce you to some of my clients. You guys can start having that call center, making calls on their behalf, and send leads over, send people who are interested in talking to them right over to these guys and you guys will start to make $10,000 a month. It’ll make you guys money and then you can worry about what else you want to do. It’ll create an income for you.”
I’d say there was that. We were really excited also about the ad network space and the vertical ad networks that we’d get, because it’s sort of this middleman business where you’re brokering between publishers and advertisers. We kind of went away from that business because it felt like you needed relationships, which we didn’t have, and that would make a lot more challenging.
One of the reasons that we chose performance marketing was because it’s performance oriented, it’s easier to get the relationships. People are always willing to give you a shot, because you’re selling people revenue at the end of the day. They go, “Okay, you seem like a halfway intelligent guy. Here why don’t you try it. If it works, great. If it doesn’t, it’s not that much skin off my back.”
We liked the fact that it was actually, what we call, low barrier [inaudible ] to get in. But then you can erect [inaudible] over time in the business. That’s how we operate today.
Andrew: What was it about the call center that didn’t draw you in?
Jesse: We just said well, what are we going to learn? I mean, we can learn how to get the call center going, but we’re not going to learn anything about something that we think turn into a much bigger idea. Something that can turn into a much more [inaudible] or technology or something that’s going turn into a big idea.
Whereas, with online advertising, we knew that it was a growing market, that it was something interesting. So, as long as we got ourselves into it somehow, we would start to see all the different opportunities, start to have those conversations, start to go oh wow here’s a really interesting idea that we could that we could potentially pursue over a long period of time.
Andrew: Okay. So performance marketing is a business that you guys decided to get into. What was the first thing that you did after you made that decision?
Jesse: The first thing we did, we started researching it and we decided, not only do we want to do performance marketing, we were going to focus on the education vertical.
Jesse: We chose the education vertical because on online education is one of the most amazing businesses in the history of the world. It gets a bad rap and there are certain things about it, but the fact that you can build a course once, and it doesn’t matter how many people are using that course, it can scale to the sky, makes it one of the most interesting and dynamic business models in the world.
It changes a lot of things. It’s lot of people that get educated who, previously, were unable to get educated because they couldn’t sit on their computer and do it. We thought that that business itself had a ton of tailwind behind, the way an investor would say it. It just has a long growth trajectory in front of it.
Then online advertising, of course, also had a great growth trajectory. So we kind of looked at it as this cross section of two amazing businesses that were in the same place, and we want to learn about both of them, as we did our thing. So, getting into what we started to do was a way to learn about both of these businesses that we thought were pretty dynamic businesses or industries.
Andrew: I understand your analytic background, agent business background, but what about your technical background? Who was the developer who was going to put all this into action?
Jesse: That’s a great question. That was the biggest challenge in the beginning, frankly. I landed October 5th of ’09 in San Diego. We are based out of San Francisco today, but we moved to San Diego because my folks live there and we wanted to bootstrap. So, me and my co-founders, we lived at home. I lived at my parents house and we were bootstrapping.
We touched down in San Diego and started hiring for someone. The first thing we realized, and we had a couple of advisors and friends who were sort of guiding us and helping us kind of technically screen people and whatnot, we realized that we had no idea how to screen people.
I remember lining up the first few weeks of interviews. I wish we had a video camera because we had every kind of person who could do anything come through. It was the most challenging thing. We actually worked for three months, kind of planned, and started to learn about the industry in order to hire somebody. A lot of times what happened was the people who we liked, I think we had like 80 interviews in the first three months, we ended up giving two offers out, and both of those people didn’t find the idea of sexy enough to join. The first three months which is a huge challenge in terms of us actually pushing the ball forward.
It’s also the learnings of an early entrepreneur. We would get paralysis by analysis all the time. We would say oh well this guy’s this good and we’d say oh but wait this guy, we really want to find perfect. I think that was a big thing that we learned, but it’s much more important to push the ball forward.
We brought somebody on in January. He was a nice guy, but, within two months, it became clear that he just wasn’t the right fit. When you think about where we spent a lot of money in the early days, there was one person, he didn’t really get the ball to where it needed to be. And we met somebody else and he was actually really quiet guy, and funny enough, probably not one of the better interviews we had. He seems like a nice guy and said hey why don’t you come in for contract and see what you can do. He just started running circles around the first guy we hired.
It was the funniest thing, we sat him down on Tuesday of his second day and said you seem really, really good. I mean, you’re amazing. How long will it take you to rebuild the system this other guys been working on for two months? He’s this quiet young guy, quiet Asian guy, and he looks and he thinks for a while, and he goes . . . and you really scared of the answer right, because he’s thinking, he thinks for five minutes before answering, and he goes oh, maybe Thursday morning? Thursday afternoon?
And we’re like, “Are you serious?” We just didn’t believe it. And lo and behold, he’s been an amazing part of the team and he’s been one of the critical pieces to the puzzle in terms of developing the business and get it to where it needs to be.
Andrew: So one of the first things that you do is you hire a developer?
Andrew: Okay. And you guys decide this is going to be a technology company, we need a developer on hand, we need someone who is going to be in-house working with us.
Jesse: The other thing that we did that I think is worth mentioning, is my co-founder Nick, really dived into the product. He learned how to get into code, he learned how to try coding himself. We really tried to make sure that we became as technical as we possibly could.
Andrew: All right. What was in this platform that you guys built? There were landing pages?
Jesse: In the early days, it was very, very sparse, obviously, because we were just getting it off. We sort of had built out what we thought would be the technology, and we worked a couple of our advisors to develop how you would be optimized, how you would take certain types of media you would buy and make it better. But we want to get going, too. We didn’t wait around to get our product in order before we started the business.
In the early days, in search engine marketing, you can’t start optimized. You can’t start by automating things. You have to get in and just do it. So that was like reading book, sitting in spreadsheets, learning how to build out keyword campaigns, and getting them prepared to actually launch.
Even if you understand it conceptually, if you don’t have the data, you can’t optimize anything. So, the first step is actually, and frankly, we actually overinvested in technology and automation up front because you need data first. Frankly, we had to run the campaign pretty manually for the first six to nine months. And even just in the last three to four months, we’ve actually started to get more automated and use some the original stuff we thought we’d be using from the get-go now.
Andrew: What do you mean? What did you learn by doing it manually that you didn’t know in the days when you started to build out the automation?
Jesse: There’s two things. One, and forget about us learning from it. You can’t actually automate it. In order to say well the bid should be changed in this direction in that direction, you have to have enough data and history in order to actually apply those algorithms and that data to what you’re doing.
We thought, that from the get-go, we’d have this intelligent thing and the next thing we know we go, “Wait a second, it’s not going to work because you have enough data to actually make it meaningful.” It’s sort of like these blind spots you have as an entrepreneur. You think that you know everything and all of sudden you go, wait a second, it doesn’t make any sense. The algorithms are useless, frankly, without building them manually.
The second thing about building manually is lots of the parts in search engine marketing are un-automatable. Writing ad copy, picking out the specific keywords, deciding how the keywords are going to be grouped together. We use a little bit of natural language processing there, but, frankly, it’s a lot of stuff that has to be done by humans, especially in the early days.
So, just learning what’s successful and [inaudible], bidding is just one very small piece of what actually drives successful platform.
Andrew: All right. Let me get deeper into the details. First client that you got was Kaplan, Kaplan University, right?
Andrew: And your goal there was to convert a hit into what? Into a lead for Kaplan, right?
Andrew: Okay. So what’s a keyword that you bid on and what happened with the person who clicked on it?
Jesse: You mean in a normal state or what happened during that time?
Andrew: In the beginning. You guys are just getting started, you have this idea that you’re going to do all kinds of wonderful things when a person sees your ad, and then after they see your ad, what’s the first thing that you do with that?
Jesse: We may, in addition to Kaplan, at three or four additional clients, we designed a few landing pages that we thought were reasonable and made sense. We were like here’s how we’re going to test them. It’s going to work really cool. It’s going to do this and is going to do that, and we had all this [inaudible] learn how to do search engine marketing, built, let’s say, a 5,000 or 10,000 keyword campaign.
April 1st, last year, we launched. I won’t get into more details, but suffice it to say that, out of thousand people who clicked on our ads, I think we got one conversion. So, if you do even a $1 per click, it cost us $1,000 to generate something where we make revenue of $40. That was how it worked in the early days. We were numb, frankly.
Andrew: What were you doing that was so wrong?
Jesse: There were a lot of things. The landing pages, you can’t optimize landing pages without data. We just sort of made up ideas in our heads and said oh this sounds like a good landing page, this might be a good landing page.
Andrew: Give me an example. What sounded like a good landing page?
Jesse: Something that says “Get your degree today” and a picture of a happy lady with a graduation, and a drop-down menu that lets her select the degree type and the course of study, and a button that says go do this. We’d send people there. Our search results, at that time, were far more complex. They looked like a lot of links and they were very intimidated looking.
We kind of said no this looks good, this looks smart, we should to do this, and we tried to send people there and get through the entire process. There was that. That was a mistake. And then, the keywords we were using, one of the things we didn’t realize was that there’s something Google has called Quality Score. It’s basically a quality ranking of how likely they think your ad is going to be to help a user. And they specifically didn’t like the space of education for a number of reasons.
This actually really hurt us because the second we uploaded a bunch of keywords, our quality score were twos and threes. This was something we had completely unanticipated. we thought that we would have just as good a shot as anybody else in the space.
Come to find out that Google doesn’t actually like these types of advertisers, for a number of reasons, and they initially start everybody off at a two or three quality score.
Andrew: Okay. I want to break down everything that you just said.
Andrew: I understand now why Google thought that your ads were bad, but give me a deeper understanding of the keywords. What was wrong with the keywords that you picked? What are some of the keywords?
Jesse: Masters in teaching online, master teaching degree, MAT degree.
Andrew: Why are these bad keywords?
Jesse: They’re not bad keywords. They weren’t bad and we’d probably still buy them today. I think what it was, we took an academic approach to it, because we had read books and we thought we were pretty smart guys at this. We didn’t take a make it work at any cost approach to it, if that makes sense. So we launched 10,000 keywords because we thought we were so intelligent, we had these really pretty spreadsheets, we had automates to the spreadsheets that worked really well.
The reality was that that’s not how you’re supposed to do it. What you’re supposed to do is get 10 keywords, make those 10 keywords work. Once you get those 10, then add 11, then add 12, then add 15. So, just by launching with too many keywords at once, it was completely the wrong approach on how you’re supposed to do search engine marketing.
Andrew: I see. So too many keywords at once, and you also said there was something about the text that you wrote that Google said, these guys are a little shady. What was that?
Jesse: It wasn’t the text, just the space we were in, actually. In education, specifically, Google just said that when you’re a new guy in the space, we don’t think that you deserve a good quality score. So they actually gave us quality scores of twos or threes.
The way quality scores work is that you multiply by bid and that determines your ranking on the page. So, if my average quality score were two and our competitor’s were seven, they would have a five X advantage over us, in terms of cost.
So, that was the other big issue that we dealt with. At the time, people said there’s pretty much no way out of it other then to buy your way out of that. Meaning, you have to keep buying into Google until you prove to them that you’re actually a high quality advertiser, not a low quality advertiser.
Andrew: And have you experienced that? Is that what happened?
Jesse: Absolutely, absolutely. It’s a structural advantage for us now because, today, we spend $3,000 or $4,000 a month on Google and we’re very high quality. We work with their team over there. We know people, So it’s something we did kind of have to buy our way out of, obviously.
What we did was, because search tends to be more competitive and aggressive, we pulled away from all search over the month of April and we started to focus more on AdSense, tried some display, tried some Facebook. So we tried other stuff that didn’t have this quality score issue that might be easier for us to start to generate some momentum with.
Andrew: Okay. So that’s what was going on, on the Google side of things. Then the person ended up on your landing page, what was wrong with the landing page you did? That you created? Looking back.
Jesse: Everything. It wasn’t as intuitive as it could have been. It didn’t draw the user in. It was too busy. It had too many things that a person could click on. When you want someone to go do something, you have to have a very directed focus. You have arrows pointing at what you want them to click on, you have bright colors, you have simple ways to focus on the landing page.
The other thing that we did was we had a lot of steps on our landing pages that weren’t easy to accomplish. Every step you had to give more and give more. Now it’s a very fluid feeling, if you go through one of our landing pages.
Andrew: Okay. When you say too busy, there are too many things people do to, give me an example. What did you add to the page that distracted from ultimate goal of getting them to fill out and submit?
Jesse: Too much text, for example. Links throughout the page that don’t actually get you to focus on the purpose.
Andrew: Links to where?
Jesse: Links, maybe, to our homepage, maybe one of our clients pages, all these kinds of things. Other than links that you need to have on there by law, there should be no additional links.
Andrew: I see.
Jesse: There should be buttons pushing them in the direction they need to go.
Andrew: Other than links that are required by law, you should have no other links on the homepage and, now, that’s what you did. You said it was too busy, it was also not intuitive enough. What about it was unintuitive? Looking back.
Jesse: I think the questions that we asked people didn’t draw them in the right way. If you look at any of these Daily Deals guys, like Groupons, the first thing they ask you is your ZIP Code. Everybody knows their ZIP Code. It’s not a hard question to go okay my ZIP Code is 94103. And then they ask you for your e-mail. You’ve already gotten them into the process a little bit, they’re happy with it.
We were asking stuff like you want to study an associated, bachelors, or masters, or something like that. And someone was like well, I don’t even know. I’m not even there yet. We see a lot of people leave on that. Most importantly, the landing page itself wasn’t so terrible. Meaning our balance rates, which is the way you evaluate a landing page, wasn’t so bad. It was the following pages which got overly complex for somebody. Those we cleaned up over time.
Andrew: I see. Actually, I could understand that. If I were to say I wanted to go to school, even at that stage, I wouldn’t know what I wanted to study, or where, what the ultimate goal was. I’m still in the exploratory mode when I’m coming to a page like yours. All right, that makes sense. The next page after that, with the process after submit, you said was too complicated. What did you do after submit?
Jesse: You’d answer a few questions and then we’d give you a bunch of results. So now you have to sort through these results, and they feel kind of random because we hadn’t asked you enough qualifying questions up front, and it’s just like oh now what do I do?
Hopefully, they would pick one of them, in most cases they didn’t. And then, frankly, after we had them pick one of them, the next page they would see would be even more information and more complicated.
So, it actually became progressively more complex as you went, which should be the opposite. It should become progressively simpler as you go.
Andrew: You know what, Jesse? You only had one advertiser. It was just Kaplan, how many options show people after they hit submit?
Jesse: We had like five schools at that point. We waited until we had about five clients before we launched.
Andrew: All right. So it’s still five schools. How many different links can you offer on the following page? Shouldn’t it just be five or fewer?
Jesse: We had five schools, plus each of them would have three or four programs which might be matches based on what the person said. So it turned into about 20 different hyperlinks on a page of search results. They just weren’t laid out intelligently. We sort of did it based on, and this is one of the biggest things we’ve learned about becoming a data driven company, is we would do things based on what we thought made sense as opposed to actual use of data to go, let’s not even think that we have the answer to start with, let’s just look at the data, or try a test, and then figure it out from there.
At this point, though, we didn’t even have enough traffic coming through that we could test anything. So that was all based on hunches, and based on gut.
Andrew: Okay. I’m going to come back to what kind of data can a company collect to help it fix that up. But I want to understand the full process here. We understand the Google. We understand what happens on Google. We understand the landing page. We understand the page after that with too many links. What happens after the page with too many links? What else can a user do on your site, at the time?
Jesse: Then they would click on something and, at that time, we weren’t taking their personal information up front. So when they would click on something, they would go to a final page that that would be branded with the school, and then we’d ask for all their personal information. It was in a skinny column that went really far below the fold. Anytime you put something below the fold, where a form goes below the fold, it becomes so intimidating for a user that they typically just click off.
So that’s an example of something we learned, to keep all the requests and fields above the fold so that someone knows what the form size is. You can think about yourself intuitively. If you see something where the form continues beyond where your browser goes, it will give what you don’t like oh, I don’t want to answer that.
Andrew: But you had all these fields that need to get filled out. Even if you understood, back then, that you need to keep the forms short and above the fold fully, how do you get rid of the fields? What do you do with all these fields that you need to collect?
Jesse: We sort of reversed the order. We started taking information more upfront for people, and get a little more of the qualifying details in a step-by-step process, and then match them at the end of the process, which allows a number of things which made us a lot more successful.
So it’s easier, we see conversion rates go up tremendously. And to be fair, we borrowed pretty liberally from competitors. We were two weeks away from being bankrupt and it was sort of the last straw, saying hey let’s just see what other people are doing who were doing this. It’s not like we’re the only people doing this. Just like Android borrows from iPhone. We said let’s see what they’re doing. Let’s try to integrate that, let’s create a multi-step, multi-version of this and let’s try it. In many ways, that’s what saved the business, frankly. It was that and it was parsing down our keyword campaigns to literally five or 10 keywords to try to get those right in the beginning.
Andrew: Ah. We’re skipping ahead to the turning point but it’s important to bring that up. You and I talked about this before and you said specifically that. We copied from a competitor, their process, and we limited our spend by focusing on a few key words which also focus our attention. That was the key to turning things around.
Jesse: Absolutely. Between April 1st, when we launched, of 2010 and May 16th, when we were contemplating what losing all of our personal savings would look like, we scrambled. We went from search to the Google content network. We tried display. We tried Facebook. At the same time, every week, we were launching a new landing page structure. And every week we were going oh we’re losing more money, we don’t have enough data to make any decisions.
Finally, we just started to say you know what, let’s not try to be too cute. Let’s try to get this to work. Basically, around this time of last year, this week of last year was our first week we generated any profitable revenue.
Andrew: Jesse, this is 45 days. I didn’t realize that, within 45 days of launching, you had nearly went bankrupt. How much money could you possibly spend 45 days that it would turn around the whole prospect for the business?
Jesse: That’s only the first time we went bankrupt. We went bankrupt second time, which I’ll explain. But, you know, we had to hire a few people at that point. I think, at that point, we were at four people, maybe five people. We had four people and one intern.
Andrew: Four people beyond the founders?
Jesse: No, no, no. Four people, total. So two founders weren’t taking any salary, a third cofounder was sort of there and involved, also not taking a salary. Paul, who’s our lead engineer, and another guy by the name of Ethan, who’s one of our friends from school, who would work on the media side with me and building some of these search campaigns and getting them up and running.
Andrew: So it’s two salaries and ads that were sending you into near bankruptcy?
Andrew: Is it because there was a big buildup to the April 1st launch where you spent a lot of money? Is that why you ended up with so little on May 16th of last year?
Jesse: There was an additional engineer that we hired, which ended up being a complete waste. He was a great guy, a nice guy, but we paid him for two months, and nothing that he built we used. So two months of his salary, which can be pretty material. It was probably $15,000 to $20,000, and all, with health benefits and all that stuff. Various other extraneous expenses.
We had actually gone to a conference, which is where we got our first clients, and that probably cost us about $5,000. Obviously, well worth it from a ROI standpoint but a cash expense. Then, these two ongoing salaries and then, to your point finally, the media expenses were just so high. We were so upside down that we started to blow through lots of money very, very quickly.
Andrew: How much money did you spend on advertising in the 45 days there?
Andrew: How much money did you spend on advertising?
Jesse: The first week we launched, that first week of April, we probably spent about $5,000 or $5,000, which keep in mind, generated four or five leads total, because we were paying $1,000 per lead. We said we’ve got to let it run. We’ve got to give it some time to do some work.
But we were launching that, and then we pulled back, and we said let’s wait for another week. Then we tried it again and the cost per lead dropped to $300 but we still probably spent another $5,000 or $10,000 all the while. There’s continued expenses, and people’s salaries, and so on and so forth.
Andrew: Okay. So you guys were close to bankruptcy, then you copy, you focus, things start to turn around for you. How do you build on that?
Jesse: Once we were able to actually have a little bit of success, I think one thing was obvious, we figured out the landing page optimization thing, at least from a basic level, which is this multi-step thing, borrowing from our competitors, and that started to allow us to expand our traffic over time.
Then it just became a function of, at that point, but just constantly looking at numbers, being more data oriented, and then finally, actually bringing in some of the really intelligent things we thought we were going to do in the beginning, which frankly didn’t happen for another probably three to six months. I don’t think we immediately were able to bring that stuff in.
We needed data, we wanted to just actually focus on the growth, and the growth was actually happening by hand-to-hand combat, it wasn’t happening by computer. It was happening by build more keywords, think about this, test this to see if it works, and just constantly, every day, getting more and more involved.
The other thing was, every time you bring on another client, there’s a decent amount of integration that was involved. So that would actually take a decent bit of bandwidth and the team, to kind of integrate that client, make sure their operating in the right way, and so on and so forth.
Jesse: I see. So wasn’t so much the software that was come up with the perfect keywords and the perfect process, it was probably you sitting there, personally picking out the right keywords, personally writing the right ads, and linking it to the site. As you were doing that, what did you learn about writing the proper ad on Google
Andrew: The hand-to-hand combat really works. I think you learn what gets someone’s attention, when to use an exclamation point, when not to.
Jesse: That’s what I mean. What gets someone’s attention?
Andrew: It’s just these little things that we’d read about, like selling benefits, not features necessarily. So when you say search through a thousand schools, people would not be that excited about that. You’re saying oh great, I have to go through 1000 schools, that’s not that exciting. When you say hey get your life together and fix stuff, that’s like the benefit, that’s what they’re going to get. Or get a better salary or whatever it would be that benefits, emotionally get people to click on things.
The other thing we learned, which I think is become really critical these days, is this whole quality score issue with Google is getting very specific around what your keywords actually are and writing ad copy that is very, very, nuanced to those specific keywords.
A lot of the times, and I can show you an example of a query some other time, but if you type in a query you can see Google folds the parts of the ads which are specific to your search, that draws somebody’s eye there. So when you have simple ads sitting there that is matching keywords that persons ad, the more specific parts of that specific query that you have, the better.
I’ll be more specific. If you do Associates accounting degree, there are people who are bidding on an accounting degree will also show up for that associates accounting degree, but it won’t say the word associates in that act anywhere. Because Google’s matching it to that. We make sure that, whatever that search is, you’re seeing exactly the specific things, like associates in accounting degree, and that drives people to click on your act more often than other people’s ads, which then drive through quality score of which lowers your CPC from Google.
Andrew: I see. All right, that makes a lot of sense.
Jesse: Writing that ad copy taught us to do that and those are things that, today, we’ve actually fully automated. I mean, we don’t automate ad copy because it’s impossible but we’ve fully automated the process of parsing keywords into very small groups so that someone can write very specific ad copy for those groups. And we’ve gotten a ton of leverage and success out of doing that.
Andrew: So, if I were to bid on Google for business degree, Google would automatically give me associate business degree right? That’s what you’re saying?
Andrew: But, the word associate would be bolted in the search results when someone typed in associate business degree, just business degree would be. You’re saying don’t buy these wildcard searches, be as specific as possible so that you actually get as many of the keywords as possible within that wildcard search.
Andrew: That makes a lot of sense.
Jesse: The Google language, to speak it for a minute, would say do only exact match searches and make sure that your ad groups and ad copy are relevant to those exact match searches. Don’t participate in phrase match or broad match, to the extent you can avoid it with Google.
Andrew: Okay. All right. I get it. I see what you’re doing here. You get this all right, and your growing and your learning, and you said you still went nearly bankrupt again. What happened the second time?
Jesse: As Wall Street guys, you would’ve thought that we had a little more sense about these things. May actually ended up being approximately break even for us. In his last two weeks we were able to generate enough PNL income statement profit to show that we were break even. We get to the end of the month and our balance sheet is zero, our cash is zero. We don’t get paid for 45 days from our clients.
So now we go, “Wait a second, we have to operate in June. We have absolutely no cash, because we’re not going to see this cash 45 days.” Essentially, we went bankrupt a second time. From success in this point, and at that point, the founders were able to, luckily, loan more money into the business just to keep things going for that month.
At that point, our cost base was extremely low, so we were able to kind of manage for that and we made it through June and then June sales doubled. Then we ran into the same problem because, again, we went collecting cash. By the middle of July, we actually got a factoring company in place, which essentially lends us money against these invoices that we have for all of our clients.
That was the initial critical turning point of being able to actually generate cash and having enough operating money to make the business work. Because we’ve done this really stupid thing of running our balance sheet to zero and, even if you’ve done that, and you’ve got to success, you still have the operating cash to run the business on a day-to-day basis.
Andrew: When you buy ads online, don’t you use an American Express card which gives you at least 30 days, and all the miles, but don’t you at least get the 30 days?
Jesse: Yeah, you get 30 days but if you do the math, and your client pays you in 45, you have those 15 days at the beginning of every single month where you’re in big trouble. And to be frank, to be clear, to solve the cash issue. And we were growing really fast. So we went from $40,000 in May, to $80,000 in June, to probably $130,000 in July. And when you’re growing that fast every month, it technically requires cash to grow that much.
I’m not going to get into the details of the accounting or the T tables, but the point is you need cash every month. So what we were doing was just credit cards. All of us had five credit cards open, all the founders. Taking them, charging all our credit cards up, and just carrying all the balances we possibly could. If we needed more, we could open up another credit card.
So there was a time where, in order to make the cash work, we were just using all credit cards and a little bit of personal cash that we had loaned in. But most of that personal cash was going to pay people’s salaries rather than going to fund the media costs. The media costs were sitting on various peoples credit cards, we’d collect our invoices, we’d paid on the credit cards, and we’d start charging them up again.
Andrew: How many miles do you have now from spending millions of dollars on ads?
Jesse: We have more every month. One of the cool things that happened to us in July, which has limited our miles, is that we got a plum card. This is a little plug, I guess, for the plum card. The plum card is managed to take our line from zero to $400,000 a month in the last one year, or the last eight months.
So the plum card doesn’t give you any points but they do give you the benefit of longer periods to pay back. And, if you pay back within time, they give you one and a half percent off your invoice, which is actually a pretty awesome benefit. So the majority of actual spending is happening on that card since then. But we probably have half a million to a million points, even despite having the plum card be the majority of our spending.
Andrew: The plum card from American Express?
Andrew: Okay. Cool. I now understand how you got here, and I’ll talk to you in a moment about beyond Google to, maybe, Facebook since that’s so new that a lot of what you did, people haven’t figured out yet.
But Google is a very competitive marketplace for ads. So many people are trying to go in there and figured out. The education space is especially competitive because there’s so much money in it. What are you doing differently from all these people who came in before you and had all kinds of experience ahead of you?
Jesse: This is a good segue into Jesse’s rules for being a good entrepreneur. The way I typically say it is it’s one third naïveté and two thirds dedication. I think, as a team, we had that, where previously in our college exploits, we didn’t have that. We had no idea what we were doing. We were so naïve about this that I said well, there’s so many million potential keywords out there, there’s got to be a small pocket that we can buy.
Literally we were deciding and going oh we can buy some clicks at a dollar, convert them at four percent, at $40 a lead, that’s going to make us profitable. That was the onus with which we started the business. That’s why we got it could be this cash flow business that would lead us into better things. We were just naïve. We said oh we can do it, it just makes sense, it’s reasonable.
I think the other two thirds of one third naïveté, two thirds dedication, I think, at that point, we just said we have to figure this out, we have to figure this out. It just became a thing where there’s tons of blocking and tackling. There’s no better way to describe it. Just on the ground, try this keyword, does it work? No, stop using it. Try this keyword, let’s try this list. Let’s try that list.
Over time of doing this, we learned what worked from ad copy standpoint, we learned what worked from a search standpoint. And we started to bring some more automation in, both technology automation and process automation, to how we did things. I don’t want to talk too much about it because some of it’s proprietary.
The point is that we go to the most detailed level we can when we build out search campaigns and keyword campaigns and we developed a process that makes that very scalable.
Andrew: All right. I’ve got to push you to be a little more specific there. Because if my audience is going to deal with all the tech issues that you and I go through as I record this and force them to listen to, we’ve got to give them something a little more than that.
I know that you can’t give everything, but give me one thing that you learned from all that blocking and tackling that you described that they could, maybe, implement.
Jesse: I think if you going to compete in a very competitive search market, I think it’s critical to make as small and narrow ad groups as you can. Try to get every ad group to 10 keywords, be incredibly specific, and make sure you’re writing ad copy that is specific to those 10 keywords.
Like we do stuff all the way to the level of therapeutic nursing degree, we’ll write get your therapeutic nursing degree and, I assure you, nobody else does that kind of stuff. They’re up on nursing degree or nursing college, which is far different from a search standpoint then therapeutic nursing degree.
Just get as specific as you possibly can. Be very patient with it, have some time and effort, and be willing to spend money and invest in doing it because it’s not easy. It’s something that comes over a period of time. Even investing in a search between when we started in April and December of last year, it was still the smallest thing we did. And now it’s actually, four months later, the biggest thing that we do.
Andrew: What was the smallest thing you did?
Jesse: According to our sources, percentage of the different places we were using, search was the smallest up until December of last year.
Andrew: So then, what was the biggest?
Jesse: The biggest was probably Facebook.
Andrew: Oh, Facebook. All right. Let’s talk about Facebook.
Jesse: [inaudible] but it was the smallest, as a percentage, of all the sources we were buying traffic from, yet it would require the most time investment from all of us. Over time, finally, after investing that much and that much now, it’s actually by far the biggest and the most stable out of all of them, is search.
Part of that is because a couple of competitors got on the market, part of that, though, was because we built up such a good process that we were able to get it to work consistently on an ongoing basis.
Andrew: All right. I want to talk about Facebook because you say that Facebook is today what Google was in 2003. I’ve talked to entrepreneurs who were on Google and 2003 and it was like a party. These guys were just raking in the cash, they were getting away with murder, and it was all happening with the rest of us not paying attention to it.
So, let’s get specific. When you buy ads from Facebook, how do you do it?
Jesse: Today we do it via a bunch of internal technology, and interfaces, and stuff that we’ve developed. But, when we first started back in May of 2010, we went to Facebook.com, we clicked on advertising, we looked at how the interface worked, and we started building ads. Just one at a time.
In the early days, it was just a question of how this is work? Is there anything we can do that’s going to make it work well? We also went to the process of trial and error quite a bit the beginning before actually getting to a place where Facebook worked well for us.
Andrew: Okay. So let’s go through that in the same way that we just did Google ads. And I know that we’ve gone way over time here, but I’ve got to dig into it. And then we’re going to bring you back and we’ll do a session where we just talk about this and show it specifically.
So, let’s start with you bought ads, what kind of text did you have on those ads?
Jesse: The funny thing is, in November of ’09, we were really early, we were just experimenting with Facebook, we tried Facebook. And then we didn’t really understand how it worked. We just targeted women between the ages of 25 and 35. And the funny story actually was, I remember specifically making an ad targeted to men between ages of 25 and 45 and the headline was go to class naked. And it had a girl, it wasn’t showing anything, but she was bent over and she was a little naked. And it said what other place can you go to school when you’re not wearing any clothes? Go get an online degree.
We had so much click volume come through that and not a single conversion. That was one of the first lessons of getting people’s attention is not always a good thing. That was our first foray into it back a while ago. Then, when we’d finally reset up our platform and gotten everything built, we tried it again probably in May or June of 2010. And there, we started to realize that the more micro-targeting we could do, the better.
So the point there was don’t try to go after millions. Facebook has 150 million people in the United States. If you build million person segments, that’s a lot of people. You’re not going to reach those people with any specificity. If you try to do the opposite and go after very targeted segments of people, you going to be able to reach people with exact messaging because you know them so well.
Andrew: All right, I’ve got to interrupt you. Hold the thought that you just had. Let’s dig into that. So, on Google, you target people based on keywords, and you go really narrow, and you write ads to that narrow target. Describe the same thing being done on Facebook where you narrow based on the person’s demographics etc. and then tell me what ads you show.
Jesse: I’ll speak it in a little bit more high level stuff because I want to be careful about any proprietary stuff that we do, especially on the education side. If you are Ford and you’re selling an SUV. With search you have to look for a keyword of Ford, or SUV, or the type of car I want. It’s what someone searching for, their intent is what you’re looking at.
The difference with Facebook, and what makes it such a unique platform, is it’s an audience based platform. So I can know, with specificity, who my audience is and then, based on knowing my audience well, I can target an ad for them, very specific around who they are.
To be specific, audience creation is the first step that we talk about. So if I’m Ford and I have an SUV, I want to go after mothers. So I’m going to go for people in New York, women between the ages of 30 and 40, who say that one of their likes and interests is my kids, taking care of my kids, my children. There’s actually about 50,000 mothers who say that on their profile on Facebook.
So it’s actually quite a few people. So, on the add, I’m going to say, like to have your kids protected? Get a Ford SUV, they’re very safe. Or looking to save money for your family? Get a Ford SUV, a hybrid SUV, they’re going to be really great. So now I’ve taken this audience that’s really specific and I’ve built a very specific ad that’s going to speak to them. And then, obviously, that’s the core of what makes Facebook work really well.
Andrew: How specific do you go?
Jesse: I think as specific as Facebook will let us. They have a creepiness factor rule were if you’re below, I think, 1,000 people, or 500 people, they won’t let your ad show because it’s too specific.
Andrew: So you’re going to keep trying to narrowing, and narrowing, and narrowing until you get to 1,000 people, or right about there.
Jesse: Our average segments are between 20 and 50,000.
Andrew: 20,000 and 50,000 is what you aim for.
Jesse: Because it’s not big enough to get some kind of leverage out of it otherwise.
Andrew: I see. I thought you were going for lots of different thousand persons.
Jesse: We do that and that’s part of the technology that we’ve built now. I could talk a little bit about that. I can put in every city, I can put it all these likes and interests, I can put in these age and gender groups, and then split it up into 100 audience segments with one button that lets us target them the most nuanced. And then bring in the creative elements, which is also [inaudible].
Andrew: [inaudible] very narrow target on the audience and their interests, and then it’s very narrowly written ad copy, so it feels like you’re saying hey Andrew and everyone else who happens to be in Andrews office. And then you addressing them by name.
Then you also talked about the picture. If it’s not going to be a naked woman, what is it going to be escorted get enough eye contact and enough clicks, but also increase conversions?
Jesse: There’s a number of things we tried, just simple things like putting a red outline on a box around the picture tends to draw peoples eyes towards it. So you want someone’s eyes to be drawn toward it.
Another thing that works really well for us is text. If you actually put text in an ad on Facebook as a picture instead of an actual picture, it ends up being the largest print an entire page, which ends up really attracting people’s attention. So those are two things, two examples of specific things we do.
But it’s all kinds of things you want to try, you want it to be a relevant picture, maybe of a mother driving a car with her kids, or whatever is relevant to that specific user, or might be relevant to them, we’ve seen to be effective.
Andrew: Facebook is more profitable for you then Google is right now?
Jesse: Than search it is, yeah.
Andrew: Than search. Wow.
Jesse: AdSense is a bit of a different animal but yeah.
Andrew: I see. What do you do with AdSense?
Jesse: I call it AdSense, it’s a misnomer, it’s the Google content network or the Google display network. It just tends to work better. It’s sort of on par with Facebook. They’re sort of cousins right. One is contextual advertising and one is audience-based advertising.
So contextual is I’m reading an article about accounting, well I’m going to see an accounting degree job, whereas we go after people on Facebook who say their profile that they’re clerks, or accounting clerks, or a junior accounting assistant. We also run ads to those people. It’s sort of the same animal, they’re sort of cousins.
Andrew: All right, I think I got a sense of how you got here. I think I understand why you’re growing so quickly because of what you’re targeting and how small you were when you got started. The interview’s practically over here. I’m going to ask you one more thing and then the interview’s over.
But before I get to that, one last thing. Is this something you can teach others? Can you take an entrepreneur like me, Andrew Warner, the founder of Mixergy who wants to promote his courses, can you teach me how to buy Facebook ads in a way that’s profitable on my site, with my situation?
Jesse: Yeah. When you say teach, I think it’s certainly something that we can help people with. Any of these things require process. I say process because beneath process, to me, is technology and just functional process of how a company runs. In order to do any of these things at scale, you need both of those things in order to do well.
So, if you want to spend $5,000 or $10,000 a month and attract users to your site who are entrepreneur specific, I could do a couple things. I could teach you and help you do that. If you want to start to pull in $50,000 or $100,000 or $200,000 a month, or more, that starts to acquire technology and process automation that’s going to make things a lot better for you.
Andrew: But on a smaller level, if I were going to be where you guys were about a year ago, when you say hey, forget the thousands of keywords, let’s focus on 10, and then tomorrow 11, and so on, you can teach someone how to make it profitably work.
Jesse: Yeah, I believe so. To be fair, we’ve only done it in a few verticals so I can’t tell you what works will always work everywhere. The one thing that I’ve learned about entrepreneurship and all this is that you don’t know until you do, because you can have this data, and numbers, and you can crunch the numbers all you want, but until you’ve actually done something, you have no concept of whether or not it’s going to work.
Andrew: I have a feeling that a lot of what you told me off-line and a lot of what you told me here would apply beyond. I’ve got to believe that a lot of people who are listening to us are going to now put images up on their Facebook ads that are text and test that. It seems like it will work for them. Same thing with putting a red box around it. And the kind of thought process that went into coming up with those two tactics, I know they’ll be able to use on other ideas.
But what about beyond the size? You’re saying wants you get to hundreds of thousands of dollars, you need automation, and you need a whole infrastructure. What about when you’re smaller and you have a things going on in your company? Can you buy these ads effectively without focusing on them the way you single-mindedly do?
Jesse: My argument would, obviously be, not as well as if you had somebody whose whole business was to do this. If you’re not in the business of online advertising, there are people, obviously, like us, who are experts in it and that’s where we spend our time. I think it’s possible.
But even guys like Groupon, who arguably customer acquisition is a critical piece to their business, they even hire other firms to run their Facebook ads and run other parts of advertising.
I guess the answer is, it’s possible, but it’s probably not the best. If you can find an expert, it’s probably worthwhile to engage that expert or to home grow that expert. As an entrepreneur in the startup world, we learned this stuff from scratch. If you asked me 18 months ago what a match type was, or click through rate really was, I probably had no idea.
So, the other thing is just finding someone in an entrepreneurial culture, home growing it internally. Long-term, that’s probably going to be a good thing for you.
Andrew: I see. So whoever it is that’s working in your company, give them this interview, let them listen to it on their commute into work, especially the second part of this interview, and then just keep flooding them with information on how to do this so they can learn it on their own.
If you and I do a course, they can take the course. If there are other courses online or, frankly, if they could just reach out to you, and talk to you directly, they could probably get some great information from you.
Your upside for doing this interview is potentially finding customers in this audience, right?
Andrew: Okay. Fair enough.
Jesse: Just to finish the story, we started back at the end of ’09 and this business, and we started doing lead gen in the education vertical, very focused in that place, to kind of build this business up that will generate cash, that will teach us things.
We were looking for this disruptive idea, we were looking for a big idea to go after and two interesting things happened. One is we realized that you can’t just start a cash flow business, that’s very challenging. So we started to pour all our efforts in and the education business has actually grown to be a pretty big business and a cornerstone of our company.
But the one thing that we did start to learn is this whole thing we talked about Facebook. When we realized that Facebook was like search in 2003, this could be our big opportunity. This could be the big thing we’ve been looking for that we would’ve never discovered had we not gotten into this.
So what we’ve done the last three to six months, is we’ve created another part of the business called Ampush Social, which is going to focus exclusively on Facebook ads and help advertisers get on Facebook ads.
That’s kind of the final part interview, I guess. The final part of the story. Now that we’ve actually found that disruptive, new idea that we think is really different, and that’s the Facebook ads business and what we’re going to build it on.
Andrew: I see. From an entrepreneurial point of view, that really is interesting. That you get into a business just to figure out the market, and to find that big idea within it. And if you do it the way you did it, by looking for a cash-based foothold in this space, then you could find your own growth going forward.
Jesse: And that’s kind of how we’ve been operating. That’s how the new part of Facebook we’ve been doing is operating.
Andrew: All right. Final question, one piece of advice. You given people so much information here, and I know that a lot of it is going to be useful, but how about saying one thing that they could do tomorrow morning that would help them get more results.
Jesse: You want a specific online advertising or [inaudible]?
Andrew: I love your online advertising. I could get entrepreneurial information from everyone. You’ve got this really niche piece of information that I can’t get anywhere else.
Jesse: Well, they’re the same. I just determine, in my head, that they’re the exact same piece of advice. Just do it. Just go out and do something. If you want to have success on Facebook ads, just open an account, put $100 in there, it costs nothing to get testing, and try it.
I think, even as entrepreneurs, it’s the same kind of advice. Just do. Do, do, do. Worry less about analysis, and coming from someone whose whole job was that for four years, just doing and actually getting something done to see how is going to work. That’s the best way you’re going to learn. That’s the best way you’re going to get results.
Andrew: What a great place to leave it. Jesse, thanks so much for doing this interview, for dealing with 40 minutes worth tech issues as we recorded this interview, and for the pre-interview, and for everything else. And for all the great information.
Jesse: Thanks so much, Andrew. I appreciate it
Andrew: The website is AmpushMedia.com. Thanks for doing the interview, and thank you all for watching.
SignalKit.com – SignalKit.com, simply put, makes your 37signals tools more powerful. Check out some of the products being created by SignalKit, a suite of tools designed specifically for 37signals users like you.
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.
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.