SILICON VALLEY INNOVATION CENTER
We help global corporations grow by empowering them with new technologies, top experts and best startups
Businesses and consumers today demand seamless user experiences and integration platforms as a service, or IPaaS, promise to deliver just that. To explain how IPaaS works we speak with Simon Peel, Chief Strategy Officer and CMO at IPaaS provider Jitterbit, in a livestream interview. Simon will discuss the trends driving growth in the IPaaS market and talk about how integration platforms can contribute to an organization’s digital transformation strategy.
It’s great to have you. So, just tell us to start with, IPaaS, Integration Platforms as a Service, perhaps just a brief overview. What is IPaaS for those out there who don’t know and why might a company want to want to implement it?
Okay, well, let’s start with integration because this is old as the hills. Ever since there were two systems, people have said, “Well, how come there is data over here as well as data over there, and how do we get this stuff working together?” And that’s often been done on-premise and it’s been connecting on-premise systems together for a long time. And we have buses to do that and the likes of TIPCO and web methods and that sort of thing. But what’s happened, as we all know, is there’s been this tremendous shift to the cloud, cloud-first for everything, lots of SaaS applications, a big explosion of technology. And so the question is, is it really the right thing to do to continue to integrate everything on-premise when there’s all this stuff in the cloud? And so how do we put that stuff together? And the answer, really, is a hybrid of the two. And this is where this Integration Platform as a Service idea has come from, where, as things move more to the cloud, we want to run more integrations in the cloud.
However, of course, the world still exists and so we need to be able to connect back down to on-premise systems. And so integration platform as a service, although it sounds like it’s probably only in the cloud, actually is a way to connect things in a hybrid way. But that’s where it’s born from. It’s this born from this massive shift up into SaaS and Platform as a Service itself and also all of the cloud-based systems today.
So how much do you say is in the cloud now? Is this pretty much everything, we just assume any business today is pretty much in the cloud? Or what’s the split? And how that’s happening?
Yeah, it’s a very good question, actually, and the answer is Yes and No. So I would say that, certainly, all of the new stuff is being done in the cloud and all of the existing stuff is largely being left on-premise, and that’s the right thing to do. So it’s often said that data has weight, meaning that, wherever it’s created, it pretty much has to sit. And if you look back in all of human history, we’re creating as much data as has been created in all of human history every year or so now, and so you really can’t shift that around. That new data from IoT devices, from everything we’re doing in mobile, and so on, it’s all being done in the cloud. However, we’ve got most enterprises still with the vast majority of everything they do on-premise and so this is where you have this thing where you’ve got to connect all that stuff.
Now, if you look back about five years ago, there was all talk about, “Shove everything into the cloud. Let’s take everything and put it into Amazon and everywhere else.” But we’ve realized now as an industry, where does that really get us? It doesn’t actually move the needle very much. What does move the needle is getting all of these new systems provisioned in the cloud, getting IoT devices, getting mobile, getting voice, getting AI, and so on – let’s get that stuff going. And then let’s just run a cable down to the existing systems on-premise or into other cloud-based systems and start innovating, as opposed to schlepping data around, which is really not that exciting.
And so, what does the IPaaS market look like today, in that case? We have all these different cloud services, AI, voice, IoT, these things only seem to be growing. Digital transformation is a very hot item on everyone’s agenda, everyone is trying to move in that direction, so it seems to me that having an integration service is going to become, I suppose, essential. I say that objectively.
Yeah, it was very nice of you to say that, I do appreciate it, but it is actually just logical, no doubt. So there are now 149,000 SaaS apps out there. As one example, everyone’s heard of Salesforce.com and NetSuite and so on, but most people don’t realize there’s 149,000 SaaS apps. But that’s what’s happening because people are no longer creating on-premise software to try and sell, they’re just creating it in the cloud.
So the big transition that’s happening as well is, you’ll remember in mobile, when the iPhone came out, Apple created the iPhone, you looked on the device and it had a few little squares on it and they were all from Apple. Well, now the App Store came along and then we heard this phrase come into our vernacular, “There’s an app for that.” So, if you want to check your bank account, you want to know what the weather is, you want to know how your retirement is doing, you want to know how to get from A to B or eat in a restaurant, there’s an app for that.
Well, guess what? Now there’s a SaaS app for that and that’s a huge difference inside any IT company. IT are used to provisioning SAP or Oracle and maybe there’s three or four or five systems. What happens when there’s 149,000? And that’s just SAS apps. If we look at AI, well, even three or four years ago, we used to think AI was about just getting an AI platform and loading it full of data and trying to do machine learning and, eventually, some rocket scientists would come up with an a-ha. You know what? There’s now 400,300 AI companies out there that are each of them doing one specific thing for AI. Blockchain is another; there’s now 140 blockchain companies and more than a billion dollars has been put into that in a single year. So technology, is fair to say, is exploding everywhere. And what you’re pointing out is that as that explosion happens, someone’s got to try and tie all this stuff together. And pretty much everything I just talked about is an explosion in the cloud, not an explosion on-premise, and so that was what’s driven this vast and massive increase in the need for IPaaS.
Now, the IPaaS market has been growing around 70% per year, a compound average growth rate. That is really good for quite some time now. And Gartner kept predicting, “It’s going to tail off, it’s going to tail off” and every year I talked to people like Massimo Casini who would look at it and go, “Simon, this is not tailing off, is it?” And I said, “We just did 130% growth in Europe, we’re doing over 100% growth year on year in many territories, so no, it is not tailing off. It is absolutely exploding.”
So, all I can tell you is that there’s a vast amount of interest and Google searches going on around, “How do I integrate all this stuff together?” And the question is, how do I do it quickly and easily? Because integrating Salesforce with SAP is one thing or SAP to Oracle, that’s another. 249,000 things and 4,000 AI, this better be fast and easy. So that’s really transition, the next wave that’s occurring now.
For sure. Given there is so much out there, thousands of apps, SaaS products and so on, what do you think for the companies out there, for the businesses looking at all of this, how do they go about choosing which of these apps they should take? Do they need to think about integration very early on in that process?
Yeah. So, it’s funny, I would say I’ve been working with Salesforce maybe 15 years or so and there was always this phrase, “Phase Two.” And integration was phase two. Phase one was you buy your Salesforce and you start using it and everything’s going to be great. And then, eventually, you’d find out that, “Where’s all the data? How do I get it? How do I connect it?” And that was phase two. What you’re pointing out is that integration is absolutely phase one now. Nobody buys stuff in the cloud, thinking, “Oh, I’ll worry about how I connect it later” – everything needs to be connected immediately.
So when we’re getting involved in deals, it’s normally when a new infrastructure is being put together, when a new application is being thought about being brought in, when consultants are doing what you talked about earlier, digital transformation. “We’re going to digitally transform our ecommerce system. We’re going to digitally transform the way we deal with customers, we’re going to mobilize things that have only been on the website to date.” Right baked into that is, “And how are we going to connect all that stuff together?” And honestly, the existing on-premise or bus-based platforms of the past are not even in the consideration set. It is, “I’m doing IPaaS. Now, which one?” So the question becomes, “How do you choose which IPaaS?”
And if you look at Forrester, you look at Gartner, you look at a lot of the research out there, some interesting things emerge. One thing is that, out of the top four big things they say you should look at, only one of them is actually product. And that’s a turn-off for the books because, as technologists, normally we’re saying, “Well, how many features are in this one? And let me look at all the checklists and so on.” But what they’re saying is, things like the cost over time, the total cost of ownership of one of these things is really important. So, “Do you charge me every time I send data? Do you charge me every time I add a developer or somebody that’s going to be doing some connectivity?”
And then we get into support. So they’re saying, not just total cost of ownership but, “How does the support work on this thing?” Because it seems okay when you’re doing a trial for it all to be online. And, “I’ll just look on a forum or something like that.” Or, “I’ll figure it out by watching some YouTube videos.” But when youe company is running on this stuff, it doesn’t really cut it when the system goes down. So, “Who do I call? What are the SLAs? How can I get fixes rapidly?” And so on.
And then it comes down to also “What is the partner network around this? Are there other people that can use it? Is it straightforward enough to use or is it a six-month training program?” So with us, most of the time it takes maybe two to three days or so to get up to speed to the point where you can actually start to do things really well. And although we have a lot of developers that use the product at one end, on the other end we have people who are sociology majors who are picking up the product and actually getting into production in less than 30 days. So you can hear that’s quite a tectonic shift from the older, on-premise days, where you had to be an expert for over 20 years to be able to use things.
So those are some of the shifts and changes that are happening. The cloud represents rapidity and speed and innovation and so the platforms like ours have to be super easy, have to be able to have a bunch of built-out templates that that exists already and so on, and they have to be well supported.
For sure. So how do you guys keep up with all this, with the speed in digital today? I mean, as that is your business, to be on top of this thing, I suppose, that you can offer these integrations to other enterprises and have developers use it but also sociology majors and make sure that it is connected with other potential partners. And there are so many moving parts, I suppose. And in digital, it all can change quite quickly. So how does Jitterbit itself, how do you get ahold of that and make sure you are keeping on top of it?
Well, there’s no getting away from the ugly truth that there’s a lot of moving parts, there’s a lot of development involved, lot of testing involved. I think the point is that you make it Jitterbit’s problem as opposed to being your problem. So, when NetSuite comes out with a new release and deprecates some old functionality that would normally take down your system, we’ve already had access to that for the past three, four or five months, we’ve been working with NetSuite through the partner program. Same with Salesforce, same with SAP – that took us over a year to get certified on – and so on. So we make that investment so that you don’t have to do. That’s the essential value proposition.
One thing I will say, though, is 149,000 times? Seriously? I mean, we’d have to have half the planet working at Jitterbit to be able to keep up with that. Well, the good news is that, as technology evolves, we get standardization and API’s have really helped with that. And if you go back 15 years, 20 years, Oracle had one way of drilling into Oracle, PeopleSoft had another, SAP with the BAPI interfaces and IDocs, and so on, it’s totally different, but there were only a small number of them.
However, now we’ve got much more standardized interfaces and the key is APIs. So APIs are relatively standard and they differ. And on 149,000 apps, sure, every REST-based API isn’t identical, but they are more similar than looking at an Oracle versus an SAP. So that’s the sort of partial secret of how we’re able to keep up so quickly. We still have to get those systems, we have to prove we can do that, we have to test that we can do there, but we’re not reinventing the wheel every time.
So tell me a little bit more about APIs, I wonder how those work a little bit. So one way that I read them being described in very much layman’s terms, which works for me, is that they are the waiter whereas you’re the customer at the table. The API is the waiter which goes to the kitchen and gets what you want and brings it back to you as the customer at the table. And so, presumably, in the in the context of IPaaS, this is happening between programs, between different data sets, applications, the API is where the work, the exchange is taking place. Is that correct?
You know, I’ve seen that analogy and it’s an interesting one. I would use a slightly different one, which is that API is more like a plug socket. Imagine everybody having to access power grids and solar panels, and having to create all that stuff, who’s going to do that? So everyone has plugged sockets around the entire world. The only thing is, if you travel around the world, you realize that, “Oh, this cable doesn’t plug in. I can’t power my laptop. I can’t get on this discussion today, on this webinar, because I got the wrong power socket.” So what APIs do is give you a consistent power socket – they all look the same – which is great. So now, heaven forbid, but let’s just say the American system took over the entire world and everybody had the American power socket. Great. There’s your API.
Now, the reason why the waiter thing doesn’t work for me is that API just sit there. Power sockets just sit there. It’s just a hole in the wall. Well, what do we plug into it? And if I want to connect different things together, I need these cables and I need some logic in between that, “Well, which power socket do I plug into which device? And how does that device relate to the next device?” And so on. So, you know, in this analogy we’re trying to build here, it would be a computer and plugging that in, and then a monitor that plugs into the computer, and then a mouse that connects and so on.
So that’s the IPaaS component. So the APIs are I just there to say, “Yes, I’ve got a pretty consistent way of connecting to me,” but they just sit there. The IPaaS systems are there to go and plug into those different sockets, the 149,000 SaaS apps, and then move the data around and apply logic to it.
So when somebody creates an opportunity in Salesforce.com, who should know about that? When it turns to an order, as an example, then you should go to NetSuite or you should go to SAP and tell them, “Hey, there’s an order.” And then those systems have to come back and say, “Well, that order is now being processed.” And go and put that back into Salesforce so that the rep can say, “Oh, look, my order is being processed.” So that’s how APIs work to and standardization really helps things like IPaaS move forward.
Thanks, Simon. And you talked a little bit about the data there going between different applications and therefore they can execute on what they need to do and so on. And I noticed on your website you mention artificial intelligence and how IPaaS can work with artificial intelligence. And it strikes me that what you said there about the data and then creating orders and so on, that can all be automated, I suppose. And I just wonder, what is the role here of artificial intelligence in all this? And where might that take us?
Well, there’s two elements to that question, which I’d like to separate out. So one is there are 4,500 AI solutions sitting out there right now that most people are not using, and shouldn’t they be able to do that? And the other is, well, when we’re doing integration, shouldn’t vendors like us be using AI to predict what kind of integrations you’re going to do and do it for you if we can or at least try and make it easier? So let’s take those two things in order.
So, on the first one, everybody already has applications and systems and mobile apps they’re using and everything else. Wouldn’t it be nice if we just applied some AI to it and turbocharged it?
So I’ll give you a quick example. The labor market is incredibly short everywhere around the world today. There’s almost zero unemployment in so many different areas and so losing staff is a really big deal. It turns out the 30% of people leave a company because of cultural fit. Not because they can’t do the job, not because they lied on their resume but because, you know, “I just don’t like the environment here. It makes me work all night or work on the weekend.” So there’s one AI vendor called Human Intelligence and what they do is a 12-minute survey asking you all kinds of wacky questions that you are like, “Where is this going?” Like what color do I like and who are my friends or whatever it might be. But it comes back with a simple magic number that says what the cultural fit is going to be for you with that company. So the pitch very simply is people can save 30% of their hiring costs with a simple 12-minutes survey. That’s millions, if not hundreds of millions, for large companies.
Walmart, with 2 million employees, saving 30% of their turnover. The AI system is sitting there, knowing what it can do. Now, AI out there is one example. How does that help Walmart when they’re using Workday as a hiring manager looking at somebody? Well, it doesn’t. So what we can do is go into the Human Intelligence system, take the information and say, “Look, for John Smith, who’s applying for this job, I’ll drop that right into the Workday screen,” so that as the hiring manager looks at Workday as they normally do, “Oh, well, 1/10. Wow, that’s not good.” Now, the system could also say, “But this guy’s a 10/10 for the Telephone Support Team at Walmart. It shouldn’t be a greeter, but it’d be fantastic on the Telephone Support line.” So there you go. One example of what we call AI infusion. It is infusing AI into your existing business processes, literally in days. So we can save Walmart 30% of their hiring costs in a few days. And it’s logically, how that can happen.
Right. And so, Simon, saying there the chap, John Smith, he might be right for the telephone team, that comes from the IPaaS, that insight.
No, that comes from the AI tool as well. So the AI tool is going to go, “Yep, you asked ‘Is this person a good fit?’ Don’t think he is. But by the way…” So what the IPaaS does, though, is say, “I’m going to tell the hiring manager, ‘Not a good fit for you, I’m afraid.’” But it can also go into the system and look at someone who’s not a hiring manager, who is actually over in the Support Team and our intelligence says, “Hey, you should know about this guy.” And that pops up on that person’s screen and they go “Wow, this person is perfect for me.”
So we’re having that intelligence. But the main message is we’re doing the infusion. So if you can, don’t look at any of the 4,500 AI tools out there, don’t change a darn thing about the way that you actually code all your systems and so on and just infuse this massive intelligence into existing business processes today and get it running in the next few days, weeks. So that’s the first category.
The other category I think you mentioned was, what are we doing with AI to make things easier? And, of course, we’re have 62,000 customers working on the system at any one time and so we’ve got a lot of knowledge there as far as what connects to what.
When you connect Salesforce to SAP, as an example, we can predict pretty closely the kind of fields you’re going to need. You are going to want order to go back and back and forth bi-directionally. We know that BAPI Customer Getlist is the right API that we call and use on the SAP front and where to put the resulting information back into Salesforce. So we’re using AI to do that and to be able to speed up the way that you do implementations.
We’re also, though, on the cusp of releasing something literally in the next few weeks where you’re going to be able to use your voice to do these integrations, you’re going to be able to ask for data. Because, ultimately, if you think about applications, those are things created by developers to try and put a screen on the front of the data you actually want to get to. Well, if I’m going to a meeting and I’m driving there, and I think, “Hey, should not be able to ask the tool? Say, ‘Jitterbit, are there any escalations or red alerts for the customer that I’m driving to??” Its Nabisco. And it should say, “Checking. One moment, Simon. Yes, there’s actually a red alert status.” Or I say, “How much is the customer I’m driving to right now bought from us over the past year?” And it says, “Well, worldwide, $1.7 million.” “Really?” So that’s what I want. I don’t want all this application stuff and integration in the middle and AI and all the rest of it. Just give me access to the data. And that is what we’re going to be releasing over the next few weeks, which I think is absolutely game-changing. It combines NLP and natural language processing together with a bunch of different AI that can then find that data, build the integrations and get it back to you real time.
That sounds very, very exciting, Simon. I think AI, conversational AI, natural language processing is something which we often hear is coming, is coming, but we haven’t necessarily seen any breakthroughs yet. So I think, Simon, on that note, we will probably draw our interview to a close today. So perhaps just as the last question there, do you see going forward that AI and conversational AI and artificial intelligence will come closer together and really make computers and systems like Jitterbit so much easier for us to us as humans, where we can interact with them, as it were, on our own level and just ask questions and get answers that we need? I mean, is this the way that you see things are going to be moving?
Simon Peel: So I think we’re entering an incredibly exciting time, because there’s so many pieces of new technology that are hanging out there just waiting to really zap. Because the information is in the cloud, we’ve got AI systems that are intelligent enough to help us, we’ve now got these integration platforms that can pull all this information together, we’ve got systems that provide things like NLP so I can actually speak to in my normal voice. You’ll notice Alexa. A few years ago, Siri comes out, it’s a little bit awkward to use and so on, but everyone uses Alexa today and it’s so simple, so easy. It doesn’t say, “What? Repeat, repeat” all the time. So these technologies are all there.
And there are these moments in technology history, like there were web pages out there and there was this sort of network thingy that people could access, basically with dial up modems and that became the Internet because of Netscape coming together with web pages together with a broad network – and our entire life changed forever. And now that is exactly what’s going happen very, very shortly with the advent of AI.
AI has been around since the 1950s, there is nothing new about it. What’s new about it is that every single application and every piece of data in the world now is having a plug socket put on it. And so things like our API platform create these plug sockets automatically now, for every piece of data in the enterprise automatically. Now, because of that, all that data is available, the plug sockets are there, we have the technology to plug into those things and we now have the AI available to be able to use the natural language processing, so we don’t need apps anymore, we just ask questions of the data that we know is already out there. So this is about to happen, about to explode big time. It’s an incredibly exciting time.
Brilliant. Well, thanks so much, Simon. It really does sound fascinating and it’s very infectious, I think, the way you tell it as well. So, we really look forward to seeing some of this come to fruition.
But it seems, Simon, for today that’s where we’ll draw things to a close. So just thank you, thank you so much for joining us today, taking the time to be with us today on the program.
Exciting. Thank you for the opportunity and see you next time.
And to the viewers out there, I hope you’ve enjoyed my talk today with Simon Peel, the chief marketing officer and strategy officer at Jitterbit. I just want to remind you that we have another expert talk coming up next week and it’s funny, today we’ve been talking with Simon about conversational AI and about natural language processing. Well, we’re going to have three expert speakers join us next Tuesday to talk just about this, conversational AI. We’re going to be delving a little deeper into that, which is one of the hottest topics today in digital transformation.
So do join us for that on the 27th, next Tuesday at 10 am Pacific and see our website siliconvalley.center for details. But for today, that’s where we’ll wrap things up. So from me, from my guest today, Simon Peel of Jitterbit, thank you very much for watching and we will see you again very soon. Bye-bye.