Drill to Detail - Drill to Detail Ep.25 'SAP BusinessObjects Analytics, AI and Digital Transformation' With Special Guest Timo Elliott

Episode Date: May 1, 2017

Mark Rittman is joined by Timo Elliott, originally of Business Objects and now Innovation Evangelist for SAP, to talk about the origins of self-service BI with Business Objects' innovative "Universe" and the role analytics now plays within SAP; why analytics is the most important function within your organization and why the vast majority of analytics is still reporting (which isn't so bad); and the role AI and other innovations will play in analytics going in the future.

So hello and welcome to another episode of Drill to Detail and I'm your host Mark Rittman. So my guest this week is someone many of you will know from his blog on analytics and for speaking events around the world and he's actually SAP's innovation evangelist, Timo Elliott. So Timo, do you want to just introduce yourself properly and let us know kind of like how you got here really? Thank you very much. I'd love to. And hello everyone. And thank you, Mark, for this opportunity to come on the show.

So my name is Timo Elliott. I'm an innovation evangelist for SAP. My background is Timo Elliott. I'm an innovation evangelist for SAP. My background is I'm English. I'm from the tropical part of the UK down there on the sunny south coast. I grew up near the beach, I think California, with a bit more rain, and then went to Bristol University to study econometrics. And after I finished my studies, did my own personal Brexit, worked in Asia for a while, a year in New Zealand for Shell, and then decided to be a good European. I came to France, did seven years in France, then transferred five years in Silicon Valley. And now I've been back in France for another 10 years or so.

Wow. And so Timo, I got to know you first through kind of business objects and the work you were doing there. So what was your role there and again, how did you get involved in that before? So I was lucky enough to be the eighth employee of a small French startup 26 years ago called Business Objects, obviously one of the pioneers of business intelligence. I was hired because I fit the primary criteria, which is I was cheap. And I knew something about computers. So my job was I did a lot of things over the years, but helping write the first English language manuals, creating the icons for the first English language manuals, creating the icons

for the first Windows version of the product. I ran competitive for several years. And slowly over the years, I became more and more of an evangelist, spending a lot of my time trying to explain the technology and finding out what our customers were doing on the leading edge of analytics, taking that information and then sharing it with other people trying to do the same things. And then BusinessObjects was acquired in 2007. And so I have come over to SAP where I'm doing essentially the same thing, but now on a broader stretch of technology. My heart is obviously still in analytics and big data and now artificial intelligence. But I also do innovation evangelism around Internet of Things, cloud, and other themes that SAP relates to as a whole.

Wow. So, I mean, Business Objects is a company and I suppose a product that a lot of people these days working in, say, big data and cloud and all that probably wouldn't have heard of. But it was massively innovative, wasn't it, at the time with the kind of the idea of the semantic model and so on. I mean, that set the scene for a lot of the kind of things I suppose we take for granted in BI now and analytics and you've talked about. We absolutely were created with the notion that end users should be able to access their information themselves without IT. Remember, at that time, the principal competition were products like Oracle Browser, which was sort of drag and drop SQL. In fact, SQL itself was sold as an end user tool at the time because it was English language like, you know, select star from customers. But we worked with an independent developer, came up with this notion of a semantic layer, a way of translating business terms into the SQL you needed to get the data out of the database. And we saw explosive growth for many, many years. And we've been around a long time, but I have to point out, we're still absolutely the market leader in business intelligence and analytics by a long, long margin.

Despite Gartner having redefined the magic quadrant, focusing on what they call modern BI or self-service BI, if you look at the market numbers overall uh sap business objects is still the by far the number one yeah i mean it's it's interesting i wrote a blog post recently on uh on looker and look at their innovation uh in this kind of new market is to add a semantic model to uh to bi and for a lot of people this is a new it's actually a new concept but it was it was so revolutionary wasn't it and um every I suppose there were companies after that that kind of copied the idea I think there were lawsuits and so on but that was fundamentally that idea that we could take sort of data and make it more understandable to business users and present it in a way that allowed them to kind of make decisions in that way I mean it's such a central

part isn't it but it is I guess it's it's the importance seems to have been less significant people now what do you think well the the big difference now is obviously there was two big differences one is the technology has come a long long way it because of things like in memory uh you can actually interact with the data in very different ways than you could 20 years ago but the fundamental business need hasn't changed at all. What people are trying to do is very much the same. I laugh at some of the marketing that I see because it's almost like they assume that we weren't trying to get people to access data. Of course we were. It's just the technology limitations at the start at the time made it particularly difficult. Now it's much easier. The other big change is that before

there really was only corporate data. If it wasn't in the corporate database, it really wasn't that important. What happened now is there is an explosion of data. And in fact, there's a slight majority of the data that business people use now comes from outside the corporate systems. So it's social data or third party data that they want to combine outside the corporate systems. So it's social data or third-party data that they want to combine with the corporate data. That in turn has helped lead to the new explosion of this need for self-service. Not self-service in the old sense, which was helping people inside the company access company data autonomously. Now it's accessing inside and outside the organization and that has meant some big culture changes necessary and not all organizations have yet

coped with that change no exactly so so we'll get on to that topic as we go on actually that's kind of interesting and but so SAP so SAP when they bought business objects what were they what were they looking for really and what do what they've done with business objects since then and and where where's that where's SAP kind of business intelligence now at a high level before we get into some of the details later on so SAP has uh for a long time sold data warehousing as an integral part of their business applications with the BW business warehouse um but it was really focused on that very centralized, trusted information, notably finance, where it was all about gathering the information, make sure that it was all trusted and calculated correctly. Very German. It's the right data.

You can trust this. And BW is a fantastic data warehouse. It can be a little bit slow to update, but in terms of the power of what you can do in multiple currencies and creating hierarchies, there's still nothing equivalent on the market. But the front end tools were really focused on those financial users. They actually liked tools like Business Explorer, BEX, but the average person who just wanted a little bit of data to do their jobs didn't really like the BEX interfaces. And that profile of users was becoming more and more

important. So SAP invested, I think it was around $7 billion in acquiring business objects to help in that market. And that was a revolutionary thing at the time. I remember that consolidation happened around that time was kind of very interesting. And yeah, I mean, it's good to talk a little bit about what SAP are doing later on in the conversation in terms of HANA and that sort of thing. And I guess where business objects is now within that. But one of the things I want to go through first of all was you've written some superb blog posts in the last kind of few years and as well as the cartoons as well actually I always think they're quite good and there's one in there that I was coupling there actually I thought be worth talking about in this episode

because I thought they kind of got to the heart really I suppose of what it is that you do and you're famous for and I suppose the drivers of analytics and the value from it and benefits from it for customers. And there's one that started off with a blog post you wrote called analytics is the most important business process in your organization. So do you want to just kind of recap what that is a little bit at the start, really, and we can talk about kind of some of the things in there. So this was, it's a blog post that was prompted, as most of my blog posts are, by my conversations with people using analytics tools. So our customers or I'm very lucky I get to go to around 40 conferences a year talking about technology and analytics. So I get to talk to people implementing the projects and find out where they are, what they care about. And one of the things that I noticed is that analytics is

changing in two very different ways. So the first is this, the core traditional home of analytics, which is about self-service access to information. And the big trend there obviously is modern BI, making it easier and more interactive to get data. But there's an entire other side to analytics that the more traditional analytics people don't always take account of. And so I start seeing cultural problems inside organizations. So this is the whole notion of digital transformation, where people are trying to change the way they do business, create new products and services, engage their customers in new ways with Internet of Things, cloud, mobile, and so on. But the heart of those new business models really is data and analytics. In many ways, it's the same types of technology, but it's used

in a very different way. Instead of only thinking about analytics as something you do after you've actually run your business, you know, I'm running my business, and I'm going to analyze it and try and improve it. That's never going to go away. That's always going to be important. But now analytics really fundamentally is inside the business model itself. When you as a customer choose to go to a particular supplier, information applied at particular touch points is now part of your product experience. So analytics is going out of the back office and becoming part of that front office experience. It's going away from sort of cutting costs and finding out future opportunities and really being an operational growth engine right now. And what I find is that as I talk to organizations, there's entirely different teams that are using technologies to do that aspect of it.

Often they don't even work for IT organization. They report up to business units. They're trying to find new ways of engaging the customers. Now they're doing an awful lot of analytics. And in some cases, they're sort of reinventing the wheel and creating new platforms, running into all the same problems that we've already seen on the analytics side. And so if anybody is out there, I urge them to go and find out who's doing that inside their organization, because I do find there's a lot of gaps between the two. Yeah, I mean, there was a similar topic we talked about with Paul Sondreger from Oracle back at the start of the kind of show back in last year with the idea of data capital that you would have. You know, there are companies now that drive everything they do from intelligence and the data you gather is a form of capital as well.

And the gist of it really was that, I guess, as kind of business is changing and as we're going into the Internet and kind of so on, that there are a whole new breed of businesses coming out there that are based on data and analytics. I guess the question to you is, is it possible for a traditional company to reinvent itself to do this? Or is it new companies and new organizations that are taking this approach you're saying? No, absolutely. Every company that I talk to, including massive companies, are absolutely investing in these new approaches to using data on the front end. They're very aware of what needs to be done.

There isn't a big organization that I talk to that doesn't have some kind of digital team to try and figure out what they could do differently. And actually at SAP, we've been working with all of our customers using methodologies like design thinking, where it's a methodology where you bring in the IT people, some customer-facing people, some finance people, and you sit down and try and rethink your product and service from the customer point of view. It's basically a structured form of brainstorming that allows you to get to very concrete prototypes and things to test out. And again, analytics and data is always very central to that new style of thinking. But they don't always invite the traditional analytics teams. At the same time, the analytics teams are sitting there. Some of them are doing things the same old way.

Gartner quips that they are B-inosaurs. You know, the comet is streaking through the sky, and they haven't quite realized that they might be obsolete. So, and again, actually, and you see, it's actually related to technology in some ways because you see these new teams rushing to try to do things with, say, Hadoop or Spark. And again, sometimes those teams are run by very different people that are running

the traditional data warehouse. And yet there's huge overlaps in what they're trying to do. Yeah, but it's interesting, isn't it? I think that you have this danger, I suppose, of, I suppose, the innovators in organizations doing it without IT, without kind of the analytics team and so on, and it going off kind of slightly wrong. But you've also got the issue, I suppose, with IT that struggles sometimes, isn't it, to kind of get business cases for this and does struggle in a way to kind of that you know I think IT is very good at coming up with a kind of a with a kind of platform and architecture but do you find in reality that it's more the business that comes up with those or can how can IT I suppose in a way participate in that and help really? Well for the last 26 years i've been urging uh analytics people to get closer

to the business to really understand the business issue all too often unfortunately i go and talk to an analytics team and i say so what are people what are you doing with uh analytics internally they'll say oh well they're doing sales analysis it's like oh great okay so you know what benefit have they got out of that what what are they doing with it? And the answer is sales analysis. They don't really know because they don't feel like it's their job to know. And yet I feel that it's incredibly important if you don't know what people are doing with the data, that you're in no position to help them take things to the next level um so yeah i mean so so i mean you you talked in in that blog post you talked about there's a quote there i made note of actually that i thought was quite good you said but organizations are increasingly reason that realizing that digital transformation doesn't just require new processes it requires a new approach to creating and implementing business processes and you talked about creating on the fly by analytics and so on but what does that

practically mean i mean that that sounds a bit of a kind of truism and certainly we'd like to do that. So in your experience with the tools you've got and the techniques you use, how does that actually kind of come to fruition then? So this is the big change between the old style of applications and the newer style applications.

So older style applications are fairly linear or they're defined points where you might branch off, but you can really sit down and draw the end-to-end chain. A good example is typically how people have thought about customer interactions. You know, there's this notion of a sales funnel or a pipeline and customers would, you know, you'd get some interest from customers that you'd engage them and then they might do some sort of trial and then purchase and you know it was nice and linear and people would try to measure that linear process but the reality is that these days customers get to choose their own adventure if you like you remember those books when you were a kid where you say if you're on page two you could choose

whether you want to go into the scary cave or go into the wood. And that at each step, you would actually determine your own path. So that's what customers do now when they engaged with any vendor. They can go to the website. They can go to the store. And companies have to cope with all of those different touch points. The actual customer journey is going to be completely personalized. It's going to be different for every person. What companies have to do is manage all of those different processes as optimally as possible and really do analytics at each point. predictive analytics going there right now what kind of ad should I show what are the products I should be showing to this customer to make them interested after they've purchased what can I do

to help them recommend my product to other people so at every touch point you're really applying analytics in real time and it's not just customer interactions you could say the same thing for logistics or any many other traditional business processes. So, I mean, one of the things, one of the trends that I'm spotting is that I suppose when we first started doing this, analytics and BI was typically a standalone tool. It was a standalone process, a context switch and so on. And we're increasingly seeing these kind of things being embedded into business processes and the applications and so on to the point where you can imagine in the end what the i suppose the end goal of this is that analytics is not something that you go and do it's just part of the process really are you finding that's happening really or is that more just kind of just uh wishful thinking again this is a great example of the gap between

the business need and what we've been able to do with the technology because the business need has always been for an application i just want to do my job and workflow so as part of my job i need some information there's a decision as part of that um it's always been ideally it would just pop up as part of the workflow the problem has always been that we haven been that we weren't able to connect the operational processes with the analytic processes because the technology was very different. You had to have an ERP for your operational stuff. And if you tried doing too much analytics as part of that, then it would just bog the whole system down. So you had to take the data out and do data warehousing and so on. But that's always been an artifact.

That's never been something driven by the business. It's always been something that we would have liked to have done in a very integrated way if we could have done. Now, with new technologies like in-memory or massively parallel big data, Hadoop and Spark, you can do that. You still need data warehousing, but it's not the separate environment from a platform point of view as it used to be. So SAP, obviously, we have our SAP HANA product. It was designed from scratch to combine OLTP with OLAP, operational with analytics, so that you can have that sort of live business process with the analytics happening at the same time as the operational steps. So analytics will never go away. But today, a lot of people whose job isn't analysis have to do analytics when, in theory, they could just do it as part of a workflow.

Again, the reality is that there are so many different types of jobs and things that people are trying to do that you'll never have sort of cookie cutter prepared business applications for every type of use. But the vast majority of people probably do just want a quote unquote interactive report or a dashboard rather than sitting down and and analyzing the data themselves from scratch so you in your job title it's kind of innovation analyst innovation evangelist sorry and uh and you mentioned about iot and cloud and so on are you finding are you seeing any examples yet of using things like internet of things things like data lakes and so on in a kind of concrete way to push forward analytics in companies and drive value? Is it happening now from what you can see?

Oh, absolutely. Yes. Many, many examples. In fact, there isn't a company that isn't doing something in this space. It's biggest in industry. I mean, they've had connected machines gathering data for decades now. But this ability to do more with that data faster means that you can do things like predictive analytics.

We worked with Trenitalia, for example, to make sure that their trains don't break down in the middle of the Tuscany countryside, stranding hundreds of people. So now there's an alert, you know, your wheel bearing is about to overheat so that they can pull into a station in Bologna and get people onto a new train,

go and fix the train while improving their customer service. So this is one of many, many examples. And in fact, when it comes to our internet of things in particular, there's two things that I like about it. One is it's revealing formerly hidden processes so a lot of what was going on in in life or in companies was uh not just not being recorded so you couldn't see it because of the

increase of cheap sensors you can now actually see what's going on. So you can track it and optimizing it. My favorite example there is I used to read books on paper, believe it or not. Now I have Amazon Kindle and Amazon actually keeps track of your reading process. So they can tell you how many minutes you have left in the chapter or the book. But so they're actually recording something that was previously invisible, and they can then use that data to start making recommendations about other books. Potentially, they could even go back to authors and say, look, sorry about this, but everybody's giving up on page 200. You know, next time, maybe you need to write your plot differently.

But that same kind of concept is being applied to everything else. The other aspect that's interesting is that the Internet of Things means that you can sell metered services. Instead of selling products, you can sell things on a subscription basis. And there's been an explosion of that. Exactly. So another post you wrote recently, which i think was very topical and timely was when uh the gartner kind of conference happened and you i think you posted something

the vast majority of analytics is still reporting and that's not so bad and i thought that was really good because there is so much um i suppose publicity and thought around things like data discovery and visual analysis and so on but the bulk bulk of things you see, the bulk of things people want are relatively simple and simple kind of like, I suppose, you know, discovery analytics and so on. I mean, again, what was the driver for that? And what was the kind of thinking behind that post really? Well, I think it's interesting

because I always get slightly annoyed at the standard business BI maturity models that you see that has, you know, know old boring reporting in the lower left and then um sort of um prescriptive predictive and what to call the last one again i don't know i forgot yeah i know you mean though it's it's very kind of uh the implication this is stuff there's no value to it but but for most people they just want to see some numbers on the screen and they want to understand you know what it means and so on, really. Well, plus there's much more secular than you realize when you look at that model. Because the reality is that you have basic reports that

tell you things. But then over time, the analytics gets more sophisticated, typically through more self-service, traditional BI. But then people go, oh, well, now that we know how to do that, let's pre-calculate it and put it in the report or in the dashboard. You see what I mean? So the first iteration of the report might just show, you know, here's what you've sold. The next iteration of the report might say, oh, but now I'm going to rank it by here are the areas that you should look at in order to optimize sales. But it's still a quote unquote report, but the level of analytics in it is a huge leap ahead. So there's this constant cycle of we provide sort of semi-interactive interfaces for business people. They learn how to use that. They then go to the next level of sophistication, typically by using more interactive

analytics. But then once we've mastered that, it goes back into the pre-prepared dashboard report. So we're constantly improving the level of analytics that we do. So reporting is never going to go away. It's the ideal, as Don McCormack has pointed out several times, that the weather report is a fantastic example. There is an immense amount of big data sophistication that goes into creating those weather models. They have some of the biggest supercomputers in the world. But the end result is a little sunshine symbol or a little rain symbol and that's all we need exactly exactly just just because it's reporting doesn't mean it's

not sophisticated no i mean the way i think of it typically is it you know it's the classic maturity curve or it's the classic sort of series of steps on a ladder really and i guess part of the problem is that in most organizations there still is a kind of a real lack of the basic fundamental understandings of analytics and the different things you can do with it and how they build on there. And to my mind, having basic reporting against what you've done in the past, understanding it to get a kind of a solid picture of where you are is important. But then looking forward, I mean, I think there are things, I guess the probably problem with that kind of reporting and where it sort of falls down is in how do we then make decisions off of that? And how do we kind of automate and speed that up as well? And something that I mean, we've got, I think very soon on the show, we've got BeyondCore coming on, Arjit from BeyondCore, talking about the work they're doing around trying to, in his words, kind of like close that comprehension gap with i suppose in a way taking

the kind of reports you get in a tool like business objects or oracle tools but automating the way it's presented to people and surfacing kind of insights and decision points off of that is that something you've looked at at all or had some thoughts about yeah absolutely again i think that people don't realize just how far we've come because this notion of taking analytics and then automating it has happened time and time again. It's just that when we do it, as soon as it becomes automated, we don't think of it as analytics anymore. So at some point, somebody used to choose, I don't know, airline prices, right? The price per seat. Then it became automated and something else started doing it. Now it's probably some sort of machine learning that's trying to do it automatically.

So the cycle has always been happening. It's just the instant it's automated, we don't care about it so much anymore. So there's nothing fundamentally new in the cycle. What is new, of course, is that artificial intelligence looks like it can bring us a huge step forward in what we can automate. But the fundamental trend has been there for a long, long time. People just want it directly in the application. If they can do it without having to think about it, that's ideal. That's great. So, yeah, yeah, exactly.

I mean, I think it's interesting. I think typically there's a lower understanding of analytics than you'd expect in organizations i think there's a kind of not an education thing there but i think it's about the the kind of the stepping stones and building blocks and so on but do you find i guess with the change in the way budgets are held now in bi and the fact that kind of business people now control a lot more of it that's had lots of good things happen because of it um but but you feel that things like the the value of semantic models and data quality and so on is is being less appreciated now so i think there's a whole bunch of things that are going on that are making analytics more important than ever i mean it's

been the number one in technology for about 20 years, off and on. There's a couple of years it wasn't, but it's now even bigger than ever. And partly, I think there's a whole lot of things behind that. One of them is a generation change. I think today's CIOs and CEOs, we're now starting to see the first real generation that grew up with a PC at home when they were kids. So that took information, some sort of information access for granted from their earliest childhood. And that I think is making a difference to how people

run companies. Second, there is just a vast amount of data now available. Before you might have wanted to do analytics, but you just couldn't get hold of the data. Now you can. The problem of bringing that data together and governing it and making sure that it's trusted, that is just as hard as it's ever been. In fact, probably harder. There is maybe a scope for artificial intelligence to help there, but it's still 90 to 95 percent of the effort of any analytics project is getting the data together in the first place to start doing the rest and that's really hard i mean that data governance data provenance you know data quality they're all topics that are hard and i think that one of the kind of the criticisms i think really of i suppose moving to kind of self-service tools and that sort of thing is that these things tend to get pushed to one side. And I think particularly within the big data world as well,

it's got a massive kind of buy so far from that kind of thing because it hasn't had to deal with that. But you wrote a good blog post at the end of the year, which was, I think, a few people passed around, which was your predictions of Cynics' Guide to BI in 2017. And I thought it was really good. And you talked there about kind of, you mentioned Timo's first law of BI as your first one what's that then just tell us that and the thinking

behind that bit so the first law of BI is that executives will always be dissatisfied with their BI systems so just realize that why it's because you imagine you're an executive your job is essentially making decisions as your BI system becomes more sophisticated, gives you just the data you need to make decisions, it becomes automatic. You're not going to sit there and go, okay, great. I don't need to work now anymore. What happens is that businesses move on to the next level of sophistication. Think about supply chains, right? There was no earthly way we could manage today's supply chains except through the very complex systems that we have depending on, you know, so Cindy Housen does regular measures. Howard Dresner also does some regular numbers.

Now, it's always surprised me, really, because what are they measuring? They're actually asking people, was your project successful? We know that the vast majority of projects didn't really have any criteria set in advance to determine whether they would be successful or not. So it's really just this aggregation of subjective views by people who weren't really getting the business benefits. So at one level, it's a really bad example of how not to use analytics. Yeah, exactly. I mean, you mentioned, I mean, I want to get onto the future stuff in a second, but you mentioned about BICCs as well. um you know business intelligence competency centers I guess there's lots of ways we can describe you know I suppose in a way a business trying to come bring these things together but to your mind what's the what what's a good

example really of organizations you know organizing where they do analytics you know internally to get the best benefit out of that you mentioned things you've done with sap but what's good examples you see out of that moment well the the best the best ones do less interesting enough so the the traditional old school bi competency centers are the ones that create the reports or the the dashboards and you you file in a request for a something and they prioritize it with a council and then they go and build it. That is exactly the kind of old style business intelligence that people have rebelled against because it just isn't enough. The modern world requires lots and lots of interactive types of analytics. Again, going back to digital transformation, a lot of what people are trying to do, engage with customers in new ways,

that's an experiment. They need to be able to adapt it really fast. They need to say, oh, that didn't work. I now need to change what information I'm giving to the customers. And that's just a very different cycle than traditional IT projects. It's very far from ERP, we're going to put in a supply chain system and then upgrade it every five years. Analytics has to be really interactive, even more so now that it's part of the product interface. And so the older style BI competency centers didn't really react fast enough. The right way to do it is for the group to be more of what Gartner calls an analytic center of competence, or a kind of like BI

community of competence, where the BI competency center experts concentrate on best practice. They don't fly the planes. They concentrate on training the pilots, running the airport, and manning air traffic control. So they ensure that people have the skills for the rest of the organization to get the throughput they need. Interesting. So let's look forward to, I guess, what's coming in the future and also what SAP are doing in this area. So first of all, just set out for us, for people that knew BusinessObjects and they know HANA, for example, what does the kind of the SAP product lineup look like at the moment in BI? And what's coming along down the line really that could be interesting for people in the future?

Okay, so we have the traditional sort of business intelligence that has steadily gotten better over the years. So web intelligence is still a fantastic product, still one of the top products sold in the market. It's great for interactive reporting. In many ways, it's like the Swiss army knife of business intelligence. It's robust. You can do just about anything you want to do with it. And we continue investing in that platform. Every version gets more and more interactive, more powerful. And we continue investing in that platform. Every version gets more and more interactive, more powerful. And we fully intend to keep with that. As we said, reporting is going on. It's not represented so much in the latest, what Gartner calls the BI magic quadrant, but it's still absolutely a huge company need. And actually, when you go to the seminars where they talk about

modern BI, they always run this question saying, oh, that old stuff, are you still going to invest in that? And every time the survey comes back saying, oh, yes, we're going to roll that out to more users and more content. And they always act a bit surprised. They shouldn't do because information is more important than ever. So traditional reporting, interactive reporting, absolutely still very important. We've got some great products there then we've invested heavily in the newer modern bi self-service so that's cp lumira we've got lumira 2.0 coming out in a couple of months and the big step we've taken there is that we're actually combining the self-service aspects of data you can grab data from multiple sources

mash it together correct correct data problems, create visualizations, and share them. But we've now done it so that it's also tightly linked to a designer environment because we've realized that there's a BI lifecycle that typically isn't taken account of. It's a bit like driving a car. There's first gear, which is maybe Excel. Second gear is a self-service tool like Lumira. Third gear might be some sort of dashboard or traditional reporting. Way up to fifth gear, which is where that information is fundamentally part of your data warehouse and your operational environment. The problem up until now is that it's a car without a clutch. Every time you want to change gear, there's a horrible grinding sound because to move from easier. So from there, if you can take something,

if it proves useful, you can then easily extend it and turn it into a full analytics application. Notably, you can add buttons to actually take actions. So rather than having to, you know, oh, there's your analysis thing. Oh, now I have to go back to this other system to do something. You can actually just add actions in that make sense based on the analysis you've just done. We're investing heavily in BusinessObjects Cloud. So we've actually taken the opportunity to rethink BI because there's all these different engines. You've got reporting, visualization, budgeting and planning, predictive. You can buy those tools separately right now and install them inside your organization,

but that's a lot of work. Because of the cloud, we can just provide those engines in the cloud, and we've rethought the user interface so that you just go into the cloud and you get seamlessly reporting and predictive or budgeting and reporting and predictive just end to end in a very new way. And then we've put our digital boardroom on top of that. And that's really the, you know, the three screens that allow you to get the high level view of your business, the top KPIs, and drill all the way down to the smallest operational metric. The sort of thing that executives have dreamed about for at least 50

years is now becoming a reality. Okay, brilliant. Well, so AI then, I mean, predictive analytics, AI and so on, you talked about it earlier on, where, I guess, where is SAP taking this? And where do you think AI with kind of analytics will take us in the future? All right, so quick definitions, because otherwise people get confused as heck. So this is the way I'm going to use the terms. If you disagree with them, I don't care. AI is a sociotechnical construct. In other words, it's like the marketing term. It's used generally to mean things that computers can do now that were typically restricted to humans in the past. Then machine learning is the word we use for any type of algorithm that learns, that improves itself based on the training data. That includes

traditional statistical approaches. So we've been selling machine learning with traditional statistics for a long, long time with the acquisition of KXEN many years ago and applying that to fraud, to predictive maintenance, to next best action and consumers. So that's been going on for a while. It's exploding because a lot more data is available. And then the exciting part is the newer forms of machine learning, including things like deep learning, where you're using more of a neural network approach for things like image recognition, natural language processing, and so on. So that's the technology. But the bottom line is, this allows us to rethink a huge amount of business technology from scratch,

really. Almost every business application is, answering a series of questions, whether it's fly chain or HR. And we can now, if we take AI for granted and we know that this exists, we can start really rethinking the workflows in a huge number of areas. So we're using it pretty much everywhere we can, everywhere from natural language interfaces. So we have SAP Copilot so that you can use a chatbot approach to your corporate applications. So, you know, Siri, book me a week of vacation. Siri, can you print me my latest job slip?

Or Siri, what budget do I have left? It won't be Siri, it'll be Copilot. But you get the same things we're seeing in the consumer world. Why not use it to make using enterprise software that much easier? But the really big opportunity is automating complex decisions. So automating some of knowledge work, the most boring, repetitive parts of knowledge workers' jobs, we can replace with AI. So we're in the process of launching SAP CLEA, C-L-E-A, and that is a series of applications devoted to particular areas where we know that we can help companies save a lot of money while improving customer service. So things like, you know, matching invoices to payments, right? It's just the sort of, if it all matches perfectly today, it's great. It's done automatically. But if there's a slight discrepancy, the payment comes in, doesn't quite

match the invoices, then there's some poor person in a, typically in a shared service center somewhere like Prague, whose job it is to wade through all of those different documents and receipts and trying to figure out what's going on. It's a great example of where we have a huge high quality training data set and a very specific decision to be made that fits in very nicely in a business process. And so it's really a no-brainer to start providing that kind of application. We're also doing it for HR to help people shortlist resumes for customer service, using the data from the customer service request to automatically send it to the right place in the organization to get the quickest response, and many, many other areas. And what I find particularly interesting about this is that it's about analytics, but it's one of the areas where the mega vendors who sometimes are accused of not being innovative,

this is an area where we actually have a huge advantage because AI is not about the technology so much as it is about having the data. And for most organizations, there's no better, higher quality training data set available than in their business applications like SAP. So this brings, you know, we have this huge advantage when it comes to automating any kind of analytics decision within a business because we have the training data and we can actually apply it directly into a business process. It's been great speaking to you. And it's nice to speak to someone that you've read the articles about for so many years and you've got such a good insight into this whole kind of industry and so on, really. So, Timo, thank you very much. Where can people see you next? I mean, you mentioned the website there. Are you speaking at any events in the near future?

Yes, I can. 40 events a year or something. Yes. What's the next one to talk about, really? I'll be in – where am I going? I'm going to Austria next week for an SAP forum. Very nice. And then we have the big SAP conference,

Sapphire Now, where we have, I don't know, something like 20,000 people descending on Orlando. I'll be talking there about the big trends in BI and analytics transformation. I'd love to hear from you there. And honestly, the blog post

is where I put everything that I'm interested in. And I'd love to hear, talk to anybody who's interested in the same things that I am. And it's been a pleasure talking to you. I've had a lot of fun. Thank you, Timo. Thank you very much. And have a nice day. Thank you. Thank you.