The company behind Silicon Valley’s dirty little secret

There’s a dirty little secret in tech, in Silicon Valley, and I bought into it.

The idea is that you use an app to take a photo of a receipt and it magically gets transcribed into the expense app you’re using. It feels like a machine but it’s not. The magic is that it’s actually people who are doing the transcription.

The guest you’re meeting today is a guy who runs a company that does this stuff for other companies.

Mark Sears is the founder of CloudFactory, which provides on-demand API driven workforces to scale your business in the cloud.

Mark Sears

Mark Sears


Mark Sears is the founder of CloudFactory, which provides on-demand API driven workforces to scale your business in the cloud.


Full Interview Transcript

Andrew: Hey, everyone. My name is Andrew Warner. I’m the founder of

There’s a dirty little secret of the internet of tech, of Silicon Valley, and I bought into it. The idea is that you use an app, you take a picture of something and magically it gets transcribed, like if you take a picture of a receipt, it magically transcribes into the expense app that you’re using, or if you record an interview like this and there’s someone who does a transcript, it feels actually like it’s a machine does it, like there’s some OCR that really takes that receipt, takes that piece of paper and transcribes it. But it’s not. It’s people. The magic is people.

And the guest who you’re meeting today is a guy who runs a company that basically does a lot of this stuff for other companies, companies that you know. You’ve probably used him indirectly and didn’t even know about it. His name is Mark Sears. He is the founder of CloudFactory. They provide on demand API-driven workforces to scale your business in the cloud. This is the company that so many of the businesses you know are using. And we’re going to find out how he went to Nepal, fell in love with the country and then that trip helped create this business.

And this whole interview is sponsored by two companies. The first will help you send out email that actually converts and gets you more customers. It’s called ActiveCampaign. And the second company will help you get your next great team of developers. It’s called Toptal. But first, Mark, good to have you on here.

Mark: Thanks, Andrew. I’m excited. It’s great.

Andrew: Can you throw out a couple of the names of companies that we might be surprised to hear are actually using you?

Mark: Yeah. I think there are kind of two modes. So I love magic–the magic is people. That’s great. And sometimes it’s humans at scale that are acting like robots behind the scenes, and sometimes we’re training up the robots. So, obviously, you mentioned things like receipts and invoices, so companies like Expensify,

We’re processing receipts, invoices by the millions behind the scenes, fairly real time. Other companies like Microsoft, and a lot of other great companies are building really cool ML algorithms, machine learning stuff. Again, we have humans at scale that are really training up those robots.

Andrew: How? ML is machine learning. How are you guys training their robots?

Mark: So the only way that robots really learn nowadays is by getting massive training datasets that really tell them what to do. So we literally–you look at self-driving cars, autonomous systems and that’s a really big vertical for us right now where we’re literally taking video from the hood, from six different sensors and we’re tagging pedestrians and intersections and everything to really train up those algorithms. That’s the way it works. It’s not kind of if-then-else type software code anymore. Most algorithms nowadays are getting trained up. It’s people that are doing that on an ongoing active basis.

Andrew: So a self-driving car doesn’t have software in it that automatically knows what’s a pedestrian and what’s an intersection. At least it didn’t know it until a human being saw a picture of what the car saw and said, “That’s a person.” And if you say that enough, then the software and the car therefore understands anything that looks anything like that is a person. That’s how you do it.

Mark: Yeah. It’s millions of those data points that basically they’re allowed–you talked about OCR, computer vision, all those technologies are actually first trained with human-created output. So we like to train up those algorithms and then also sometimes augment and kind of fill in the gap when they can’t quite get the level of accuracy they need.

Andrew: So you know my interviews and therefore I don’t think you take this the wrong way, but dude, what about Mechanical Turk? Amazon’s got tons of these people. What does the world need you for?

Mark: Yeah. They were a big part of the reason we started CloudFactory. I mean, ten years ago Amazon came out with Mechanical Turk really to fill their own need, right? There are massive amounts of product listings that need to be scrubbed, tagging things so their search worked better, all these uses. They did it for themselves and said, “Hey, let’s put it out there for others to use.”

Unfortunately, I think the world has spoken that kind of this anonymous crowdsourcing is not the future of work. There’s a piece of it. And that’s the piece that we’ve captured, right? So having the ability to have an API-driven workforce, send your work to the cloud and have it come back down in minutes absolutely is part of the future of work. But our big thing is the kind of anonymous crowd that has no training, no security, etc., etc. versus actually having trained, vetted, managed professionals that are really building up that workforce.

Andrew: These people work for you?

Mark: We have a combination of independent contractors and full-time people. So it’s kind of a hybrid workforce, but they do engage. So you can think more of like Uber, where there is more of a direct ability to engage. It’s not signing up online anonymous. You never meet anyone. Your account gets blocked, you just created a new account, etc. So it’s kind of enterprise-grade alternative to Mechanical Turk.

Andrew: Yeah. I don’t want to put down Mechanical Turk, but here’s the problem I had with Mechanical Turk. I used to have my transcripts done by Mechanical Turk. If you go back early enough, you’ll see that. I got really good at it. I created this whole system that would take my interviews, chop them up into bite size pieces, send it to Mechanical Turk for people to just transcribe a small piece, bid it out, the whole thing. And if you look at some of my past transcripts, you’ll see random stuff that people put in there.

And I didn’t have a system in place to catch it. I kept trying to come up with more systems to catch it, “Maybe if I send the same piece to two different people and if there’s a big enough variation, my software will catch it and say this needs to be looked at by a third person,” and then I would send it to a third person. And that’s what we were constantly doing with Mechanical Turk. They’re good people, inexpensive. But it’s hard to depend on it.

Mark: Yeah. It’s close, right?

Andrew: Right.

Mark: It’s the same as AI, actually, right? You think about it, it’s sort of that idea of getting 60%, 70%, 80% kind of accuracy or quality of work is actually kind of useless. In most situations, the difference between even 80% and 95%, 99% is the difference between having a real usable product, a real user experience. So that’s exactly what–that’s what we’ve built. You learned from–you want it to feel like AI. You want this idea of elastic workforce, global, all those pieces are there, but it’s kind of how you actually create that workforce is really, I think, the different take that we have on it.

Andrew: Yeah. It’s interesting. I always thought I was the one who was at fault. And you’re a guy who–I record these interviews via video. We also turn them into audio and then we turn them into a transcript. You’re a guy who reads the transcripts, who’s actually passed the transcripts to people on your team, so you know the value of it. For someone like you to see a little bit of a mistake, maybe you can look the other way, but if the whole thing is junky, I look junky because that’s your whole interface into me. That’s why I get where you’re coming from.

Mark: Yeah.

Andrew: Let’s talk dollars and cents, though. How big a business is this? What kind of revenues did you guys pull in 2016?

Mark: Well, we’re growing real quick right now. Our fiscal year is kind of messed up, but in terms of where we are kind of scale and scope, we’re at about a $5 million run rate. We just had a great quarter. We feel like we’ve hit an inflection point. We were about 40% last quarter versus the previous quarter. So we’ve had ten consistent quarters of growth. We started building this thing in 2010 and really focused on getting an engineering team, getting a platform, getting a product–

Andrew: We’ll get into it. I know you love the details here and everyone who’s listening loves the details and I want to get into that. But I’ve got to ask you something that I probably should have asked my college professors back when I was in school. Maybe I did and I didn’t remember it. What’s the deal with companies that have a fiscal year that’s different from the regular year? Why not start your year–

Mark: There’s a reason.

Andrew: Teach me, please.

Mark: There’s a reason. Yeah. So we are a micro-multinational, right? We are based out of Hong Kong with offices in Nepal, in Kenya, US and now we’re just opening up UK. I just got back two days ago. So, in certain places, it’s just–so, Nepal is kind of an offset. So we have a June 30th year end. So it was kind of triggered by certain places are not on a calendar year.

Andrew: What do you mean? Nepal doesn’t use the–I was going to say the American calendar–but the Christian calendar, is that what it’s called? They don’t use that?

Mark: I don’t know what it’s called, but it’s related to the lunar–yeah. It’s totally different.

Andrew: They use a different calendar?

Mark: They use a different calendar.

Andrew: You ended up using that?

Mark: Yeah. Well, we ended up mapping as close as we could to that, exactly. So it’s just kind of half-year shifted off. Yeah.

Andrew: Okay. This whole thing in Nepal is you and your wife, 2008, went to Nepal to do what?

Mark: 2008 was a two-week trip. I’m Canadian. Me and my wife got married in Canada. She got an opportunity to go work in Doha, Qatar in the Middle East. We said, “Hey, we don’t have kids yet. Let’s do it. We can go travel. Let’s do it.” Big part of that, top of the list, was getting to go to Nepal. So we had the opportunity, two-week vacation. We actually met some really good Nepali friends in Doha. So we went back to meet one of our friend’s family in the village.

It was just a beautiful time. It’s gorgeous. It’s the Himalayas. It’s a gorgeous, gorgeous country. It was kind of towards the end of that trip that I found myself in the village and then the contrast of being in Katmandu, third world, very urban city where I’m eating pizza in a café talking to a bunch of computer guys. I started finding out like, “What do you guys do? What do you guys make? What do you guys know?”

And I was blown away at the talent, right? Just a little naïve. I’ve done companies in Canada and the US and just was blown away. These guys were super smart. They were making literally almost nothing. So it was just like, “Wait a second. You guys are way too smart to be making that little money. There’s something wrong here.”

Andrew: I see.

Mark: So they wanted to learn Ruby on Rails. I extended my trip for about three weeks, bought an iMac. My wife’s like, “You’re going to have to sell the iMac.” But it’s not a big deal. It was a little bit of money. So it was kind of a cute pet project she thought at first. And then during that time, people were saying, “Where are you?” “I’m still in Katmandu and I’m trying to get these guys on Rails.” They’re like, “What? You’ve got Rails. Those are like unicorns. Can you do a project?”

Andrew: Oh, because you trained these people. So they said, “Can they work for us?”

Mark: Yeah.

Andrew: But wait, Mark, they’ve been trained for days, not weeks, not years, really, and they’re ready to be hired by your friends?

Mark: Not at all. Not at all. But that’s how businesses start.

Andrew: I see. They didn’t assume that these guys were as good as DHH. They just said, “Hey, you know what? You’ve got people who can do this? How about you train them up and we hire them?”

Mark: Yeah. I think our first project was literally at like $5 an hour. It was like $5 an hour, $10 an hour, $15, $20. . .

Andrew: What were you doing at $5, $10 an hour? What kind of coding work?

Mark: I don’t know. It was a couple web apps to start. Yeah. It was very small projects. But the idea was it was paid work. It was paid work, and it allowed us to kind of pay to train and start growing a team. We grew to probably about 12, 16 engineers and really it was a software outsourcing boutique, Ruby on Rails, building MVPs.

Andrew: And when you talk about $5 an hour, we’re talking about an environment–you told our producer that you had your first meal there in an open area. Can you describe what you saw and what this whole buffalo thing was there?

Mark: Yeah. That was one of those just like moments where you can like remember the literal like smell and the senses, right? It was actually back during that first two weeks where we were out in the village, like very, very poor. It was my good friend and I’m at his family’s home. They had prepared a meal. We’re literally eating outside right by the two massive water buffalo that they have. They give the plate and it’s like I’m the guest, so I eat. They had to like go to their neighbor to borrow a chicken because I’m a guest, important guest.

So I’m sitting there eating this rice and lentils. There are flies everywhere, and the buffalo crap is right there. I literally started shaking because I’m trying not to cry thinking, “I’m going to die.” It was just that moment of being like the naïve westerner that has no idea and you’re like, “Wow, I really want to see how I can help, but I feel so helpless. I’m a computer geek. What can I possibly do?”

There was such a visceral kind of feeling. So it was kind of from that moment to then being in Katmandu and realizing, “Oh my goodness, these guys are asking about Rails and contracts. I actually know something. Maybe I can actually add some value,” right?

Andrew: Yeah.

Mark: So that really was that contrast of like, “Hey, you know what? Maybe I can’t teach somebody how to grow things better, but there are a lot of really smart, talented people. I’ve received a lot of opportunities. If I can give some opportunities back, let’s do it.” So, yeah, it ended up being that project, a second project. We ended up staying there. Me and my wife had our two kids, so two kids, CloudFactory.

Andrew: You ended up having your two kids in Nepal?

Mark: Well, they grew up there. We lived there for six years.

Andrew: Wow. Yeah. I told you before we started your Skype showed that you were in Nepal. Like you were so connected to Nepal, that when I look you up, you’re the Nepalese guy.

Mark: It’s home for us. We’ve been here helping get the sales and marketing office into orbit.

Andrew: Here meaning you’re back in the US. What city?

Mark: We’re in Raleigh-Durham, North Carolina.

Andrew: Okay. And the reason that you had this experience is that you were an entrepreneur before. Is the company before this the one where you went to JavaOne Conference and you did the whole–which one is that? Is that Godspeed Computing or is that Sprout Technology?

Mark: No. That was actually Zucotto. So I came out of computer science, was working for Sun Microsystems. The manager there started a company. That was probably my two years, two and a half years of just crazy growth.

Andrew: 2000 to 2002.

Mark: Yeah, something like that.

Andrew: So what did you do in those two years? I think this story about JavaOne Conference will give us a sense of what was going on there at the time.

Mark: It will. It was the typical–there were five people in a room. We grew to 140 in like 14 months and opened up offices all over. Wireless Bluetooth, we actually had like native Bluetooth and Java, actually in hardware kind of thing and I was in product at the time and had the opportunity to do this keynote at JavaOne.

So we just had too much money too young. I bought this like $3,000 massive blimp, and we were going to like put our technology and have it like come over the crowd and it gets close and the Bluetooth connects and all these things are going to happen. It was this beautiful scripted dramatic thing until the fire marshal reviewed the fact that there’s actually flammable, combustible gas inside the blimp. So the demo didn’t go. I think we literally flew the blimp once.

Andrew: You already had bought it at the time.

Mark: Oh yeah.

Andrew: But you know what? I’m impressed you would even think of that. That is really dramatic. But at the same time, we’re talking about after 2000, after the dotcom market crashed, it’s a really hard time to spend that kind of money. You’re much more careful with your money now. You saw this opportunity, right? You told our producer, “I found developers get 6,000 rupees per month, which is the equivalent of $60 a month.”

So you said, “There’s an opportunity here. These people want more opportunities. The fact that they don’t live in the US shouldn’t hold them back. Let’s find a way to teach them. Let’s find a way to bring them into the market in the US and let’s find a way to open up the eyes of Americans and people in the rest of the world to the possibility here.” You were starting to train them. You told us you got to about 1,000 people. You were now a tech firm, essentially, right?

Mark: Yeah.

Andrew: That’s at least at this point in the story. Meanwhile, today, you’re not doing that. You’re doing simpler tasks. What was the period? What allowed you to say, “You know what? I have all these developers. The world is aching for developers, but I’m going to shift them towards a simpler task?”

Mark: So it was 2009. We’ve been doing this for about a year, year and a half. All of the strategies we were building MVPs for were asking us to do data work. They’re like, “Hey, can you research all this data on financial aid organizations in the US. Can you watch this video and tag every time somebody scores a goal or gets in a fight.” It was an NHL amateur hockey video.

So all these different things were very routine, repetitive work. So we started, actually, as the developers trying to use things like Mechanical Turk and other options out there. We’re like, “This sucks. We can’t do this.” So, at the same time, we saw the demand from the market. Everyone who’s building software needs data to make their business model work.

At first, I’m like, “We don’t do your data crap work. We are a high end Ruby on Rails boutique.” At the same time, on the other side, we’re living in Nepal. We’re hiring these engineers, but I’m meeting literally, like my wife is going to buy vegetables in the market every day and people are like, “You guys have a business. You must hire me.” And people from the UK, people that have massive education, super, super smart and just like 40% to 60% unemployment, like literally so many people that are idle.

So, yeah, the whole thing, the thesis for the business is that talent is equally distributed around the world, but opportunities are not. So how do we use technology to really flatten the world because we talk about the world is flat, but it’s not yet. It’s flattening quickly, but it absolutely takes a lot of innovation and technology and different models to really connect growing businesses to this global talent, like how do you do that at scale.

That’s where we pivoted. We’re like, “There’s a demand. There’s a supply. What’s in the market right now is fundamentally broken in terms of the thesis of how they do their workforce model.” So that was the opportunity and we pivoted in 2010. Everyone that was on the bench that wasn’t billable started building version one of the platform.

Andrew: Building the platform first and then doing the tasks?

Mark: Yeah. It was all about like 2008 to 2010 was building the engineering team, and then 2010 to 2012 was building the platform. We launched at TechCrunch Disrupt about 2012. And we hired our first 25 cloud workers. We call them cloud workers. We put them in teams of five. We were kind of learning from like the Grameen model, the whole microfinance, like solidarity of like how do you loan to people and actually get really high payments. You put them in teams and you get this idea of solidarity. You loan to a team and the whole team has to pay back.

So we built this team-based model, and we’re learning how to motivate people. So you go from being a tech company with a bunch of engineers to actually now I’m reading about behavioral psychology and organizational–breaking out my old textbooks that I thought I would never actually need again. That was 2012 to 2014.

Andrew: Before you go beyond that then, I’ve got a bunch of notes here. I’ve got notes to ask you about what was in the platform, about what you learned about behavioral psychology. I also want to find out about this thing you did. We talked about Mechanical Turk and how you have better quality results than they do, but still there’s this interesting thing that happened with Mechanical Turk in 2010 that I want to ask you about.

But first, I’ve got to tell people about my sponsor, which is ActiveCampaign. The reason I have to tell them about it is this. Right now, what many people do, Mark, with email is they send the same email to every single person on their list regardless of what country they’re in, what time zone they’re in, what they did, did they buy or not buy, did they click a link or not click a link. That is a huge insult to your audience and it also is a wasted opportunity for business.

So ActiveCampaign has come out with a product that says, “Look, we’ll let you add people to your mailing list on ActiveCampaign, but we will enable you to actually talk to them intelligently.” So, if somebody has bought, you could send a different message to them. If somebody has click a link and then another and another, you send them a different message from someone who’s never clicked a link. Based on what they clicked, you could decide what goes into the message.

If you write one email to everybody, maybe the PS line could be different for buyers or non-buyers, for people interested in one segment of your business and not another. That is dramatically, dramatically impactful for a business’ sales. They do some really cool things too. They you allow to A/B test emails with that. If you just collect an email address from someone, they can fetch the person’s age, their gender, their location, social interest, social profiles, all this stuff.

We’re talking about really sophisticated software that they are paying me to talk about for one reason only–people don’t realize how powerful ActiveCampaign is because ActiveCampaign is so simple. A lot of these features are new and integrated into their software and people who have been for years don’t recognize these features exist and these features exist in ActiveCampaign.

I really want you if you’re listening to me to go check out ActiveCampaign. They are going to help grow your sales by making your email marketing much more intelligent. We’re talking about, as I said, based on what people have done on your site, every interaction that you have with them can change.

If you’re listening to me, they have got a special offer. They are going to give your second month for free. They’re going to give you two one on one sessions with your people so you actually get to work with a consultant to think through and strategize your email marketing strategy. And they will even migrate you. So, if you’re with another software and you’re not happy with them, they will migrate you over for free, which is–frankly, it’s fairly easy to do on your own, but the fact that they’re going to do it for you is going to help make that transition a lot easier for you.

They’ve got so many other things that I can tell you about like CRM. They have workflow to enable your team to sell. I could go on and on and on with the list of features. But what I want to emphasize is this one thing. If you email your customers, you want to email them more intelligently and ActiveCampaign will allow you to do that and do it in a way that’s so simple that even your new virtual assistant without much training will be able to maintain your email communications.

For the rest, all the other features, you can just go to this special URL where they’re giving Mixergy listeners and unbelievable offer they’re not giving anybody else. The URL is–write this down because you’re going to need it for somebody else if not for you right now and frankly, you should check it out yourself to–it’s, really get to know ActiveCampaign. This little flow chart that they have on the middle-right of their page will show you how helpful they can be to your business–

Mark, the platform–what do you mean by that? As I was going through your service, I was trying to peak in and see what your team sees when somebody submits a task, but I don’t think I fully got it. Can you describe what the platform looked like version one?

Mark: Version one was really about how do you provide an API to clients so that they can send tasks into a system, and then how do you get those systems–how do you get that routed to the right worker at the right time and then how do you then get those results back? So it really was more about how do you just get a very tech-driven so we know where every task is at every time and it’s just an API-driven workforce.

Andrew: When you say to the right worker at the right time, do you mean if they’re actually sitting at their desk and don’t have a lot of other projects, you want to route it to them? If they are not there because they’re at home that day, you don’t want to send it to them. That’s basic stuff, right?

Mark: Yeah. Well, that’s really where we start to evolve. It’s more the right worker is who has the right skills. It was really this idea of okay, we’ve got tasks on one side and workers on the other. Tasks are coming in and you see who’s online, who has the right skill set, who has the right skill score if you want to send it to the best workers that are available.

A lot of it is definitely around deadline-driven turnaround time. If you’ve got five hours to do a task, you’re going to put a lower priority than if you have to get something back in five minutes. So kind of the right work at the right time, that real time matching and dispatching of tasks was a key part of the technology–

Andrew: You talk about Grameen Bank wanting to create teams of people who then get to borrow. I understand it because the borrowers have no connection to the lender, so why would they want to repay? Where’s the obligation? But if they have a team of people, they don’t want to disappoint the team by not repaying their loans and then it looks bad for the team.

In that context, it makes sense, but in your context, I’m not fully understanding it. Why can’t they just be responsible for the software and if they don’t do their work, you guys let them go or before then, they get an alert saying, “Hey, we’ve assigned you this many tasks. You’re not completing them or you’re not completely them right. Consider yourself warned.”

Mark: Yeah. I think we take two approaches. One is definitely using software and data and algorithms to know and measure how each worker is doing. That’s absolutely essentially. But if all you’re doing is just kind of policing, you’re missing out. So that’s exactly where we come in is saying, “We want to actually motivate and equip and train and invest and really have workers fired up to do great worker.” So, when they sit down to do a one-minute task, the level of ownership and care and pride they have in that work is the difference between 95% and 99%, which is the difference between us beating every competitor in the world at this.

So, for us, it’s that solidarity. So we’ve got team leaderboards. Every week, you know where your team is. When you get together, there’s a weekly meeting. Everyone gets together. They work remotely. Eighty percent of our workforce is distributed or semi-distributed, but then they come together and meet together with one of our–

Andrew: In person?

Mark: In person with one of our representatives. So you literally get that effect of, “Dude, what are you doing last week? You’re killing us. We fell from 134 as a team.” So there are motivations from that perspective. It’s just the whole, “Who’s my tribe?” What does the future of these gigs look like if we’re all freelancers that sit at home and don’t have any social contact? A job is more than a paycheck.

So we’ve been trying to design something where these people that are like friends. We’ve got people getting married that have been meeting through the company. There are lifelong friendships and other things that form when you put people together in groups. We have them go out and do community service projects and we do leadership stuff. It’s like how do you really design the future work from home kind of thing?

Andrew: I see.

Mark: We want to be the best gig and town wherever we operate. That’s Nepal, Kenya, etc.

Andrew: I wonder what else this team approach could actually work for. I’m imagining even online education would work better that way, where it’s not you interacting with the site. It’s you and your team interacting and who finishes an assignment or not determines how many points you get. How do you guys score people? You have so many different kinds of tasks. How do you assign points?

Mark: Yeah. Well, there are lots of different ways, right? Typically within our platform, we’re doing things like sending them known answers or gold standards. So, periodically, our algorithm will send you a question that already knows the answer to really test how much does it trust you.

We’ve got a reputation algorithm that does certain things like that or sends your work to a more senior worker for a review and all these different signals that we get that allow us to give you kind of–calculate a score, which is a skill score, but essentially it’s how much does our system trust you with these particular tasks. That’s one aspect.

The other aspect now is we’ve actually brought it up a whole new level because a lot of our clients are spinning up what we call a work stream, which is basically a remote team, like spinning up a server on AWS or spinning up a team as a work stream. They’re having our workers work on their platforms.

So we’re not just people sending us tasks. We actually have our workers now logging in working on a lot of our platforms. So that work doesn’t flow through us. And so we needed to find a way. How do we keep our reputation algorithms going? So we actually created a browser. So the work stream browser is something that all of our cloud workers use now.

So it’s just based on Chrome, based on Chromium. We’ve put some software in that actually literally looks at their keystrokes and looks at their mouse usage and even turning on the webcam looking at posture and engagement and facial stuff, like really simple signals that actually allow us to measure the engagement and the quality of work.

Andrew: But you see engagement that way. You can’t tell if somebody’s in my system and entering correct information or not. You just know that they’re fully entering information, right? Is there any feedback mechanism to tell you this worker did the job right versus this worker did it diligently?

Mark: Well, yeah. So we have that. That’s exactly–we’ve talked about training. That’s how we train our algorithms. We know the workers that do really good work. We have all the results and then we look at our best profile workers and that’s how we then train our algorithms to detect the signals and the patterns that really indicate when people are doing good work.

Andrew: I see. So, if someone’s logging into someone else’s website, you can’t get that feedback mechanism that says, “This specific task was done right,” but you can tell if this worker has all the traits of somebody who does it right. I see. Okay.

Mark: It’s even–early indications are that it actually can even be a better way to really understand how well people are working.

Andrew: Really?

Mark: Absolutely.

Andrew: If somebody is coming into my site and entering a transcript into my site, not yours, isn’t the best way to know whether it’s done right to actually look at the result or to see if I as the customer have kicked it back?

Mark: Yes. That would be the best way. But the problem, right, is all you can really do is a random sample. So when you’re doing literally millions of tasks every day, you can’t look at the results of every single one. So, typically, what people do and what we do is you take a random sample of those and that allows you to estimate the level of accuracy.

So, if there are 100,000 transcripts going through in a day, we’re going to look at a small percentage and guesstimate how overall it’s doing. But actually being able to track every single signal actually allows us not just to estimate and measure accuracy but to really get granular and to understand which workers are mapped to which transcripts.

So, yeah, it’s actually a different thing. The most power is of course when you put the two together. So you kind of capture this reputation signal, how much do you trust workers, quality at the browser level and you also have knowledge of the actual work and some of the other mechanisms you have. You put those two things together and that’s where we’re having fun. It’s a never-ending thing.

The technology of just managing a workforce at scale–we get to build a jet engine because you talked about building your own thing on top of Mechanical Turk, right? That’s what a lot of people do. It’s great when we can take that consolidated investment and really focus on solving that problem. That’s what we get to solve for our customers.

Andrew: Right. I shouldn’t have to figure out how to do that on Mechanical Turk. That’s not my business. My business is doing interviews. My business is selling courses. My business is not figuring out how to outsmart Turkers all over the world.

Mark: Exactly.

Andrew: Behavioral psychology–can you teach me one thing that you learned about behavioral psychology that’s had a big impact on you?

Mark: That’s a good question. The first thing that came to mind is actually related to the whole team thing. We did find out through–our data scientists came back. We had certain teams that were outperforming other teams. We weren’t sure why that was. Then all of a sudden we found what the correlation was. The correlation was that there was an existing social relationship between the workers before they came.

Andrew: I see.

Mark: So we had people that came in, we just put them in their own team. We had other people that came in and said, “Hey, these are my buddies from college,” or, “Hey, can my mom and my aunt and my neighbor work with me?”

Andrew: I see.

Mark: So we found that actually when you have some existing social relationship within that team, it actually does change the behavior. They were saying that they were more satisfied when we did surveys in terms of engagement. So they’re more happy. Their results are better. Yeah. It’s definitely a combination of using data to measure it but then to understand, “What is the behavior psychology that’s driving this team dynamic, this individual dynamic and how do we adjust our model and our technology to maximize that?”

Andrew: You built from 2010 to 2012. Even though Mechanical Turk is essentially a competitor, you guys built your platform on top of Mechanical Turk.

Mark: Yeah.

Andrew: How? What was the connection there?

Mark: So you had to focus on one problem at a time. For us, it was building a technology, but we didn’t have workers. So, initially our whole purpose was to try to create meaningful work for people in the developing world and really make this flat world a reality, but we had to start with getting a working platform. So we actually were sending all of our work initially to MTurk just to really get things working.

Andrew: You were Turkers on Mechanical Turk’s platform?

Mark: No. Actually, we were sending our work to Mechanical Turkers.

Andrew: And you were hiring them?

Mark: Exactly. Yeah.

Andrew: Why hire them when you had so many local people who were looking for work?

Mark: Initially, it was because we were a bunch of engineers that were just focused on the tech and then secondarily, it was like that transition. We had a 25 workers and then you have 50 workers and you have 200 workers. But then all of the overflow would be able to go. So we could take work from our customers and we do all of the work as much as we can and then we send the excess overflow to Mechanical Turk. So there’s a bit of a transition period until we can get a critical mass and get into orbit.

Andrew: I see. I remember Matt Morales–

Mark: Yeah.

Andrew: Was a guy who I’d known for a while. He was a Mixergy fan. He told me he was building something that would transcribe my interviews and at the time, I wasn’t making money with the interviews. So I didn’t want to spend a lot of money on transcriptions. But I loved what he was doing. It was called SpeakerText, right?

Mark: Yeah.

Andrew: And the beautiful part about SpeakerText, he had this great example on the website where I could not just watch a video and read the transcript, but the transcript would scroll with the video and if I tapped any section of the transcript, it would go to that section of the video. This whole interaction was beautiful. I’m only bringing that up because early on, you guys acquired his company. I can’t figure out why. What was the connection? What did you need?

Mark: That’s a great question. So Matt had a great team of engineers. They built a couple key tech assets. So they raised from Google Ventures, a bunch of other people. So they started with SpeakerText, like you said, but then they started building a more generalized technology similar to our platform. Yeah, we found a little bit about what they built and got very, very interested. We did the acquisition really for their tech assets. We took their technology. We took our version one of the platform, and we actually put them together for CloudFactory version two.

Andrew: Their software?

Mark: Yeah, their software. They did some really good things.

Andrew: Like what? Give me an example. The guy looks like a fun loving guy. He’s always a hangout guy. Frankly, if I just wanted to talk real with someone and also have a good time, he would be one of the guys that I would go to.

Mark: Well, it wasn’t Matt. It wasn’t Matt.

Andrew: I never got to the business part of it. What did he build that was so exciting?

Mark: I think frankly they didn’t know how to monetize it. There are a lot of other things going on, as there always are in startups. But they had some guys out of Carnegie Mellon that built some really cool things. So we talked about that idea of the reputation algorithm. We talked about that task dispatching.

They had some very interesting technology that just kind of it was like a perfect puzzle piece with what we had built. We built a lot of great technology to manage the workforce and do some other things, but we saw those two puzzle pieces and we put them together and that was version two of our platform. They’re great guys. They built some great tech.

Andrew: And now Matt owns a piece of your business?

Mark: No. It was just a straight up acquisition for cash.

Andrew: For cash?

Mark: Uh-huh.

Andrew: Wow. How much cash did you guys raise? How much did you raise?

Mark: We’ve raised about $7 million to date. We’re just closing a series B hopefully in the next few weeks.

Andrew: What was the first one and who was it from?

Mark: We did $500k was our seed round, ended up getting an extra couple $100k from a private equity firm actually here in North Carolina.

Andrew: Which one?

Mark: It’s called Sovereign’s Capital. It was a small fund, micro fund. I think their first one was $12 million. This one is $50 million. We were their first investment. Yeah. Our costs have always been obviously low. We started the company in Nepal. That’s where 80+ percent of our employees are. We’re not outsourcing to them. They’re outsourcing our sales and marketing to the US right now.

Andrew: What do you mean? Who’s outsourcing your sales and marketing?

Mark: I think a lot of people assume when you have offshore offices that it’s the US outsourcing there. It’s kind of this weird thing where we started the company, we’ve got people from different countries that have moved and live in Katmandu, Nepal. That really is our headquarters. The last office and our smallest office has been our sales office. So we’ve actually kind of outsourced sales and marketing to the US.

Andrew: I see. You’re saying, “We’re essentially an international company that’s outside the US. It’s not like we’re a US company that hired these guys outside to work for us.”

Mark: Yeah.

Andrew: It’s an international company that happens to have sales and marketing in the US.

Mark: Yeah.

Andrew: By the way, you mentioned your low costs. So I’m on your pricing page,, which, dude, the little thing on the left that I can click on the on the left and move around with all the people, I can’t stop playing with that. But the thing that stands out for me is I thought you guys would be super cheap. You’re not. At every level, almost every level, you are above minimum wage in the US.

Mark: Well, I mean usually what you’ll see is we’re about $5 to $8 an hour is kind of the range that you’ll see based on volume. So that $5 to $8 an hour, that’s everything in. That’s project management. That’s client success. That’s training. Everything is included in that.

Andrew: And if I ask someone to transcribe a receipt today and the receipt takes them just two minutes and they don’t do anything else today or tomorrow, do you round up to the nearest five minutes or the nearest hour or anything?

Mark: No. Absolutely not. So we do subscriptions that are basically subscribing to a block of hours, which I think what you’re looking at right now. It’s like anything, right? That’s kind of the starting point. A lot of our customers have grown much beyond what’s on the public pricing and they actually subscribe to a block of units. So a unit is actually a pay for results. So, when we get into a larger contract where we processing millions of receipts, for instance, they’re subscribing to a million receipts a month.

Andrew: On a per receipt basis and not a permanent.

Mark: Exactly. Yeah. So it’s expense per receipt. Even in that hourly model, that is captured like everything, to the minute.

Andrew: Okay. Your first batch of clients were people who came to you saying, “Can you do this for us? Can you go beyond coding?” And that’s where you got them, right?

Mark: Yeah.

Andrew: Okay. Before we go into the next batch of customers and I want to talk about sales and marketing, I have to tell people about Toptal. And you know Toptal. You guys have actually written an eBook, I think, about them, you said?

Mark: Yeah. They were involved in one of our on demand economy-type books. Yeah.

Andrew: The thing about Toptal is they recognize that there are some people that want to be fantastic developers, want to be tested and they want to focus on like that top three percent of people. What most people don’t understand about Toptal, Mark, is this.

There are a lot of development shops, a lot of creative shops, a lot of agencies out there that if you call them up and say, “Hey, can you do an iPhone app for me?” They say, “Yeah,” and they actually have an iPhone developer there. And if you say, “Can you do an Android developer for me, for example?” They’ll say, “Yeah, we have a developer for it.” They’ll sign your contract, but they actually don’t have an Android developer. They only focus on iOS.

For the Android developer, their secret is they go to Toptal and they say, “Here’s the project we have. Here’s who’s on our team. We need somebody or a team of people to work on this project to hit our numbers and to hit our deadlines.” And then Toptal sends them the developers. For many agencies, Toptal is that secret weapon, that secret sauce that allows them to act bigger, to sell more but be so lean that they only have an operation of in some cases literally two or three people, but they can then take on many bigger projects.

That’s what many of the companies that if you’re listening to me, if you’re hiring companies, there’s a very good chance that they are using Toptal to beef up their team. I had no idea this was going on until I had a conversation with someone at Toptal and this whole job was to interact agencies like this and to be the point person for them because there are so many who are doing it.

So, if you’re out there and you’re looking to hire a developer to help grow your project to focus on something and you don’t have the team to do it, you might as well cut out the middle man in some cases and go straight to Toptal. The cool thing about Toptal is you don’t just go on their website and sign up.

You actually go to their website and schedule a call with a real person. You tell that what you’re working on. You tell them how you like to work. You tell them a little bit about your culture, frankly, and what the output needs to be. Is it a long-term project, short-term project, whatever? I had them just do search. I said, “Look, here’s the thing. Search is driving us crazy. We can’t get out of our own heads about it. Can you bring someone that can help us think differently about it and help us rethink it?”

They gave us someone. We said, “It’s a good fit, but I don’t think it’s perfect.” They brought us someone else, we hired them, perfect fit within I thought would be a few weeks. It ended up being a few days. The person redid our whole search top to bottom beautifully. That’s the kind of thing that you could do to Toptal. You talk to them. They match you up. If it’s a good fit, you can often start working with them within days.

If you’re listening to me, they are going to give you something they’re giving nobody else and that’s because the founders of Toptal are long-term Mixergy fans and they want to support what we’re doing here at Mixergy. They got a lot out of it. They know you’re getting a lot out of it and they want to support what you’re working on.

So here’s what they’re giving Mixergy listeners only at a special URL just for us. You’re going to get 80 hours of Toptal developer credit when you pay for your first 80 hours. That’s in addition to a no risk trial period of up to two weeks. So, if you’re listening to me and you need this or you know someone who needs it, please refer them to the special URL and it is–top as in top of the mountain, tal as in talent, And I shouldn’t have to say this because you’re listening to me, but Mixergy is spelled M-I-X-E-R-G-Y. I had somebody complain to me yesterday that I got it wrong and I want to correct it.

Mark, how did you get your next batch of customers when it was time for you to actually go out and bring them in? Oh, you’re still on mute. I know you were muting so you wouldn’t cough into the mic as I was doing that spot.

Mark: Yeah. Thank you. You can picture me, a technology founder, sitting there in Katmandu, Nepal, trying to sell over Skype. So it wasn’t easy, but that’s reality when you’re getting going. The first wave is your network. The second wave, it was just scrappy, just trying to find different companies that were fast growing. We did different things like looking at people using Mechanical Turk that were unhappy.

Andrew: How? God I wish you would have called me. I had no idea this was even possible back then.

Mark: Yeah. It was totally not scalable. It always is when you get your first wave before you hire sales. So it’s looking around online. We were able to find people that were using Mechanical Turk, looking at forums, people that are unhappy.

Andrew: I see. That helped you do it. What else did you do to figure out who could potentially want to work with you?

Mark: I think it really moved–it felt early on it was a little bit more like a needle in a haystack because you have to find exactly the use case. What we found was actually just starting to have conversations with venture-backed startups, you realize that, “They’re just looking for a better way to get this work done.” I need an easier way to get repetitive work done.

They’ve already tried outsourcing it, crowdsourcing, hiring interns, freelancers. They’ve tried all these different things that aren’t working. We started having conversations and being like, “What can do you do with thousands of college educated people on the other end of a browser? You guys are growing. You’ve got tons of routine, repetitive work.” Everyone’s got routine, repetitive work.

So it was in the midst of those conversations that we just started to get like, “Yeah, we need somebody to do this. We need somebody to do that.” So it was really an inbound conversation of like, “Here are use cases. Can you guys do this?” We started doing more and more of them. We started getting more tech and getting trained workers and just got better and better.

We’ve been going after a lot of different what we call kind of micro markets, so processing millions of receipts, millions of invoices, tagging tons of images for self-driving cars autonomous systems, all these different kinds of things where we’ve gotten really good at dominating one by one.

The way we’ve done that is by having conversations with mostly tech companies that are growing and just like you mentioned Toptal, a lot of people are using companies like Toptal to–call it agile talent. How do we focus our core strategic work and then augment that with these types of on demand online work companies? So, I’m going to get some developers from Toptal. I’m going to scale up the operations and data work with CloudFactory. So, it’s been kind of that process.

Andrew: So a lot of the conversations were you saying, “I have all these smart people. They can do work for you. What kind of work do you have that’s repetitive tasks that we can assign to them?” And they were brainstorming with you. Is that right?

Mark: Yeah. Exactly.

Andrew: How did you get these people to be willing to brainstorm with you? You’re talking about venture-backed companies. They have to move fast. They have to grow quickly. They have no time to brainstorm with anyone. Why you? How did you get them to do that?

Mark: They’ve got time. They do have time because this is the work that everyone is trying to get off their plate so they can focus, like you said, like, “How do I innovate? How do I grow?” Well, you’ve got this mass of stuff. I want to sell. Well, I’ve got tons of data that I need to get enriched and cleaned. So it’s actually amazing that it’s a big problem. It’s a big problem for fast growing companies is I’ve got tons of stuff that needs to get done and there’s no good way to do it.

So it was amazing, but people are really willing to listen and give things a try. That’s really what’s happened. A lot of people, we make it really easy to get going, spin up a work stream, get going right away as little of 200 hours a month is like $1,800 a month. So I could hire a full-time employee, or I can subscribe to CloudFactory and get people starting to do some stuff and it works [inaudible 00:48:29]. . . my business process. And so it was all about like removing the barriers, reducing friction, getting people going, getting them trying different use cases. It’s been that rinse and repeat. We’re at a different place now, but that’s really how it starts.

It was like finding exactly who may be doing a perfect task, that’s a needle in a haystack. We just went to who are our early adopters. Who are the people that are just looking for a better way to get this work done? They don’t want to do old school outsourcing. They don’t want to–they’ve tried Mechanical Turk and it failed. That’s actually been a majority of the venture-backed startups we’ve talked to.

Andrew: I love that when you talked to our producers, they take these notes and they feed them to me. Here’s one thing they wrote from your conversation with our producer. Strategic shift happened when you went to Kenya. You locked yourself in a room and you went through all your customers. What was going on there?

Mark: You know, as a founder, that was one of those crazy moments where I feel like I’m always going into the meetings and you have to have like a plan. Even with your own leadership team, it’s like, “I have to have a plan.” I went in with a plan. We literally came out with something completely different that I never thought of. I had this like physical like–I was like shaking and sweating. It was crazy. I would never use the word pivot because that would be really strong, but that was a really big strategic shift for us.

Andrew: From where to where? Where were you before and what happened after that?

Mark: So, a year ago, if you were a CloudFactory customer, you were using our platform. So, if you built out on Mechanical Turk, if you built your own tools, whatever it was, you were going to have rebuild on top of CloudFactory’s platform. So, as the technology founder, I’m like, “Dude, we’ve built the best like jet engine. Why would anyone build their own thing? It’s ridiculous. We’ve built the best mousetrap. It’s unbelievable.”

Andrew: Right.

Mark: And everyone was like, “Oh, yeah, that’s great.” But when you’ve invested so much–even you said, you invest so much to get things working on a particular platform, the cost and time and risk and everything to move was really hard. So we had a lot of people like, “I love you guys. We need an on demand workforce, but we actually don’t need your tools. We don’t need your platform.

So we started experimenting. We lost a couple really big deals, like really big deals. I felt like I’m the CEO and how did we just lose top five tech company who loved us, wanted to access our workforce but didn’t want to do it through our platform? So we started just iterating with some trials. So we sat there in Kenya, everyone coming from Nepal, US, Kenya, executive teams, strategic planning. We came out the other end looking at the data and saying, “Oh my goodness, this thing is exploding.” And the sales team is saying, “We need to shift.”

So now customers, some of them are taking advantage of our platform, especially if they’re just getting started, if they have high volume and efficiency is the most important thing, but a lot of people are just getting going with a work stream and subscribing to a block of hours to have people come and work in Google spreadsheets, work in their platform, whatever that is, right? And we realize fundamentally it’s like, “Who are we? Are we selling a platform? What are we selling?”

So, for us, it’s like we’re building the world’s workforce. We see this like cloud computing is cloud labor. So AWS goes to strategic locations, builds these low cost data centers in low cost areas. They put virtualization software to basically rent out slices to the world. We’re doing the same thing. We realized that’s what we do. Our business, our workforces are data centers.

Andrew: People. That’s why I like even on your website you say spin up a team of people or some variation of it. And this explains why when I was–I always use SimilarWeb to kind of hunt through people’s traffic to see what’s going on. A lot of your traffic comes from one login and now I understand why. What seems to happen is you guys one login to allow your team of people to access your clients’ websites without handing over a username and password that you can’t revoke.

Mark: Yeah. That’s funny. Exactly. Again, security is a really big thing for us. So for us to be able to ensure that we know exactly every worker, how they’re accessing, what they’re accessing, be able to turn that off, it’s been important for us, but yeah, that’s exactly it.

Andrew: What if it’s not repetitive tasks? What if I have here at Mixergy like I want to do research on every customer and it’s not repetitive. It’s a research-based task. Is that something you guys do?

Mark: It is. That was a part of the shift, right? We basically have opened up the ability to subscribe to the number of hours and we’ve got kind of a Slack-like channel-type system. So you’re directly connected to your team and you’re able to drive that team, kind of Trello board and Slack.

So we do have people. We kind of call it repetitive and routine. So routine is still in terms of skill base, like we’re not doing programming and graphic design and a lot of higher level skill stuff, but yeah, things like research and kind of one-off projects. We have clients that are driving their remote teams on a day to day basis and changing their priorities. So it’s not always just sort of click, click doing millions of the same task.

Andrew: But it could be like, “Here’s the new set of customers. I need this information on them. Give me their Twitter accounts, their Facebook accounts,” more importantly. “Tell me school they went to, what they like, what business.” I got it. Okay.

Mark: Yeah.

Andrew: You mentioned Kenya. I’ve never met a guy who’s been to so many countries who’s running your business–every time we talk, it feels like there’s another country. You’ve got the UK, the US, where you’re talking to me from right now. You meet in Kenya. You started in Nepal. Cultural intelligence is an issue. You’ve had issues with it before, right?

Mark: Yeah.

Andrew: What’s this one example about family doing harvest? I don’t even understand it and I understand that you’ve had an issue with this?

Mark: Yeah. So cultural intelligence is one of our six core values. The reality of working across different cultures like we do, it’s super, super important. So, for me, again, young Canadian guy trying to live in Nepal, it was time after time of sticking my foot in my mouth and not understanding.

I think the time that I was talking about was a stand up meeting early on where I wanted to celebrate one of our employees because I thought he was doing a really cool thing. I was like, “So and so is taking a couple days off this week, so he won’t be in. He’s actually going to be helping out his family with harvest.”

I was like “That’s awesome. That’s a good value. They’re using their vacation not selfishly, but they’re actually going to help their family. That’s really cool.” So I thought it was a great thing until he came up at the end of the day to my desk and quit and was crying because I had shamed and offended him because the idea that he wasn’t able to hire people to help the family was embarrassing.

Andrew: I see.

Mark: It’s like a known sign that if you have to do that yourself, that it’s–I just had no idea. I had like shamed him in front of the whole office.

Andrew: What can you do about that? How can you possibly know all this stuff and the more countries you get into, the harder it is. What do you do about that?

Mark: Well, cultural intelligence, just like emotional intelligence, it’s something you really can develop. That’s something we do spend a lot of time, more than the average startup, for sure, spending time on. You can. It’s things like–

Andrew: How? How do you do that? how do you train someone to know this is a problem? I would never know.

Mark: There are always going to be mistakes, right? It’s like anything. Learning how to ride a bike, you’re going to fall off. It’s going to happen, but you want to try and reduce the amount of times you fall off. So, yeah, a lot of it is–

Andrew: How do you do it?

Mark: A lot of it is just, again, curiosity. So we share different meals. So a Nepali meal will be served in all of our offices or a Kenyan meal will be served in all of our offices. We celebrate different holidays and share stuff back and forth.

Andrew: I see.

Mark: We just have open conversations. Again, that team model for our cloud workers, we actually do that internally. So we go through different kinds of leadership and character principles and things like this come up. We talk a lot about the differences and just force people to rub up against each other and all of a sudden, we’re all learning.

I mean, for me, it’s like one of the best things about what we’ve been doing for the last year. It’s changed my life, my family’s life. It’s amazing. Certain people love it. Certain people just–it’s the last thing they want to do is get on another plane, learn a language, learn another culture, but I think people at CloudFactory, that’s one thing in common is curiosity.

Andrew: How many languages do you speak?

Mark: Oh me, English would be the only one. I speak Nepali but really bad. We had to learn it obviously to survive.

Andrew: And now you’ve got so many people working with you that you guys have a whole bus, apparently, just for your team at times?

Mark: There were a few buses. We located our office in the suburbs of Katmandu, so just on the outskirts. There’s nothing there. It’s just small little stores with like one person working in it and then we’ve got a company with hundreds of people that are coming every day to work and get trained. It was really cool. We see coffee shops and restaurants. There was nothing there.

Andrew: Just for you?

Mark: Well, not all of it is just for us. I’m pretty sure a couple of the coffee places was pretty much for us, but just even the bus lines.

Andrew: A city bus line, just to be clear. We’re not talking about you chartering a bus. You’re talking about the city realizes, “There are so many people going to work here. We need a bus to take them in.”

Mark: They’re basically private. They’re ran by the city. But it’s other entrepreneurs. It’s just that ripple effect. That’s what we’re going for. You bring work, digital jobs by the internet into these places where they’re just talent hot spots, super smart people and all of a sudden, we’ve had over 7,000 people that have been earning money from CloudFactory and there’s a ripple effect that happens when you’re inserting earnings directly in the pockets of people.

Andrew: Let me close out with–well, maybe this shouldn’t be the last thing, but one of the last things. There are other people who do this. I interviewed Bryce Maddock.

Mark: Yeah, Bryce.

Andrew: You know him?

Mark: I know Bryce. Yeah.

Andrew: How is his business different from yours?

Mark: I have a great moment remembering Bryce. We met a conference, CrowdConf many years ago in San Francisco and both were pretty early getting started. I think that they have taken a very people-only culture-based approach. So they’re based in the Philippines and they work on more higher level work now. So they’re mostly doing call center-type work. But very similar, they targeted venture-backed startups with the idea of like, “We just need to build a better culture.” Outsourcing is so old school, and a lot of tech companies don’t even know how to engage or speaking two different languages.

I think that’s one thing we have in common is we’ve both targeted the same kind of startup tech company. They’ve done it again with trying to build a better culture in the Philippines so they’re getting better talent than the average call center there. I think we’ve tried to do a lot of the same thing with culture, but we’ve also paired it with technology.

I think that’s kind of the key thing. We’ve invested a huge amount of time and money to build technology to facilitate a lot of it. I think a lot of people would kind of say they’re doing higher level work, definitely a lot higher cost and we’re doing more data, repetitive back office type stuff now. That allows us to bring a lot more technology into it.

Andrew: They used to do it. I used to see on their website, for example, receipts featured pretty prominently. When I interviewed him, that’s what I saw. Now they don’t. Now when I look to see how people compare the two businesses online, you guys are basically like AWS, like the spin up people fast. You even have prices on your website. The API is featured pretty prominently. Their site, it’s like a, “Contact us and we’ll talk about how we can do business together,” as opposed to just spin u p a team of people.

Mark: Yeah. We’ve tried to productize. That’s been a huge push this past year. Our sales cycle has gone from about 150 days down to 15 days. So, it’s gone 10x in terms of accelerating the sales cycle. A lot of that is, again, we’re using technology to make an online facilitated sale, just productize stuff as we can. We’re using tech to make it easier to buy, tech to make it easier to send and receive work via API, tech to manage the workforce, tech to manage quality control of the work. So, for us, again, a lot of it is the roots of the company. We’ve got over 50 engineers that have been working of–

Andrew: Right. You guys started as an engineering team.

Mark: Six years now, right?

Andrew: You know what I wish I had done? I’d love to know how many of my listeners end up being customers of yours. I don’t want to use like an affiliate link or a special URL all the time, but man I’m so curious.

Mark: We can do something. You inspired me with the–

Andrew: With Acuity Scheduling–I mean not Acuity Scheduling. Acuity Scheduling is how we booked the interview. You’re talking about the ads that I did.

Mark: Yeah. We can throw something up.

Andrew: Let’s do it. I don’t want to get a commission. I’ve never gotten a commission from my guest. The reason I do it is because I don’t want you to hold back on what you say here. I don’t want to be worried about talking about your competition. I don’t want us to be in league together. I want to do the best interview and that’s my number one priority. But I am curious how many of my people end up going to CloudFactory, how many people end up coming up with an ideas. So you have an idea for what we can do. What do you have in mind?

Mark: So this is definitely winging it. But we’ll do We’ll throw up a landing page right after this. I’ll get the marketing guys to put something up. What we’ll do is we’ll do for the first six months–so anyone who spins up a work stream, we’ll do it for this month, for February, anyone who spins something up in February, we’ll give six months at 25% off.

Andrew: Okay. Let’s do February and March because I’m not going to publish it until–let’s do it for a month after the interview is published. You and I are recording, just so everyone knows, we’re recording February 1st with the idea that it’s going to be published a little bit later in the future. All right. That sounds good. I’ll follow up and find out how many people ended up signing up and if it fit or not.

Mark: We’ll let you know.

Andrew: Cool. if you’re curious about it. I’m wondering what you guys will end up doing with it. I’m really fascinated by the business. And of course the two sponsors that I had, the company that will help you hire incredible developers. We’re talking about Google-level developers. It’s Top as in top of the mountain, tal as in talent, for that special offer. And if you want your email to actually go out there and bring in customers, we’re talking about heavy duty email that will work with well that’s stunningly simple, go check out I’m grateful to both of them for sponsoring.

I’m glad that we had you on here. Thank you so much, Mark.

Mark: Yeah. Thank you, Andrew. That’s great.

Andrew: CloudFactory. Impressive. Bye, everyone.

Who should we feature on Mixergy? Let us know who you think would make a great interviewee.