EIC 2012 Session: The Kuppingercole IT Model and the API Economy

So as we all think that you have seen many, many slides during the last few days before, we'll only have a few informational ones to, as an inspiration after that, the three guys are going to kick in with lots of discussion about the details, I guess. So fire away, Martin. Okay. So maybe I start, haven't been talking for at least on 10 or 50 minutes.

So, so we added a lot of talk about covered in my keynote, the it model and the API economy things. So it's a great job, probably a little bit about uniting choosing, which on one hand were initially very much driven by Craig on the other hand. So when it comes to the API economy, on the other hand, very much driven by me when it comes to the it model thing. And I think there's, there are a lot of things which are overlap. I just wanna start with a very, very, very short presentation before we start into discussion with Kim in addition. So y'all have seen this picture of the it paradigm.

The one was the typo. And so did, do you find the typo and it's really about how can I provide business services by integrating a lot of different services by really making the best out of on and cloud use and this middle layer around services? I think it's the important point. Where's no about a lot of things around services, but one very important point is the service orchestration thing. And the service orchestration thing really is the right tailoring and combining, and so on services making more out of this.

So that's really this part where I will say the API I economy sorts really come in and play a very vital role. So this idea of saying, okay, I don't have a big fat SA app only, but I'm able to consume the app and use the parts I want or integrated with whatever I have on premise of premise.

And so, and I think that's really the point I've touched this slide quite a few times. And so that's really the starting point for this.

And yeah, I think that's one of the terms maybe Greg wants to say some, some words around using microphone, number three, by the way, before we go for forward. Yeah. This is the core thesis of the API economy document that we published a few months ago and that organizations I, we believe are moving into a position. If they're not baking their core competence into an open API that they're not in the game or it's an economic imperative that they be doing that now there's a lot of implications in a statement that bold, it means that that organizations are becoming SaaS providers.

And that means well, that means that they have to figure out how to make and build and maintain software. What, what, and what happens when, you know, 20 million hits or, you know, if it went into the billions of hits like some organizations have had, so we've got some, a lot of issues about security and how to protect the data legal matters, who owns the data? Is it in the cloud or is it still on the premise? And I love that we have Kim here a lot here today, cuz he's been thinking about those issues in a lot of depth for Microsoft and we've been talking about it for many years.

So I'm excited to hear what he's gonna say. And I think that there are two points which are very important around this.

You know, you might always say, why should an organization becomes, or enterprise should become so often sort of on whatever cloud or SaaS provider. I think there are two points. If you look at the it paradigm it's about on premise, it acting like they are one. The other thing is I've used the term identity explosion in my keynote, I think. And I think the CI thing, you know, we are not dealing only with this, let's say 28. So I had a number recently for customer and they have 28,000 employee and four, 5 million customers to deal with.

So it's, the employees are less far, less than 1%. And I think that's really the point you have to look at. So unwilling things, making them more open is important because you have to deal with your customer. So it's an economic imperative and that's trust leads to the another slide that trust wanna quickly look, look at. It's really the point I've been asked and you probably as well, why is relevant for an enterprise? I think it's for relevant for everyone it's relevant for SA providers because the big fat SA apps are insufficient. It's relevant for other types of cloud providers.

Even if it's more maybe on about the administrative side, it's relevant for the on-premise it providers because they have to act like a cloud for the end user organizations because they're the consumer number one for the individuals, because they might rely on those maybe based on the future life management platform. So there are a lot of things in there and full of later on will go much deeper into the consumer side of things in his presentation. But I think that's, that's a lot of things.

And the thing I really wanted to show is only this, the last slide, because I had question yesterday on a panel, I think we have been Bo we had a progression on someone. Why should I believe in this?

You know, Cobra failed, soap is not growing or so well overall is not drawing that massively. So why should it happen? And I think that's the answer. And right now I think maybe it's really latest time to inter include Kim into our discussion. Let him talk a little bit because I think he's char is the Fu he's really waiting, oh, when can I start talking?

So, so, so Kim, you know, as being historically working or working for a company, which has been historically been a provider on premise solutions, what is your view on that?

Oh, well you mean around the API economy and yeah, it's funny, you know, I'm, I'm just an engineer and you know, you know, you, you might, I sort of came into this through the restful programming model, which, which I thought was brilliant because I always like anything that is a step simpler is, is good, but, and so I really, you know, understood the technology of it, but I, I didn't get the message, you know, I, I, I didn't actually, and I, I knew that Microsoft had to do it because it was simpler. Anything is simpler, you have to do, but I didn't get the message.

The message is that actually there's this message of there being a new economy and, you know, basically, so it was really Craig who, who, who, who enlightened me on that. And I mean, I just think it's so incredibly true. All of our customers will be offering their services through this, this mechanism in a new economy. So I have to adapt to it in several ways.

One is, you know, everything we do has to, you know, our software has been very well layered in the sense that, you know, over the years we learned, gee, your UI has to be on top of a different layer than your backend services and so on. And so we actually have, you know, we did things like power shell, which surfaced all of the functionality in a very simple mechanism. And so really all of that stuff has to be then now given restful heads so that it can all be be, and it's, it's not really a hard it's, it's, it's just, you know, it's not hard from a theoretical point of view, it's all doable.

So that's, that's a question of, of making everything we do like that. For example, the directory, the cloud directory has to be accessible in, in precisely this way.

The, so all of the configuration, everything, everything about identity management has to be available through these, this type of API. So it's in fact, moving away from we are provider SA is and different levels. So it's really about saying, okay, the message you're telling is yes, we are moving forward from providing one, let's say more or less moralistic thing towards having all the options the customer might need.

Yeah, exactly. I mean, you know, I think that the main point, my big aha this year was the, the real essence of the cloud economy is specialization.

It's not, we can run our machines in a bigger refrigerator, which was sort of the original message that we were given, you know, by the, by the cloud thinkers, it's really the, the whole economy is gonna be, is gonna be distributed and specialized. And, and in, in a way that, that the API economy facilitates or, or it's like, which came first, the cloud or the API economy, that, that that's what future parents will, when they want to amuse their children, they'll say, which came first, the, the cloud or the API economy, because they're so tightly and intricately bound. It is.

I, I, I just, I, I can't emphasize the significance of what Kim has said that they're gonna take the cloud ums and make it accessible through restful APIs to any application from anyone that wants to consume and use that directory as an element in their programming. That's, it's, it's, it, it, that's such a shift from the thinking of any vendor in the past or any notion, and the ability for that to fuel the economy, like he said, is, is, is huge. The implication of that is, is really exciting. Yeah. And I think that what we usually trust that came around it, it allows really to specialize.

And I think that's really where if we put, take together these things to cloud and really cloud as a specialization thing, if you take the API economy and the big world of interest realization of it, and I think that's really where these things come together, because if you look at the, in industrialization real world, some very, very long time ago, it was also a movement forwards towards specialization and towards the ability to inter-operate.

So specializing things without specializing, without having the ability to then integrate back things which came from different specialized providers didn't work. And so it's really about moving forward to broach, which supports doing this. And I think that's really where these things come in. And so from that perspective, it really has a very, very strong logic because when we are not able to integrate, we don't need to specialize at the end of the data and it's trust at best half the week.

And so I think that's a very important point from that perspective that we need to move forward in that direction. And, and I think that's the three, I think the point I'm really honestly, still thinking a lot about is what will be the provider role of the consumer in the future. But I think we have some answers in our life management platform report.

And I think really disability for him to provide some things in a way which is exposed probably by something which allows them to access this information in a very privacy and security related way will be a part of the answer, which is just starting to, to answer to a question, just starting to become clear, I would say A personal cloud as it were. Yes. But I don't like the term personal cloud because I think the it's over stressing the cloud paradigm and maybe it's too cloudy.

Well, I'll tell you around the specialization a few weeks ago, or I was at this conference and I haven't really spoken about any of the stuff I spoke about here in my keynote to anyone. It's really just one other group of people, which, which was this conference of, you know, people in the aeronautic, aeronautical and defense industry, if you can believe it very, very high level people. And I was giving them this, this, this explanation of the cloud is about specialization, as you were saying.

And all of a sudden I realized that everybody in the room except me understood specialization because, you know, they're in the aeronautics industry, you have one guy who produces a, you know, some piece of an engine and another guy who produces another piece and everybody's specialized. And, and like, it's just the software industry where we kind of had this monolithic.

We, we didn't make it into the, in, after the industrial revolution yet. Yeah. We didn't Make, it's just happening now. That's what this is. Yeah. And hopefully we do it a little faster than the real world did. So otherwise it'll be 2,200 something until we come that way. And there's a couple other things I think besides specialization that this enables one is scalability and the other is change management.

I, one of the most frustrating things in it is the, the pace that it's able to embrace the need for change. Yeah. It is incredibly in that trial and Business. This Yeah. The businesses asking for agility.

In fact, I think when we are looking at an, so I think I had this discussion during these days sometimes say, oh, that's so much integration and things to do. I think the point is it's getting easier. And if you build the right organization, if you build the right model, that's what we've described. And if you enable these things, then you will become better able to do this. And I think the point is, it's a little bit about liver. It's called liver or something like this the other way around.

So I think the point is either you do those it, or you will be replaced at the end of the day, if you're not agile enough, then business will bypass you. And that's what, what the cloud enables. So to become the better. And it's better to, to, to do it structure in an organization, what we described in our paradigm. And that means you have to go to pass down somewhere. I think there will be a lot of learnings around us over time. So we are really in the early stages, but I think it's becoming more and more consistent.

And that's really what businesses want agility because they are changing requirements and that's not going FA slower than it has been the last year. It will become faster and faster, I think, in a globalized economy. And that's really what it's about at the end of the day. I sort of see it as, you know, the, it expands to consume the budget allocated to it and, you know, and Then moves the Well, no, but it expands to consume the budget allocated to it.

And, and that will continue. It's just that the economy itself, you know, because things will become the it component will become more and more central to the nature of what an enterprise is through the API economy. And so what's interesting right now, a lot of it is spent doing things that are not core to the essence of the, of the corporation.

It, a lot of the, it is just spent, you know, sweeping the floor and, you know, mowing the lawn as opposed to doing the actual thingness of that enterprise. And so I think that's, what's gonna happen through the cloud, all the mowing, all the lawn mowing and so on will be done in the cloud by all of these other specialized services. And the enterprise can take its it resources and use that to be itself. Yeah.

I think, I think we are all in the same line here. I think we have to, We don't have an argument. This is a problem for a panel, right?

We, Yeah, no, no. I think we anyway have to move to the next presentation, which is full loop. So I ask Sebastian and to come back to the stage because he lost stage in so much discussion or probably around live management platforms and other things, which is good by the way. I also hope that some people were heavily Twittering right now because I think we had some very nice bold statements there. And if they don't tweet them, we will forget them again. It would be very helpful to have them tweeted Everything that is, is remembered Hand over to the best yeah.

To move forward in our agenda. Otherwise we absolutely is a run out of argument or we will spend the entire hour and a little bit more without Phillip having his present. You so next on stage is my valued colleague Phillip fall. And he'll give us a, a short glimpse into the consumer view of the API economy. Welcome on stage, Phillip. Thank you. So I don't have voice today and it's not because of the ping party. I may promise. So I'm not a pilot, you know, I'm a sailor, so I tend to stay on ground. And Kim says a technologist. I'm also a technologist.

So I feel more confident with things that we can really implement that with big abstraction. And so that what I want now. Yeah. Sorry.

And so my, my presentation is really about, you know, what is the usage of, of those API? You know, what's the consumer view about it? So the funny thing is I was expecting that this would define the open API, you know, it's, it's quite funny, everyone is talking about it. And if you Google on, on, on the internet, it, you find very few definition, which is pretty aware for new technology.

You know, mean people fight about what should be the right definition. I think everyone agree on what is API.

You know, we, we know that open is, is more complex. Okay. Doesn't mean it's real. Does it mean it's open source? Doesn't mean anyone can use it.

You know, what is Beyonce is open is a, is still a good question. I think nevertheless, that what people agree onto is on the objective as your open API.

So it's, it's exposing a subset of your data outside of your firewall in order this to be used by external party. Okay. And I think this is very important. Kim is talking about breaking the domain border.

You know, I, I'm talking about breaking the middle age castle architecture, which is the same concept at the end of today. It is working with people that are not within on security domain. But another element that I think is agree with everyone is that it should be simple and secure. Okay. And simplicity is a key element down there. And we talked about many previous initiative who failed in this arena. And one of the main reason was the complexity. So coming back to the consumer view, and there are definitely two vision about the, who is consuming zoo API.

You know, we can't use a consumer as an enterprise. Okay. And in this case it's probably more, better and easier integration to the cloud application. Okay. Is that really what was the main motivation for them? Because they had no answer with the traditional approach of, you know, accessing directly to the data storage, but it's also very good to open to partner program, whether it is to expose or to consume, you know, information from others and, and last but not least enabling new channel.

And we've seen that mobile phone was definitely one of the, the key enabler for people to move to the world of the open API now from a consumer or from, you know, citizen basic and user point of view, obviously zoo people should not see the open API technology. It should be transparent for them. Nevertheless, they still use that.

And, and what they use first is probably the delegation of authentication, you know, with the Facebook connect button, but also it's allows them to embed social service in many application, all those switch sets that allow you to have smarter application, whether it's for geolocation, for search, click, to buy or any kind of things that can be provided by others. Now what's a real mean, you know, of the API.

It, I think it's really the opportunity to connect, you know, to any type of application. You know, since the past 10 or 20 years, we've been stuck with the browser.

And, and finally someone came with announcer that the, the browser was not a limiting factor anymore, but we could use any type of application to connect to our datas. It also provide unique new opportunity for developers. Okay. Because you get access not only to raw data, but you get access to pre prebuilt information. So it allows you to build much faster and much smarter application.

It's, it's a definite gain in efficiency. Again, it's, it's allowing you to access to new channel new capability and it provide a, a unique user experience, whether it is for the mobile, the widget or, or mashup of applications. So about consuming API, when do we want to consume API?

I think, you know, every time we build business that rely on some data owned by others, whether it is geolocation, whether it is a list of product, a price, you know, airline ticket, whatever. There are many, many cases where we need information that are owned by others. And this is really where we need open API.

It allow us to build smarter application with mashup composition orchestration of services, and obviously the capability to integrate with the cloud because open API support natively is a cloud because they don't do any difference in between application in the cloud and internal application. And I think that's also very important to understand that open API also make sense and is valuable inside the enterprise. So when do we expose open API?

You know, if we consume them, it means that at some point we should expose them as well. And I think every time we have to exchange standard data with external party, it's much easier to expose that through a set of API than it is to open some part of your database to some external party. It increased business opportunity by expanding the exposing of your brand.

You know, you may have, let's say some products that are going to be obsolete in two days. You cannot sell them on your website because it would not be good for you, but it may make sense to expose those products who open API. So some special company can collect that and make a business out of it. You may also expose real time data you have available on your, on your side to the outside world.

Not everyone is going to connect to your website to know some fun, you know, tiny things about your company, where is, you know, the train timetable, whether it is, you know, how many product you have been stuck, whether is the current price of what one, one company and you're willing to sell. And in improved business opportunity with new class of applications that can leverage Eden data. Okay. I think in every enterprise we have a lot of data that are just hidden. Not because they would not be useful for others, but because it's too complex for us to expose them to the outside world.

So who do we provide API to, or who is consuming API? Because at the end of the day, this is the same question. I would say the first one is the first party.

You know, it's internal people that use open API just to boost internal development. And I think that's a very important concept okay. In the enterprise building new applications that really answer to the business demand way, it takes too long. It also allows you to open to partners, okay.

People, you know, so you may have some special contract and it's easier to provide them only a subset of data. They're entitle to see through the open API.

And, and obviously you can share services. And, and at the end of the day, you may also provide API to people you don't know, which is what, you know, all the social network are doing with your location, your list of friends, how to connect, you know, even Facebook connect is a good example of APIs that is used to be designed to be used by people that don't know.

Now, many people will say, if it's open, we have risk. Okay. So what is a risk? I think effectively is a risk that is clearly on the security side. You have to make sure that you deliver the right information to the right person. Okay.

And, and, and the, the bad thing about this is the security model is very different from what the traditional enterprise are used to, to, to implement it's based on, on token, on authorization, on per transaction is definitely not a per channel security model. Reliability might be another issue. Okay. If you use someone else API, or if you provide API to someone else, you have to make sure that this is going to fulfill your SLA. Okay. If you rely all your business on some API, and certainly this API does not scale, you're in big trouble. And we know that, you know, that thing happen.

You know, even people like Facebook or Amazon at breakdown in, in their API services and someone like Yahoo shopping disappear. So it's something you have to take in account. And at the end of the day, this is really depending on your business vision of your data distribution, you know, are you implementing or using open API just because it's trendy and, and you build something for only a few weeks, few months, you want to create a burst, but it's not a long term strategy, or, or whether you willing to really build something that is designed for your business.

And depending on the case, you probably not going to use the same class of API. Another risk is a dilution of knowledge. If you stop doing something because you take it from somewhere else. After a few years, you don't have the knowledge anymore on how to do that. Which means that the person who is providing the service is in a very good position to impose its price. Okay.

So that's, that's a definitely one risk and customer dis intermediation is another one. Many companies want to keep a direct contact with the customer. And obviously providing open API is a way to lose this direct contact with the customer. So what are the evolution and trend?

Okay, definitely more intelligent. Inter middle layer, you know, on open API. On one side, we have a consumer on the other one, we as a provider, I doubt that they're going to talk very often directly. Okay. So our cases where they will do, but not.

So we, we see a value there for mitigating risk, performance, security, complexity, to have a middle layer. We have a generalization of model API. It was already said everyone is going to have it. It's a must a feature for cloud. And it's only the only valid way to address the next generation of connected device. And when I I'm thinking about next generation, I'm not thinking about a phone because the phone, you know, you can work with a browser on the phone. It's not very convenient, but it works, but in a car, it is impossible. Okay.

A car has up to 60 input device to, to the entertainment system and, and, you know, no one is going to expect using a browser to provide information in your car. So you need open API for that. And it's still a maturing technology. Okay.

It's, it's not completely subtle. Okay. I don't know if it's going to take 100 years as you say, or little shorter than that. But one thing issue is that many actor are going to disappear within the next five years. I think that's obvious, you know, things are going to change a lot.

So what is the recommendation first is for people exposing API authentication, authentication, authentication, you have to make sure that you really know the user, the applications, the channel, you know, it's a per transaction security model, leverage external Intermed imagery, you know, in, in the middle layer, not only to handle, you know, the securities, the generation of the key and so on, but also to, to do the monitoring, the caching enables the scaling. Okay.

If you open your, your database with open API and someone suddenly start to scan all your database is not obvious that, you know, your it system is designed to sustain such an extra load. And for people consuming API, you know, use orchestration platform to consume services. Okay. Favor simple, basic service to big flat, you know, big fat traditional cloud applications. And this is going to be my last slide. Wonderful. Thanks. Yeah. Thanks. Phillip Martin asked me to, to give him room to make one comment. Yeah. I Think that there's one thing.

And then currently there's a lot of talk about big data. I personally like smart data more than as a term, because it's not about getting big, but about getting smart regarding data.

But, but one thing I just thought about when listening to Phillips presentation is, do we really need big data in the open API economy? Or is it not about really being smart enough to consume the right data when you need it? Which by the way, would be one of the most interesting business aspects driving open API economy, short term, just as I thought, I'd like to draw in here. I think it's an interesting point to think about. Yeah.

And in, in the meantime, I'd already liked Mr. Steven Wilmer to prepare to come up on stage Phillip. One thing that came to my mind when I looked at your slides, the very first ones I thought, Hmm. Like many years ago, when the developers were going crazy about object, orientation and code reuse, you could have shown those slides back then because it it's all about reuse. It's all about simple. It's all about optimistic.

Ain't, ain't that sort of the, the same model. But I, I don't reuse the code. I just use the information that's provided by that interface. Yes and no.

You know, I think that at some point where ideas pop up and they are good ideas object-oriented is a good example. And we had many discussion about small talk and C plus plus and FL and other languages.

And, and at some point, you know, someone came with a Java implementation that was just complex enough for people to use it. And then people started to use it. And I think it's, it's true that, you know, we had this open API vision with, you know, the previous, you know, component core and others. That was the same idea. It was just too complex. The question here is, is it simple enough to be used now?

There's another element that I think is going, you know, this innovation is not coming from the enterprise it's coming from from the, the, the consumer sector is coming from the Facebook, from the Google from, so it's flowing, you know, not from, you know, the foundation of the enterprise that's stinging. Yes, we need that, like was designed COBA by the bank and a big player.

You know, it's something that is available and already used by millions and millions of people. And then the entre are looking at that. Yeah. Maybe we could use that as well.

You know, I think it took for military people more than almost 100 years to realize that they should stop producing their own cars. Okay. Until very recently they had specification for their own cars, their own computers, their own, everything. Now they just buy element from outside.

And, and I think we probably getting old nav. So we also start to realize that, you know, it's might be time to stop reinventing the wheel and let's try use something we know is working Mature enough. So next on stage, we have Dr. Steven Wilmer of three scale, and he'll just give us the opposite view, the provider view of the API economy. Welcome in on stage. Thank you very much. Perfect. It's this right. So thanks very much for the invitation. I really appreciate it hanging off on stage with these guys and to, I don't know if it's gonna be a completely opposite view.

So, but I think it'll be very complimentary. So I'm the CEO through scale. We're a company that provides infrastructure for APIs. We're one of the top infrastructure providers in the world. We power about 90 public APIs right now there's about 300 other organizations that use us internally or privately for various things. You can see it's extremely diverse. So you have communications infrastructure, you have public data, government data, you have music, data, all sorts of different things that are here. And some of those are right APIs. Some of them are read APIs.

So we see a lot of different business models. We see a lot of different things, so people are doing with APIs. And so we'll try to share a bit of that with you. So I'm gonna cover three things, just the economy, then a dive down into how it sort of works and what people are getting out of it and, and why they're doing these things and try to cover as much and many examples I can. So we've seen some, some discussion already about what the, what the API economy means. So I just wanted to give you an example. So if you look@amazon.com, they started off as a website, right?

So it was a website, but now they just have a huge number of other properties that drive value to them. There's their mobile properties. If you look at the thing on the right, that's an affiliate system, so you can actually pay 50 bucks and it'll create a complete amazon.com for you and drive traffic back to Amazon. And you can arbitrage the cost of your Google ads to your affiliate earnings. And the company is famous for actually making all its internal developers use their own API.

That's one of the things that Amazon has done, it's allowed them to open pretty much every part of their business in a certain way, back to the public and to earn money from it and to arbitrage kind of the costs that they had. So they've created really a revenue ecosystem around their core competencies. We saw a graph earlier on, and I think there's actually even a bit more and the numbers even more impressive. So public directories of APIs have now reached sort of around 6,000, but we think there's actually about 20,000 APIs out there.

There's way more than, than actually just listed as open APIs. There are backends to mobile applications. There are business to business APIs, there's all sorts of interactions. And if we see projections, so we, we really believe that there's around 250,000 APIs we're gonna have within the next five years or so, which is a huge number from where we are, right? Maybe more, it may be less, we don't know, but there's a hundred percent compound growth right now. And people are opening data all over the place in various ways.

And, and the technologies we hear like rest and Jason and so forth, they're really powering this and more and more platforms are being enabled. Another thing to say is we see around 10% growth in the volume of traffic to the APIs that we power per quarter. Right now that's an average. Some are growing extremely fast. Some are doubling per quarter. Some are more than doubling per quarter, and some of them are growing more slowly. So you've got a number of APIs, which is increasing.

You've got the numbers of companies that are opening number is increasing, and you've got the traffic volume increasing, I guess, it's that that's really a big deal, right? If you think about the way the web works right now, and the way mobile is kind of moving and becoming a, a way of consuming web content, and you can see the traffic shifting from the current web to this kind of API web. So there's really something going on, which is pretty important. Next thing is diversity.

So one thing is that if you look at APIs, you, you kind of see headlines on Twitter and Facebook and how many billions of transactions they have and all these kind of things. And those are very important.

They're, they're kind of leaders in the space, but the really interesting things to ask is how diverse it is. So again, this chart comes from programmable web, which is probably the biggest directory of, of open APIs. Those are just the top segments of, of APIs that, that they had at a certain point. And that this, it hasn't shifted very much. So it's really diverse. And you can see across the bottom different types of APIs.

So I have a Nike sports watch when I run at the end of the day, I plug this in and that talks to Nike and it uploads where I ran to GPS and tells me how fast I was or normally how slow I was. And, and so NY has built a product ecosystem around their it infrastructure. And by the way, they're a shoe company.

It's like, wow. You know, that's pretty crazy the fact that they're integrating themselves into our lives in this way and all the way across the bottom, you can see like from internal APIs to kind of partner APIs business, to business, whereas data sales, which is really the, the, the driver all the way to the big public APIs, where you see really large volumes of, of transactions. And I'll talk about some of those examples.

So of course I'm a little biased, but we really believe that APIs are gonna become the major way of doing business on the web and not just for a, kind of an eCommerce transaction and content, but also for a lot of physical goods and physical things that happen in the world that are gonna be mediated by these transactions. And the way to think of it is to think of your enterprise inside out.

I think most of it, developments in enterprises have been focused on gathering those internal resources together, turning them back out, producing a very finished product and delivering it to a customer in a very focused way. What you're doing with APIs is you are opening out your internal processes in various ways and allowing other parties to do a lot of parts of that value chain for you and take partial products even, and recombine them and turn 'em into other things. So you're really turning things inside out.

And that's pretty important if you think about who you're authenticating and who you are authorizing to do all of this stuff, and who's the end consumer, right? So I'm just gonna cover a couple of common models. So we see kind of four different, let's say business models that underpin the way APIs work.

So we, this is kind of across the customers and we always try to classify kind of what kind of API people have when we talk to them. So the first one is the API is the product, right? So that means you have something where someone is literally making a call to a system and they're paying for that right there.

And then, so if they're sending an SMS with Twilio, for example, that's what they're doing. It's essentially your API is a product. If you go to the website, the only thing you're gonna find is an explanation of how to use API, right? That's all you get. The second one is that I'm gonna put the examples up actually makes it easier. The second one is that the API projects, the product that means your company has a bunch of core competencies that people are using maybe via web interface.

Maybe you might another interface and the API allows your customers to use that core competence in different places, be it on a mobile device, be it embedded within a partner system, be it completely mashed up and turned into a different type of interface. And the third one is you're promoting the product.

So you see a lot of APIs that allow companies to take their data and give it to other people for the purpose of basically promoting their brand or pushing affiliate links back to themselves, or essentially taking something, which by itself doesn't have much value, but links back to the core. So I can't use, so if I take the open platform, which is the guardians content, I can take snippets of guardian content and put them anywhere on the web, but I'm obliged to put the guardians logo there.

So basically I'm sort of getting some use of the product, but really what I'm doing is feeding the guardian. And the last one, which is probably not so relevant for most of this audience, but it's APIs that power and feed the product. So Twitter receives more than 75% of its traffic by its API and Twitter wouldn't exist without its API. Essentially all those tweets that you're sending, they create the product, right? You've created a feedback loop.

And if you think of something like YouTube, YouTube's APIs have a lot of different upload tools that allow people to put live video and all sorts of things and just chunk it up and put it out there. And if those weren't there, the amount of content going into YouTube would be much lower than it is, which some people might say is a good thing. But I presume that YouTube kind of likes that I'm gonna skip this a little bit detail, but you'll, you'll have them in the slides. There are a lot of different reasons why these have different types of value.

And I wanna touch on something that Phillip was talking about internal developers as well. So if you think of your core assets and the beginning, you have the actual asset itself or the service or whatever it is, it might be data. It might be a process that you link to. And often your first step is to internally reuse that. And most of that was done with Soer tools and, and things like that, especially in later enterprises, but you want to get that data to different departments. There's actually some value in that there's some significant value. The next step is this kind of customer reuse.

A nice example is Bloomberg. For example, they've always delivered their data through a terminal. You have to buy a Bloomberg terminal to actually Bloomberg terminal to get this data, right? So they've locked you into their terminal model.

Recently, they've added an API to that, which is free. So with your subscription, you can use the API as well. So what they're doing is they're getting their customers more hooked. They're getting their customers more hooked on their data and getting their customers to spend effort and resource to integrate that. So they'll never remove them. And I can bet you. Now it's a side show. I can bet you in five years, there'll be a significant flip in where the real stickiness for their product comes from.

And those terminals one day, they probably won't be there anymore, but Bloomberg data will still probably still be there. The next step is kind of using the API to distribute their content. So finding partners, finding resellers or the service, so it might be a service. So we have number of customers that provide SMS services and they essentially resell at and T's infrastructure, but they have their own API and they insert their billing and they add some value.

And here's the last one that most people think about when they think about open APIs, it's this kind of idea that I'm gonna put the API out there and then thousands of developers will come and build lots of cool apps. And that will be really good and I'll get ideas and I can maybe acquire the best apps and so forth. And I think people tend to focus on this on a lot. It's an amazing thing when it works well, and it's really something to aim for. But the most important thing about this picture for me is that you get value from each one of these steps, right?

You don't have to go all the way to the last step. You actually have a slow opening of the kimono as you, as you kind of can think about it and you can control them all in a pretty closed way. So I don't wanna take too much time. I might skip down these a bit, but they're essentially taking each of these four models that we had before. There's some examples, the main point I wanna make about each of them is that the value is different. So if you're a, a Twilio or a Skype or something like that, your main business is essentially making these transactions within the system.

And so really your business value is that you are increasing the usage of your core service. So you're getting dollars out. If people are calling the API more, right? Every time they call you, you get paid. If it's a projection type of scenario, if you take something like Salesforce, you have more than probably actually a lot more than this already 50,000 developers who are writing applications. So you're projecting the product into places where people might not be able to use you.

So your real value is you're increasing your customer spend because they're spending more time with your product. They're using it in more ways, they're using it in more parts of their organization because they have different new interfaces onto it. And your secondary one is that you might get more customers because a customer who didn't like going to the salesforce.com website, they do interact with the mobile app, or they do interact with this integrated into some other system. And so they're more likely to use you. Third one is the promotion.

So I have Amazon example, but there's actually plenty more. Netflix is a great example, Netflix.

And in, in Germany, it's max stone, they are content companies. They sell video on demand. That's their core business. They don't really deliver that by API they're have proprietary technologies. So what does their API do? Their API does metadata. Their API does all of the data about programming recommendations, ratings who likes what, when things are on etcetera, et cetera.

And that is extremely powerful because now people can build applications that play around with Netflix, metadata, do cool stuff with it and drive the core product of net of Netflix and of, of, of max stone, which is to use the video streaming service right now. I'm gonna skip this one a little bit just to make sure that we too much time, but I, I think I said it before, just kind of couple more key messages. One is that what's happening when you do this is the people that you're authenticating to your service.

The people that are using it are almost certainly very, very different from the people that you're authenticating. Right now. You may have internal developers. You may have different departments of your organization that are using it, but really you're getting into partner ecosystems, ISVs contractors, affiliates, when it's customers, it's their, it departments, their applications, their integrations that you have to interact with, and that you're authorizing. And that may have to do with your actual business model.

It may be tied into different tiered pricing plans and things like that, which you normally don't need to do within your own organization. And on the wider world, you have independent developers and you have the end user who's often involved. And Phillip mentioned a great thing before, which is this disintermediation issue. So that's increasingly very worrying, especially for content companies that Google Facebook, apple, and, and so forth are in between them and the customer.

So actually it turns out that APIs can help them get closer to their customer by actually managing those subscriptions on an API basis, and to keep in contact with their customer and know who they are rather than having that, that, that wedge driven between them. So this is just the last, how many programmers there are in the audience, but the kind of last idea. So does anyone know what NBC is?

Model, model, view controller. Some people at the back the program was always hired at the back, which is great. So model view controller is a paradigm. That's just really been the dominant paradigm of building web applications in the last number of years, it's been extremely successful. It's the idea that you separate the data model from the way it's visualized from the way the business logic works. And so you can change these things differently separately, independently. And I think, and then the way we see it at three scale is that this can happen at a business level.

So you can think about, are you primarily a company that produces a core product or a piece of data or something like that, or are you a company that primarily has an audience that has a view that has a way of shaping content and showing it to people, or are you a company that orchestrates that aggregates that, that combines, that can go from anything from a consultancy that looks at data that's out there and makes recommendations to a company that takes, for example, a smart car and customizes it for a particular company as an advertisement.

So I think it's worth as a company when you're thinking about APIs is thinking about what your core competence is and thinking about whether or not you also try to cover everything. Like, do I also manage all my channels? Do I also provide all the transformations or are there other companies that can actually do a lot of those things better than I can? And can I be the one in our se in my segment that really builds up that ecosystem before other people? So I think that's the power is to try to look at, I M M an M a V or a C.

And if, or am I two of them, and if, if I don't have all three or I don't want to have all three, how do I build the strongest possible relationships? And normally there's a lot of data flow by API that you want to put to make that happen. And so that's pretty much it. So we think that we're very early in the game, even if there's 20, 30,000 APIs out there, that's, you know, there's a lot more, that's gonna be exposed.

And I think people tend to think that APIs is just more or less, just the last part of the chain that opening up to all this innovation of, of small developers building apps, but it's a huge shift in terms of data flow. So that's it. And thanks very much. Thanks so much, Steven. I think there was a great overview of how the, the business side actually looks like today and how many aspects are there.

So if someone in the audience may have asked themselves how to make a business out of API, I think you're the role model for that, that really rocked my world just as I, as I had with Phillip doesn't, that NBC model enable me to concentrate, really concentrate on what I do best and leave all the rest to the others, throw it into the market and, and make them my product, a better product. I think it does. I think it allows you to partner in a much smarter way than you did before. Of course you wanna mitigate risk.

So if you're incredibly dependent on just one player in one of those other dimensions that causes you another problems, but gen I'll give you an example. There's a very large printing company in the us that we've worked with probably about 10 years ago, they did some massive integrations with two or three partners. They cost about one to 2 million each integration. And then on the other end, they have a widget system, right? So that you can download a widget and you can do print on demand with this widget, but whole of the middle of the market is gone, right?

There's no way you can pay a million dollars to integrate a 500 person company that would use that service. And that 500 person company is no way happy with the widget, right? So I think you can, you can focus on your core competence, which is printing and you can export it to as many places as possible and build audience and, and have other people resell that service for you and so forth. Yeah. Okay. Thanks so much. I'd like to invite full back on stage and please take a seat and we'll just have a few final questions to discuss.

So we've, we've seen the very basic discussion about what the API economy looks like, how, how it develops. We've seen how a Phillip described how the consumer view of this actually is. And we've seen how organizations really make a business out of providing APIs for so many different purposes.

What, what I'd like to know from the panelists now is how, how could this be fostered in the, in the economy? Because right now, I only see lots of those medieval castles that Phillip was talking about.

And, and what I rather see is that they slowly turn from building rectangular castles to now more round castles, because they found out somebody is coming with a cannon and they bounce if we build round castles, but they, they, they didn't recognize that it's actually better to tear down those walls anyway and provide little sneak, peeking holes. The APIs that would actually enable me to provide what, what I can make the world better with what I can make business with and keep all the rest away.

How, how can we actually get to the point where the growing number of APIs actually revolutionizes, what, what we do today, Martin? I, I think it's the other way around, you know, and that's what, what you also said. I think the point isn't, I think Phillip said matters. The point is this thing is not driven by it. So this thing is driven by business. If you look at printing example and many other exams, Amazon, and all the others, it's not that they change their business because it opens up some APIs, they open up some APIs because business is changing.

And what we have here served an answer to a question we are dealing with for quite a while, and a better answer than I've heard until now. And I think that's really the point what it is about. So I think we, we don't have to convince business. We just could say, Hey, finally, we have an answer on what you are asking me for quite a while.

I, I just wanted to add one point that that we've used interchangeably, the word organization and enterprise. And the reason that I was really specific about saying organization is because it's, it's more than just companies that are involved with making money. We're talking governments, schools, public utilities, all sorts of organizations that, that are involved with the ecosystem of our existence need to be involved with the open a P economy and getting access to that data.

So it's, and it's more than poking holes, holes in the round walls of the turret or the keep it's digging up the moat and opening the gate. So, so it's, it's rather like enabling factors so that organizations, organizations, not, not companies are, are able to show the world what, what good they can do for a very, very low price, because they just say, Hey, here's my API. If you like that service do it.

There's, there's very little cost attached to it. And this enables even NGOs and all those non-profit organizations to really have a big impact. Is that right, Kim? You know, I, I, while you were speaking, I, I was thinking about the, the French, the, the, the Parisian bus system.

So the, in Paris, you know, they have GPS in all the buses and then they have a restful API. So you can actually, you know, ask when is my bus gonna be here.

And so, you know, I just wrote a little app that would, would ring my bell on, on my phone when it was time for me to go down to the bus, which took me about two minutes. I just wrote an app that rings my bell, But the, you know, but then I went back to north America where of course we are missing a few of these modern, I don't know, conveniences, but it, you know, in, in most north American cities, the public transit is unusable. But if it were combined with the system to ring, you know, to tell you, okay, now it's time to go to the bus, stop.

It suddenly can completely transform what public transit is. We have a better mode here in Germany.

The, the public transport actually is on time. There's a schedule Doesn't have to be on time. On time is old fashioned that's the on time is pre binding. That's the pre binding site model. All it has to do is, is talk to you.

And, and so like all of a sudden the whole infrastructure can become, you know, communicative it, Schutze down your computer. I think this is, this is really beneficial. There there's a personal benefit for a single man, but potentially for all the Parisians for, for, for all users of public transport, you know, there's, there's okay.

There's some investment into infrastructure because you need to know where the buses are, but the, the actual providing of that API might have cost a fraction of, of the infrastructure cost and everybody could, could just build All they need to do is put an iPhone in each bus. Well, actually they could put in a windows phone for me, much better.

This, this is where we see it just wouldn't is Cangen because he does not take the taxi used public transportation. Yeah.

It, I think, you know, in, in every single city in Europe, we had middle-aged castle, they had in Paris, you know, in the city I'm living, we still have part of it and all those castle fall down and, and they just fall down now and will be same thing for the enterprise. Now, when they fall down, some people want new market opportunity and some people disappear.

And, and I think that's also obvious, you know, some company are going to get advantage of this new model of, of economy and some people are going to die. That's that's also something we have to accept, but it's a big shift. Okay. And it change a lot of things. It's not only the way to develop application or security. It's a big, big shift on the way we can, we can handle the business inside the enterprise. I think Steven, I think a good way to think about it. Like the bus example is great. So we have Ottawa transport and also in, in London, the same thing.

And if a city is investing in helping its citizens use the bus, it can do a bunch of different things. One of its could do is it could build a bunch of mobile apps, which a lot of cities have done. So they've invested in iPhone apps and, and, and Android apps and all of those things.

And they, they, they work well and, and people like them, but you have to update them. You have to secure them. You have to keep, keep building those things. And I think there's a trend that, especially those kind of public kind of data, it makes sense. What's the city's obligation is to get the data out there in a stable way. And most of the rest of it will take care of itself and you'll get 150 niche apps that do exactly the thing that different populations need rather than having the cost of maintaining all those interfaces.

So, so that actually boils down to that. The new approach of the API economy is a, a business enabler for very, very low cost. Right. Wonderful. So I'd like to thank the presenters I'd like to thank the guys in the panel. Thanks for giving us such a brilliant discussion about how the API economy affects and impacts our future it model. Thanks guys.