Everything you wanted to know about API but never asked, with Augusto Marietti (Mashape)

What are APIs?

Why and how you should develop your own API strategy today.

How an API works? What are the most interesting trend and tools?

40 minutes of pure learning with Augusto Marietti co-founder of Mashape the Cloud API Hub!


About Augusto Marietti:

Augusto (born March 12, 1988) is an internet entrepreneur, best known for co-founding Mashape, the marketplace for Cloud APIs.  Mashape raised $1.6MM in funding from prominent institutional investors including New Enterprise Associates (NEA), Index Ventures and individuals including Jeff Bezos and Eric Schmidt‘s Innovation Endeavors

Raw Transcription:

Q: Hello everyone. Marco Montemagno here, the Tech Alchemist, and here with me is a good friend of mine, Augusto Marietti, co-founder of Mashape. Hi Augusto, how are you doing?

A: Ciao. Hi.

Q: So, my idea today with you was to try to talk about this magic world of API. Everyone’s talking about API now, but I’m fully convinced that 99% of traditional businesses, of course, they still don’t get it. Probably because it’s not so clear what it’s all about, what the opportunities are. So I would like to go a little bit deeper in this API world, and then I have tons of things and questions to ask you about yourself and your vision of the future and so on. So first, where are you located now?

A: I’m now in San Francisco.

Q: All right. Great. And…

A: In South Park.

Q: By the way, Mashape, three founders, two founders, three founders?

A: Yeah.

Q: Yeah. Okay. Three founders.

A: Marco, Michele, and me.

Q: Excellent. And you got funded last year. I just don’t remember exactly when, but I think it was last year, something like that.

A: Yeah, we had the first venture around in 2010, two years ago. And then, a seed around last year, that most [inaudible 01:40] noticed.

Q: Alright. And, by the way, the round for people who don’t know Mashape, just go and check Mashape, and you’ll see the round was with excellent investors, Jeff Bezos and other guys like that.

A: Yup.

Q: So, Augusto, let’s try to go straight to this API topic. So, first of all, when you say “API,” what is it all about? Explain to a human being and not to a tech coder that obviously knows what it’s all about. But if you could just say, “Okay, API, we’re talking about that.”

A: So, first example, API has always existed in information technology. See, the problem in this new revolution of API is that we are talking about Cloud APIs. So it means APIs that run over the Internet, and they run in the Cloud. So they did this to build services, data, whatever.

But API as a concept already existed. A simple example could be Windows 95 when it came out. They already had APIs, but it was not on the Cloud. For example, Skype, the Skype that we are all using, used the Windows API to actually build on top of Windows, right? But those are the old way of APIs. But they really always existed in information technology.

This new way is actually Cloud APIs, and we make a distinction, because there is a server, somewhere, and there is the client, the other part, and then they ping the server, worldwide, wherever the IP is, to access content and data and services. And those are the new revolutions of Cloud API.

So, technically, API means Application Programming Interface, the interface where you can program applications, or whatever. In simple words, it’s like a Lego box. It’s like a Lego, or a puzzle. So the API is just a piece of the Lego, a piece of the puzzle, and you can use the Lego you want to build your infrastructure, your project.

So for example, you want to build an application on the iPhone, and you need SMS notifications for some actions. You don’t want to build your entire SMS infrastructure that’s [inaudible 03:56] application [inaudible 03:57] two or three engineers. What you can do is you go out and you look for an API that actually can send SMS for you. So you search for this API, you integrate this SMS component API into your project, and then in a couple of minutes, you can basically have an entire SMS infrastructure in your app and send SMS notifications or push notifications without reinventing the wheel on your own, and that’s actually the advantage, because it’s basically pushed down cost and time, as well. And small teams are able to create amazing things way faster and way cheaper because of the API.

AWS, Amazon’s, is probably one of the biggest examples. People don’t have to go out and buy and build a server and keep it in their house or their office, and spend a huge amount of money to manage that real, physical infrastructure. They just go to Amazon, and via their API, they have storage, they have easy [inaudible 05:00], they have bandwidth, they have everything through APIs.

Q: Excellent. So basically, you just choose the features that you want, the Lego blocks that you would like to add to your service, and then you just get it and you include it in your offer. One thing, I’m starting really at the basic level.

So imagine that there is a company, and this company says, “Okay, I get it. I’ve been watching Augusto Marietti explaining the topic, and I think that I will go for it. I will build an API.” Today, to building an API is still developers’ stuff, is it coding stuff, is there a lot of coding, or is it possible for human beings, for companies to create and develop their API the same way they can open a blog, or they can have their own social network, or features that once required a lot of coding? How is the situation at the moment?

A: There are, coming out, a lot… probably two or three tools that allow you to basically have an API if you upload your code, base, [inaudible 06:15], and they’re reaching, so they’re not really trustable 100%. But the idea is the developer that wrote the code of your app, also is the guy that is going to publish or have an API on top of that.

So it’s pretty technical to publish and create APIs. It’s actually part of the code. If you create a code… the concept of APIs is actually that there are, for example, SOAP APIs, the old way of enterprise, and now there’s the RESTful API, which is a more simple way to develop and to access those services and that data. And the idea of the API is that they are stateless.

So what does that mean that they are stateless? Once you have an API, and why you build an API… you build an API because you want to reach as many people as possible, and an API is actually shapeless. So that you can use an API if you have an iPhone app, or an iPad app, or a TV application on Google TV, or a website, right?

So the powerful APIs are built once, and they’re good everywhere, and they can be used for every scenario because they are just the fastest way to transfer data and services now in the world. It’s like the veins of your body, right? They transport blood, and your blood can be your services or your data, in a very fast and easy way, wherever the company is going to be: mobile, not mobile, TV, whatever. So that’s why they’re building that, because it’s way cheaper to than reinventing the wheel and do something from scratch and custom for the iPhone or the iPad or whatever.

There needs to be a really big distinction between private APIs and public APIs. So private APIs are APIs that you use internally. So, for example, when people say there are 1,000,000 mobile applications, so let’s say iPhone and Android, for example. That means that implicitly, there are 1,000,000 private APIs behind that. So that’s the underlying architecture. And some of them are also public. Public means that third-party developers can access them and use them. Private means only the company, or the partner you want to. So for example, Netflix, which is not in Europe yet, but it basically distributes on-demand…

Q: I think it’s coming, by the way. In one month, it should come.

A: Good. It’s basically driving on-demand film and movies through your TV, via the Internet, through the IP. They distribute all of these APIs, in more than 260 devices worldwide, all through APIs. And they have a private API, which is basically the one they use in China with the team, a partner API, which is kind of private and public, so they give access only to big companies, like Apple, when they did Apple TV, Google, when they’re doing Google TV, so they can have a better relationship.

And then the completely public ones, for the long tail of developers to kind of keep on top of that. But bear in mind that nowadays, the way you build software is changing, and it’s becoming API-centric. So the first things you build as a software engineer are an API, private, and then on top of that, you build the clients who want iPhone and iPad on the website. But the first thing you really build is a core code base, and then a private API. And then obviously, you can integrate third-party APIs into your product to get done faster, but the first thing you have is an API.

Q: How about… Okay. If you have to use APIs from other companies, fine, you just go there and you choose. You did the example of, I don’t know… phone API, and I’m thinking about Twilio, for instance, so there are companies that give you this service. So the old infrastructure is on their own cost, and is their problem. That’s it, you just use their information and you just create your stuff.

A: Right.

Q: How about if I’m a company, and I want to develop my own API? What kind of infrastructure do we have to build? Is it something massive, monumental, or is it something that I can handle in an easier way? Because I think a lot of a companies have the fear of building their own API, because they will say, “Okay, how can I handle all the requests?” and so on. So what’s the situation at the moment for you?

A: Building an API is actually, now, pretty quickly, just RESTful, or [inaudible 10:42]. But the problem is for the rest, so as you say, the entire infrastructure that you want to build on top of that when you decide to make it public, because you need to have permission, you need to have the [inaudible 10:44] developer, you need [inaudible 10:45] developer analyze, with an analytics developer, because there’s going to be a lot of things that you need to publish your API to be sure that it works well with security and things.

So all of these things, you don’t have to build them, you can actually access a third-party company’s, and have this already done for you, and then use that, and you can be up and running quickly without having resources, or engineers that [inaudible 11:20] infrastructure to make an API public in the good way, and Mashape is one of those, for example.

Q: Alright. Can you give us more examples, real-life examples, of where I can use an API, what I can use it for, and maybe, also, if you think that there are areas, sectors, industries, where you wouldn’t recommend using an API, if any?

A: So, for example, there are becoming API’s in every sector, every market. [inaudible 11:53], for example. So even Starbucks, even companies that are not related to technology are actually now publishing APIs.

An example of an API, for example, there’s one coming out which is called EasyPost.co, and it allows you, through simple APIs, to send packages worldwide. So they basically built this API on top of the complicated infrastructure of UPS, FedEx. They used their old APIs. On top of that, they’ve made EasyPost API, which is pretty simple.

So every kind of developer, for example, has marketplace needs to ship things. As being a part of where there are shipments included, he used this API, which, once an order is placed, it automatically sends a request to Easy Post API, about the content of the package, the weight, where it’s going be, and the postage you want to buy, so you can compare between UPS, FedEx, DHL. And then this API takes care of it and sends the package by the price required to have it shipped, in three days, ten days, and then you’re done.

So you don’t have to deal as an owner of, let’s say, this marketplace, with all of the shipping process. You just use this API, and you integrate it into your project, and once an order needs to be shipped, that API does all the rest. And that’s amazing, right? Because you can [inaudible 13:17] other things.

Q: Unbelievable. And what does [inaudible 13:18] do, which kind of API can [inaudible 13:20] use?

A: Data, data. And they’re using it more as a partner API, so they don’t have a public API, but they have public. Because private APIs help drop down the cost of your development. So they have iPhone apps, iPad apps, and behind all of these apps that they use, they have an API behind that.

And where we should not use an API is probably in the banks and stuff, because the security is not… I mean, banks are obsessive with protecting some data that would never be in the Cloud, for a lot of reasons. So there will always be some parts of the world that will not… but that’s not really about an API problem, it’s more about a Cloud computing problem, because all the things that would not be accessible for security reasons, and I think banks or government-some data from the government, for example, will also not be able to use APIs, because of the same reason.

Actually, the government, now, in the U.S., is putting out a lot of APIs, and they are publishing more than 100 APIs in the next year, to basically distribute all data. So the ones you see, like the [crime network people], or CD, you access all this data through the APIs. The Minister of Economy is doing the same thing, opening up all this data about economies in the United States, and you can go back in the history and access all this data. Agriculture problems, you can see everything, [inaudible 14:59]. The government is opening all of these APIs, in the next years.

Q: Now, last week, I was with a friend of mine, [inaudible 15:05] on this topic, we were talking about APIs. And this friend of mine, he works in an ice cream shop, okay? So he doesn’t know anything about tech or API. And I said, “Yeah, because for instance, you can access, you know, all this data with this API.” And he said, “Yeah, but what’s the difference, if they use this thing that you are talking to me about, or if they just upload a PDF, or an Excel file, what’s the difference?” And for me, the difference is obviously monumental, but I want you to give back the technical answer.

A: Yeah. It’s the structure of the data, right? Well, wait, wait. If you are a consumer, you could always use the PDF, or Excel, or whatever. What we’re talking about here is targeting developers in [conditions] that need to have software somewhere. So for a developer, or an engineer, it can’t be on Excel, it needs to download the Excel, structure the data, put it out, analyze things.

So it’s really actually taking the PDF structure and putting a developer structure into it. And it takes a lot. In the API, it’s just an API code, right? The data is already structured the way you want it, so developers love JSON, which is a format to transfer data, and they love JSON as a format. And the APIs, RESTful, is a simple way to do just that. It’s just selling and distributing data and services through JSON, so it’s developer language, and they can just make a call and retrieve all the data the way they want to.

Also, with API, the data are still owned by the provider, in the server of the provider. I just make a call when I want the data. So access is more important than ownership nowadays, right? And I think also, [inaudible 17:05] is a real example, that access is more important than ownership, and the same thing for the APIs. With the Excel and the PDF, once I download the PDFs, I also own the data, and I need to manage and maintain this data [inaudible 17:16].

And that’s what’s good about APIs, it’s not only that I can use an SMS API and be up and running quickly, but also it’s because I don’t have to maintain that part of my product, because there’s another company doing that, focusing on that, like [inaudible 17:29] , you know, industry, a company doing this thing. They really care about those things, and they’re really, really, really good with one thing, and I do my thing, my ten percent code really well, without creating a [inaudible 17:30] third party components and services, because they do that for me. So it’s the managing that’s important.

Q: Let’s talk about the cost, and in this way we can also go deep in Mashape. I mean, first of all, Mashape is a Cloud API marketplace, where you can put your API out there, or you can find an API. So it’s very useful, in my opinion, because it gives you a solution, a very simple place where you find all APIs, and you can move in this world much easier. So guys, just go Mashape, and have a look.

How about the cost? I mean, are companies now charging for using their public APIs? On Mashape, do you see, I don’t know, only free APIs, or APIs that are sold, the access is sold? Can you elaborate on this topic of the cost? Because I think there is an idea that all of those protocols are free and you just access, and that’s it, you do what you want, a little bit in a chaotic way, when it’s not really like that.

A: So when you decide to open an API, and give it out to a partner… obviously, if it’s private, it’s always free [because you use it.] But when you decide to open it, you can keep an API free, and there are some reasons to do that, or you can add a paid usage model on top of your API. It depends on the API itself.

In the last, I think, five, six years, we have seen an opening, [inaudible 19:18] opening APIs, and now we are entering a wave where it’s monetized APIs. And Mashape helps really, really well with that, because it allows you to monetize your API in three clicks, and accept money now. I also have a community and a marketplace for that. So this is what we believe in, that if something has value, then they will never keep it for free forever. So at one point, we started to charge for it.

So now we are entering a wave of people charging for APIs. And there are really good examples there. So there are actually three different categories of APIs. There are Cloud APIs that have infrastructure as a service, like [Twilio]. So I provide you an infrastructure, like in this case SMS. There are data APIs, that just distribute data, like the government APIs, right? The crime in San Francisco and in New York, the comparison, and things like that.

And then there are platform APIs, which is basically like Facebook, Twitter, Foursquare, where they use the API as a strategy to leverage third- party developers to build applications and services on top of Facebook, on top of Twitter, to basically increase the distribution of Facebook itself, and also reach the Facebook community by third-party applications.

Right. So there are three different kinds of APIs. What we are focusing on is more of the first two, so the selling data and services, and not on the third ones, like Twitter, Facebook, Foursquare, because in these APIs, there is revenue, but it’s not direct revenue.

So when Zynga deploys and builds games on top of Facebook, there is not real, clear, transaction revenue, but there is a collateral aspect, when people buy credit, or get credit to play Farmville. And all that is happening through the Facebook API, but it’s not really for us simple to market that.

So instead of Twilio, it’s pretty simple, because Twilio says, “Okay, you pay one cent per SMS that you send through us, or using our infrastructure.” And that’s pretty simple, or data, that most [inaudible 21:18] are free, but some of them are pay, like the unit could be data points. Every time you retrieve 100 data points, it’s ten bucks. So this is more clearly simple to monetize.

In our marketplace, the majority are private APIs, like in the world. There are a set of public APIs, and in that set of public APIs, there are APIs with a business model, with a revenue generator. And they charge, for example, there is a face recognition API. So if you want to do an iPhone app, but you need some face recognition, [sentimental] analysis, you use search us. Because we’re not just a directory, so you search an API, and then you can actually consume that API through us. And you pay through us, so it’s an entire one-shop stop to do everything you want to do. Search, consume, pay, analyze. Everything in one place, from a consumer standpoint.

And so there is this face recognition, and the way they charge is they do not charge per API code, it’s the unit [inaudible 22:13] , so it’s API codes, [inaudible 22:14] codes, which means, if I’m using an API, how many times I ping that server, that’s an API code. Bump, bump. And in the API code, there can be a lot of objects, so in the face recognition, the object they charge for is face recognized, not API codes.

So every time I recognize a face, that’s another. And they charge, like, ten bucks, I guess, for 50,000 faces recognized per month. You can add more if you want, you can upgrade to different plans. In an API world, you pay as you go, like SMS, one cent per SMS, or in a [inaudible 22:51], where you have subscription mode, where you pay an amount per month, and you have included an amount of face recognizing API codes per month.

Q: Augusto, what’s the best advice, what’s your top recommendation to a company that wants to approach this API world, but still has to start, and they have to start from scratch. What would you recommend, what’s the right guideline to follow, in your opinion?

A: The problem with doing RESTful API is that, it’s not like SOAP, where there’s a clear path to follow. But everyone has their own interpretation somehow of RESTful APIs. So my suggestion is just to go and read as much documentation as possible, like Googling and things, on how you build RESTful APIs. And then you get pretty good knowledge in a couple of days. And then you start to design APIs, and at that point, probably.

Also, at Mashape, we provide an API doc editor, which kind of forces you to design APIs the right way. So using that, you not only get the advantage of having five client libraries, distributed APIs, but you also get a function that you force yourself to build a RESTful API, because it drives you through the, not the best API in the world, but as REST as possible.

Q: Right, super. Augusto, how did you get this idea? Because talking now, 2012, about these APIs, great, you know. We’re here to explain what it’s all about, but for the tech community, the tech world, okay. This is, you know, API is a matter of fact, it is. It exists, and this is the direction. But when you started, I remember, no one really understood what it was all about, so how did you come up with this idea?

A: Yeah, I know, it was a vision, right? We did have a lot of problems in the early days explaining what we were doing, because when we were saying we are the app store of APIs, we’re a marketplace for APIs, in early, early 2010, people were saying, “What do you mean? An API can’t be sold, right?”

Because at the time, they were thinking about those APIs of programming language, APIs of Windows 95, not the Cloud APIs that revolutionized… There was already Twilio, but few engineers knew about this kind of API, and the concept of it.

Fortunately, mobile drove all of the API revolution; the entire growth is because of mobile. Because as I told you, every new app on the iPhone, it means that behind that, there is an API. So that actually was [inaudible 25:34]

We came out, actually, it was a Pbot integration from another product. And the other product, I don’t know if you know, it was like a [inaudible 24:44] where you have to assemble and combine different components together. We also had a graphical layout, so we would allow you to drag and drop component functions, and use our interface to create your own app.

But to do that, we were [erupting] all other existing APIs at the time, and then what we discovered, the explosion of APIs, and we decided, ‘”Oh shit, there will be thousands, millions, and we need to unify all this mess, and create this marketplace and this app.”

So it was an idea that came out from another idea, and it took us a lot of months, [and iterations] to find the right one. But this vision of having a marketplace for APIs, which is even ahead of it, so it’s not only opening APIs, editing them, it’s also monetizing it. This is also very revolutionary, what we are trying to do. And actually, we are creating a market, as the one and only that lets you monetize APIs by click.

And it was very hard, two years ago, to explain that because people would say, “What do you mean, you can’t sell an API.” And now, it’s way, way easier, to get way, way involved, because this explosion of infrastructure as a service, like [inaudible 26:46], send email, doesn’t want to do email, [inaudible 26:47] to send an email.

Q: How was getting funded for this idea? Because it looks easy, and you just say, Okay-many guests at the Tech Alchemist, they may be from Silicon Valley, Seattle; they’re very well inside the network environment. You started from Italy, which is not really Silicon Valley; it’s not exactly the same. And then you, I mean, you had a great story, you went there, and you did it by yourself.

So how was the path to arrive there? I mean, did you just start there and say, “Hey, hi guys, we have this idea,” and someone came? How did you find your way in those couple of years?

A: I think when I came to Silicon Valley the first time, in 2008 at a Stanford [inaudible 27:47], it was February. And actually, that was the first time I felt [inaudible 27:50], saw how powerful things are here, and said, ‘Wow.’ And then I came back to Italy, and then started to work two years trying to raise money for the original idea of [inaudible 28:025].

And, you know, at that kind of stage, you actually don’t invest in the idea, you invest in the people, right? The sooner you invest, the more you invest in people. Because if you like, let’s say, if you invest in your [inaudible 28:19] three years ago, you were investing in people. You invest in your [inaudible 28:22] now, even if you take out all the founders, and you put other CEOs and other funders, the product is already running, and goes wherever it goes. It may be a little bit lower, but it’s gone. It’s going. Now you’re investing in your [inaudible 28:31]. So that’s the thing.

And we had a serious problem in Italy with the funding, because the team, in the region, was only Marco and me, and then Mike also came in later. But the problem was the team, we were too young, we were 20 years old, and they didn’t want to fund a 20-year-old team with no experience at all. I was trying to explain that having no experience is good because nobody tells you how to do things, so you can invent your own way, and most of the time, it’s a better way.

And that actually was the problem, so in 2010, I saw this [inaudible 29:05], which I can summarize, [inaudible 29:06], everything happened so fast, from funding to partnership to business to building, I think speed is the right word to define Silicon Valley. And then in Italy, to the contrary, everything happens so slowly. But it’s not an Italian problem, it’s a European problem.

And then we moved here, and then this regional theme of YouTube, they were just the first guys that gave us a couple hundred grants, with also one investor from Italy, [Sigrelli], because he actually, the way we found him, we didn’t meet him for over two years in Italy, we met him in San Francisco in 2010 when we were raising funding from these guys, but we were sleeping on the floor, basically, in a house, with an air mattress.

So the YouTube guys saw the commitment of three desperate people who basically came here to find their dream. And, so they made the commitment, and they invested in the team. And then also, when the Italian guys came in, he invested in the team, because he saw Italian guys. So it an investment in the team, and actually a little bit of an idea, but it’s always about the team at that stage. In the [inaudible 30:25], a year later, it’s more of a mix, where we have [perfection of the idea] between group, and that’s when you invest in both team and product.

And over time, you still invest a little bit more in the team, and in the product as well, and in the [inaudible 30:44], you probably invest mostly in the product.

Q: Well, we have a few minutes more, Augusto. But after you told me, you were the first person telling me about Airbnb, and Airbnb… just, guys, if you don’t know Airbnb, just go out there and check out Airbnb, and it’s now a massive business. But you were the first one to tell me about Airbnb, and in the beginning, I thought, “Come on, that probably will go nowhere.”

A: That’s what the investor told them, in 2009, when they were trying to do fundraising. Sorry, late 2008, when they were trying to do fundraising. He said, “No one is going stay in a house with people they don’t know.”

Q: Yeah, and now it’s going great. So after that moment, every time that you say, “Oh, maybe have a look at this startup or website,” I immediately go to check. So do you see… do you have your favorite three startups of the moment? Is there anything interesting that you see for people doing tech, digital business, some cool website, cool app, cool startups to have a look at?

A: I don’t know about three. I mean, I am personally invested in a startup here in Silicon Valley for [inaudible 32:03], which I think may become really big and disrupt eBay. It’s basically like Craigslist, but mobile, and that’s simple. You take a picture of the object you want to sell, and you sell it, localize it via the iPhone app, and then other people can browse items through the location, and go pick up and buy the items through the iPhone app.

So it’s called Yardsale. GetYardsale.com. So it’s like, I take a picture and send it in 20 seconds and you buy it in 20 seconds with the iPhone. It’s an amazing user experience, like Apple-style. So it’ll completely disrupt Craigslist’s business, and eventually they go up. This is something that I would keep my eyes on, because it’s…

Q: Get Yardsale.

A: But, other consumer things that I… you know, now, in Silicon Valley, a lot of investors are coming back to their roots, and they’re-it’s hard, now, to invest in a finance consumer product, because of the iPhone market, and because there is not a clear path. Instead, money is going back to new finance enterprise companies, and [inaudible 33:12] companies, in a faster and powerful way. I would bet on Vox. That’s an amazing company, high R&D, it’s an amazing leader. I would bet on Vox. Vox is going to be a huge, huge, huge, huge company to work at. They just started.

Q: Okay. And tools that you absolutely would recommend for people doing business online? Tools, I don’t know, for promoting your business online? By the way, something that you had to do, but your target was probably more developers, or promoting through developers. Are there any tools that you say, “Gosh, if you do some digital business, you have to use these tools for your productivity, for your promotion.” Are there any particular favorite Augusto’s favorite tools that you could recommend?

A: Well, Mixpanel should be used, because it’s like a Google Analytics bar for real time and tracking. Optimally, to do AB testing, [that's a simple way.] So we’re really obsessive with the data and methods, and so we use a lot of those things. I don’t know about marketing or outbound tools. I prefer to do those things manually, instead of using software. But yeah, Mixpanel is something that I think everyone should use to track their menus. This also works for a mobile app, so they also have Mixpanel mobile, which tracks mobile usage on top of your app and tracks every evidence you want to track.

Another thing we’re using now is InstaCart. It’s only in San Francisco, so it’s going be developed very slowly because of location problems, but InstaCart is basically, I open my iPhone, I browse a list of Grocery Kings, it’s actually with Safeway, and I say I want a latte, I want milk, I want a sandwich, I want the meat, and how many pieces. Two, three, four. I want Cornflakes, whatever it is. And I can browse the groceries in a way like Pinterest, with all of the things, and then I can order, and the guy comes to the door and brings me, for just nine bucks, so I pay whatever I want to pay. The flat rate is nine bucks per order, no matter how much I ordered, and he brings me all of the things at my door, in less than two hours.

Q: Cool, cool. I was thinking, when will it come to Italy? You know, probably never, because they will steal the milk before he is coming in.

A: Sure.

Q: All right. Last question, Augusto, just try to open your crystal magic ball, and tell us what will happen in the next couple of years, in your sector. I mean, API will be everywhere, it will be a totally connected world. Is there any particular direction that you see that you say, “That will happen, for sure?”

A: Yeah, it’s the software going everywhere, and then APIs as well. But it’s probably driven by the APIs, because, as I told you, they cut down costs in distributing software in different platforms, different objects. It’s really simple, because you need just one API, to do [inaudible 36:18] in every other apps, even in upgrades, you need just the same API, that you used [inaudible 36:20]. In iPhone, you can use this again [inaudible 36:21] upgrade. So I think that the invasion of software in every corner of your life is what’s happening in the next five years. So mobile will not be only the mobile, the phone and the iPad, but the mobile will be also your car, which is also mobile. So everything’s going be in every aspect of your life, and software and apps are going to be everywhere.

Q: Super.

A: I think we’re just getting started with this.

Q: Okay. Augusto, thanks a lot, Augusto Marietti, Co-Founder of Mashape, and good luck, keeping in touch for everything.

A: Thanks, Marco, thanks. Ciao.