Data Radicals logo

The Beginning of Business Intelligence

Bernard Liautaud, Managing Partner, Balderton Capital

Bernard Liautaud

Bernard Liautaud founded BusinessObjects in 1990, where he was CEO until 2005 and chairman until 2008, The company became a business intelligence leader and one of the world’s 15 largest software companies. He joined the venture capital firm Balderton Capital as a partner in 2008 and was named managing partner in 2016.

Bernard Liautaud

Bernard Liautaud

Founder, Business Objects & Managing Partner

Balderton Capital

Satyen Sangani

As the Co-founder and CEO of Alation, Satyen lives his passion of empowering a curious and rational world by fundamentally improving the way data consumers, creators, and stewards find, understand, and trust data. Industry insiders call him a visionary entrepreneur. Those who meet him call him warm and down-to-earth. His kids call him “Dad.”

Satyen Sangani

Satyen Sangani

CEO & Co-Founder

Alation

Satyen Sangani: (00:05) In the late 1980s, Bernard Liautaud was working for Oracle in Paris. He noticed that many of his customers were struggling with getting data out of databases. And he wondered if it could be easier. Understanding all that data required people who could tell a story. It was as if the database fragmented the real world. And then analysts were the people that put Humpty Dumpty back together again to retell the story that created the information.

Satyen Sangani: (00:30) Newcomers had to track down the people who could tell them that story. And all of this took a ton of time and a ton of effort. So, it got Bernard to thinking, what if you could enrich data with language that non-technical people could understand? What if you could turn random fragments of data into Lego blocks that people could piece back together again? His imagination catalyzed the genesis of BusinessObjects. A company that you may have heard of, but perhaps know little about.

Satyen Sangani: (00:58) BusinessObjects was a pioneer — arguably the category creator in a foundational market, business intelligence. Founded in 1990, BusinessObjects took four years to become the first European software company to IPO on the Nasdaq. And 17 years later, BusinessObjects ultimately sold to SAP for nearly $7 billion to become one of the largest acquisitions in software at the time.

Satyen Sangani: (01:21) Today, Bernard is managing partner at Balderton Capital, one of the leading early stage venture capital firms in Europe. He’s also a member of the supervisory board of SAP. He has received a number of distinctions, including the knight of the Légion d’Honneur, in 2007 in France. Bernard is our guest today, and he’s going to talk to us about data — past, present, and future.

Producer: (01:49) Welcome to Data Radicals, a show about the people who use data to see things that nobody else can. This episode features an interview with Bernard Liautaud, Managing Partner at Balderton Capital. In this episode, he and Satyen discuss the origins of business intelligence, Bernard’s experiences as a founder and creating healthy cultures.

Producer: (02:08) This podcast is brought to you by Alation. Alation enables people to find, understand, trust, govern, and use data with confidence. More than 25 percent of all Fortune 100 companies use Alation to support data-driven decision making. Organizations like Cisco, Pfizer, and US Foods drive data culture and innovation with Alation. Learn more about Alation at A-L-A-T-I-O-N.com.


The Founding Story of BusinessObjects

Satyen Sangani: (02:33) Bernard, you founded BusinessObjects. Tell us a little bit about that story, and tell us the founding story: What caused you to do it, how you got your start, and why you did it?

Bernard Liautaud: (02:45) It starts in 1990. That’s when we started BusinessObjects. But prior to that, I was at Oracle. So, I started my career at Oracle back in ’86 when Oracle was a fairly small software company doing about $100 million of revenue. And it started its French operation in Paris. And I joined at that moment. That was my very first job. I joined when I was 23. And Oracle grew at that time in France from about 10 people, which was the original group, to about 500 people in three years.

Bernard Liautaud: (03:18) And it gave me sort of a desire to at some point start a software company. But I was just still searching for the idea. But what I saw at Oracle was that all these corporations were putting in place databases and the message that Oracle was, “Oh, it’s very easy to access data. You just use SQL.” And SQL is a super easy language for any business user. But the reality was very, very different. The most important thing was that the database structure was very complicated for people.

Bernard Liautaud: (03:48) And so, I always thought this is not an easy thing for people to do. And at the same time, I met an independent developer in Paris who came to Oracle with an idea and a product — actually, a prototype of a product that facilitated the writing of SQL queries. And he said, “Hey, look at my product. What do you think? Would Oracle be interested in it?” The product in itself was interesting, but was very thin and consisted of a thin layer to just select the table or select the column.

Bernard Liautaud: (04:20) So, it was not that interesting, but I worked with him for about six months and we built basically a much more elaborate product that used a concept of a semantic layer on top of the database. I met this other gentleman, Denis, who was my co-founder, who was in sales at Oracle. And we both decided, “Okay, we need to get this product, start a company and we’ll see what happens.” So, we left Oracle. We negotiated with the developer so we could acquire the intellectual property from him and then we went off and started BusinessObjects.

Satyen Sangani: (04:59) Before BusinessObjects came along, working with data required specific programming skills. Bernard broke down how the company created a revolutionary way to consume and model data.

Bernard Liautaud: (05:10) The idea was that BusinessObjects was enabling business users to access data very easily. And that was something that didn’t really exist at the time, because also the concept that we had was very new. This concept of manipulating, not just entities in a database, but manipulating terms of your business vocabulary, hence the name BusinessObjects.

Bernard Liautaud: (05:33) So, if you’re in sales, your vocabulary are customers, prospects, revenue, potential, etc. If you’re in HR, it’s basically employees, it’s seniority, it’s salaries, comp, etc. And if you’re in finance, it’s a whole different kind of vocabulary. So, we enable people to create these vocabularies in a corporation, enabling people to just assemble queries by just putting these objects next to each other. And so, that concept was quite new — actually completely new because we had invented it from nothing. And then I remember we did the very, very first sale with a customer in France.

Bernard Liautaud: (06:16) And this particular customer had picked Sybase, which was a competitor to Oracle at that time. And the customer said, “Sybase is better than Oracle. It goes faster, uses less memory. I want to use this software.” And then we worked with a salesperson at Oracle and said, “Hey, why don’t you try to present our product on top of Oracle. It only works on top of Oracle. And maybe you have a chance to win it.” And then the customer agreed to let us do that single demo. We did it, the customer said, “This is exactly what I want. I want BusinessObjects.

Bernard Liautaud: (06:53) “So, whatever database, I don’t care actually. What I want is the end-user interface.” And so, that worked. And then Oracle won the deal over Sybase. And then afterwards, obviously, we told that story thousands of times to every single salesperson at Oracle saying, “Hey, we’ll work with you. We won’t work with any of your competitors. Get us in. It’s a win-win.” And that’s how we managed to get the initial traction.

Satyen Sangani: (07:20) If you look back at the story … you created a category. You did this in a world where nobody even knew what a semantic layer was. People barely knew SQL at the time. Certainly, it wasn’t a world-dominant skill. And here you are creating this idea of this thing that people needed. And then they only realized they needed it because they saw it. Did you think you were creating a category at the time?

Bernard Liautaud: (07:45) No. I think, to me, partly maybe because I’m not a coder myself. I’m an engineer, but I always like the concept and I like easy things. I always look for convenience. And I had to deal with SQL because I started as a presale. So, I did demos and stuff. And I was trying to convince the customers that things were easy. But I recognize they were not that easy. So, to me, it was like, I was really searching for a way to make a query much easier.

Bernard Liautaud: (08:19) And it may be a little accidental, but I realized because having done so many demos and so many little presales engagement, that the main thing was really always … the struggle was always in understanding the data for an end user. It was not that much a language. The language you can learn. Okay: Select from etc., where…? But it’s always like, where is the data? What does it mean? What is joint structure? Which joint should I use?

Bernard Liautaud: (08:49) So, I was thinking, is there a way to sort of eliminate all this and really be in the language of the user? The concept of semantic layer sort of came afterwards. It’s like you basically abstract a concept from a couple of feature points that you developed. Which was, “Hey this thing called sales. So, revenue, which is a sum of price times quantity. If we could encapsulate that into something. And then we realized that we can reuse it. The same formula, we can reuse it in different places. So, it’s more like, can we take little pieces of SQL statements and reuse it, combine it on the fly?

Bernard Liautaud: (09:32) We sort of named it afterwards. We said, “Okay. Well, this is… We’re going to call these things ‘objects.’ We’re going to give them names. And we’re going to put them into a little container that we call a universe, because it’s for the universe of that particular kind of user.” And then that was it. Then afterwards say, “Hey, how do we call this thing?”

Bernard Liautaud: (09:53) Well, over time, we sort of thought about calling it a semantic layer, but there was no idea, no conception of a category at the very beginning. It was just like trying to solve a problem, which I certainly was encountering all the time. And then I think you discover, or maybe it’s different in the other case, but certainly in our case, we discover how big this issue was as we went along.

Satyen Sangani: (10:19) Yeah. I guess it would be 30 years later when I’d be pitching Alation and people would ask me, “What is it?” And I’d say, “Well, do you know what a semantic layer is in the BusinessObjects universe? Well, we’re trying to make that its own software product.” Of course, you did that 30 years earlier. So, shame on me.

Bernard Liautaud: (10:41) But what’s interesting is that the concept that we created at the beginning, a lot of people thought it was a bad idea because all our competitors — so people who were trying to create convenient query builders — they were saying, “Well, BusinessObjects is bad, because with BusinessObjects, you first have to create these things, these universes. And so, you have to do some work, whereas we can work straight out of the box.”

Bernard Liautaud: (11:10) But “straight out of the box” meant that you had to deal with a database structure and so on. Whereas we made that abstraction, but there was an upfront work. At the beginning it was a fight to convince the world that this concept of semantic layer was a good one. But that initial work — the upfront work of building this universe of objects — was gamechanging in terms of the ease of use of the product for the consumers afterwards, for the end users.

Satyen Sangani: (11:39) Yeah. And relative to what was available at the time, probably opened up the addressable market for who could use data significantly. So, tell me about the journey of the company itself. Started in 1990, you finally exited in 2006? Is that right? Or 2016?

Bernard Liautaud: (12:02) 2008

Satyen Sangani: (12:04) ’08. I’m sorry. I apologize.

Bernard Liautaud: (12:04) 2008

Satyen Sangani: (12:04) So, that’s a 28-year journey. I think my math is okay there.

Bernard Liautaud: (12:11) No, sorry. Let me just put it back.We actually started in 1990 and then we finished in 2008.

Satyen Sangani: (12:18) Sorry. 18-year old journey. So, my math is not that good.

Bernard Liautaud: (12:21) Yeah.


The Evolution of BusinessObjects

Satyen Sangani: (12:21) I should not be running this company. So, what were the stages of the company? If you look back on the story, what was the evolution? So, you had two guys and the proverbial garage at the very beginning. What were the shift changes?

Bernard Liautaud: (12:41) So, there were several phases. The first phase were startup phases. So, startup phase, like two guys in a little business center outside of Paris. And it’s basically convincing the first customers with this product and really making the product real. And also we had to build an R&D team because we just had this freelance developer at the beginning.

Bernard Liautaud: (13:04) We started getting the first customers in France and then we realized, “Okay, we need to develop professionally.” And this is when we’re going to say, “We’re all going to try the venture capital route.” And I had some acquaintances that were in venture and some in California. Some of them being at the origin of Oracle, they were business angels in Oracle. And I presented. And one of them was Don Lucas. Don Lucas, who is … It’s funny to …

Bernard Liautaud: (13:33) I was just watching the Theranos [documentary on] Netflix the other day. And Don Lucas is very much prominent there, but he was one of the first guys with other folks, Arnold Silverman and a few others who basically believed in the idea. And so, they were business angels. And so, we managed to get them to invest a little bit of money. And then we had two venture capitalists in France put in some money. But the first round was basically a million dollars and six months after we started the company.

Bernard Liautaud: (14:03) And then with that, we were able to start, and we had the idea that if we want to succeed, we need to succeed on a global scale. So, we started the business in August 1990 and in September ’91, we opened our U.S. office and our U.K. office in parallel. So, basically very, very quickly, we were just barely 10 people in France. We started our U.S. operations and it actually clicked quite well.

Bernard Liautaud: (14:33) I hired a few sales people, relocated myself in the U.S. and then started the U.S. business from the ground up. We didn’t hire a big GM of U.S. operations. We just hired sales people, presales people. So, we took the best Oracle person in New York, the best Formic person in Chicago, the best Oracle person in Dallas, and the best in San Francisco and we started like this. And that worked out quite well. We did $200K the first six months. Then we did $1.5 million in ’91, then we did $5 million the year after. Then we did $15 [million].

Satyen Sangani: (15:14) At the time of the IPO, things were going incredibly well. They were a stock market darling projected to grow at over 100 percent, but as with all great adventures, things would take a turn. Bernard called it a near-death crisis.

Bernard Liautaud: (15:32) Basically ’94, ’95, beginning of ’96, everything is going great. We’re doubling every year. We’re going from $15 [million] to $30 [million] to $60 [million]. The profit increases, we’re viewed as a bit of a darling on the stock market. We went public at originally at a $125 million valuation. That was our IPO price. And a year and a half later, we were worth a billion. So, we thought we were on top of the world at that time. And then that’s when difficulties began to happen. So, that’s phase three, which is basically the crisis of BusinessObjects, where we experience basically a near-death crisis, where several things happened at the same time.

Bernard Liautaud: (16:15) We felt that the product had to be rebuilt because A, the technology was a bit old. Second, we were confronted with a completely different competitor, which was S-Space, with a product that I’m sure, Satyen, you remember, which is all OLAP-based. And it was this concept of online analytical processing, which was coming on the market and everybody wanted that. It was fast, it was multidimensional. And we didn’t have that. And we just felt like, “Okay we need to completely change our powers so that we can compete.”

Bernard Liautaud: (16:47) So, we wanted to build a new version, which we did, but we had all sorts of issues with it — which I’m sure we can go into the details — but basically the product. Our new product, everybody wanted it, but it didn’t work. Didn’t work on the old architecture of Microsoft. And the launch was a complete disaster. In addition, we had an issue with a contract in Germany that we thought was a good contract, but ended up being a bad contract that had to back out of our accounts.

Bernard Liautaud: (17:20) And when you’re a public company, this kind of thing is lethal. So, we had a terrible time: a combination of sales problems, competition, product issues, and basically everything arrived at the same time. And our stock price was heavily impacted because basically we’re missing our results. So, instead of growing at 100 percent like people were expecting us to, we grew only at 40 percent.

Bernard Liautaud: (17:45) And therefore, our stock price … well, our market gap went from a billion to a hundred million. So, we lost 90 percent of our value in a period of three or four months. So, that was a crisis, a really dark moment for the company.

Satyen Sangani: (17:58) What year was that happening?

Bernard Liautaud: (18:00) That was in ’96. So, we were about six years into the company, two years into the public life of the company. And everything is going the wrong way. At the same time, my co-founder Denis decides that he wants to leave the business. He was interested in the first phase of the buildup, the IPO for him; sort of the story was less interesting afterwards, but I was really interested in continuing the story. And basically I’m by myself now.

Bernard Liautaud: (18:31) I had to rebuild the company. I was based in Paris. The entire management team was based in Paris and I have to engineer a turnaround. And that’s going to be from many different angles. I’m going to decide that, first of all, we’re going to have to rebuild the product, fix it, completely change the product development process. I’m going to change the center of gravity of the company, move from France to California to be closer to our core customers, our core partners, but also to the financial community, because we were public on Wall Street.

Bernard Liautaud: (19:06) We’re not public in France. Rebuild the management team because that move and all the troubles basically meant that half of the management team was going to leave and it started from there. And that is basically this new phase of the company from the end of ’96 to ’99, where it’s basically the turnaround where we were going to, again, rebuild the existing product, build also a new product on the web, because it’s the very, very beginning of the internet and web-based products.

Bernard Liautaud: (19:40) And we’re going to come out of this brilliantly. Fortunately, it was through innovation, like obviously fixing our old product, but also innovating through a brand new internet product and becoming again what the customers wanted.


The Basic Elements of the Product

Satyen Sangani: (19:57) At the time by this phase was the product still the semantic layer with the business modeling? And it sounds like you then added that same OLAP capability or at least something approximating it underneath the hood. Were those the two basic elements of the product?

Bernard Liautaud: (20:17) Yes, we came out with a brand new innovation, which was really interesting. Which was the concept of an integrated product that did OLAP and query and reporting in one engine. So, basically the idea is that you query the data always through the semantic layer, but the semantic layer is intelligent enough to design that they should bring back more data from the database than you actually need.

Bernard Liautaud: (20:45) Let’s say you want revenue by customer and you just put that query. Well, actually, the query is also going to retrieve the granular data by year and by product. So, if you want to drill into the data à la OLAP, like do multidimensional analysis — so, that’s drill into revenue, not just by customer, but also by customer by year or by customer by year and by product — you can do that without querying back the database. You work now in a little mini-cube. So, we created the concept of micro-cube, which was like an OLAP cube but was downloaded onto your desktop.

Bernard Liautaud: (21:22) So, that combination of query, and OLAP, and reporting became… So, once the product, something-version four, which had real trouble getting launched because it was complicated and it was buggy and so on, but ended up being an enormous success afterward on the market. And then on the side, we created a version of that called Web Intelligence, which was always based on the same semantic layer, but doing a very thin, very easy-to-use query and analysis product just from a browser.

Bernard Liautaud: (21:58) And the combination of these two products really enabled us to grow substantially. So, we went from 60 million [dollars] to 85, which was a year where people were expecting us to do 120 or something. But we went from 85 to 120 to 210. And so, our growth started to pick up again and then we became profitable again. So, we didn’t lose money through this period. We went from 17–18 percent operating margin down to zero then down to then back to 15 percent and 20 percent.

Bernard Liautaud: (22:35) So, finally, from a performance standpoint it wasn’t bad. The lowest point of growth was 40 percent. The lowest point of operating margin was break-even, but the market was expecting so much that it went to 100 million, but from 100 million, we went back up to, I think about 3 or 4 billion in a period of a few years. And we became, for two years in a row, one of the top 10 tech stocks on Wall Street. So, the turnaround was superb and really enabled us to be back again, the number-one business intelligence product on the market. So, it was basically a really important phase of the company.

Satyen Sangani: (23:23) It’s funny, because I hear your story and there’s so… And I don’t know if it’s just because all startups are all similar, but there are so many different parallels. We both started as product marketing managers at Oracle. The revenue trajectory is actually quite similar in many ways. And so, it’s just an interesting thing to watch this and then even some of the re-architecture. So, I guess, were you maybe from a product perspective, you guys had the semantic layer, you had the visualization layer, you had the OLAP-processing layer. All of which now are like, there’s probably 20 companies doing each of those things on some level at this moment. Did you foresee this next-generation … as you left, did you sort of think, “Oh, there’s going to be this flourishing number of companies in this space”? Or how did you think about the world evolving?

Bernard Liautaud: (24:11) For me because we’re now at generation three or four of BI products. First generation was BusinessObjects and Cognos and S-Space, and MicroStrategy. And then next generation was Qlik and Tableau. And then now, it’s Looker, and a few others, ThoughtSpot, and a bunch of other companies, but each generation replaced the prior one with the exact same message. It’s basically, “Oh, the other product before, for the IT they’re complicated. Nobody can use them. And what you want is to really liberate the users so that they can have access to information at their fingertips when they want blah, blah, blah, very easy.”

Bernard Liautaud: (24:51) That was exactly the message that we started with. So, it’s funny how it reinvents itself from a product and architecture standpoint, but each generation is still trying to solve the same issue, which is how can you let business people have access to information in real-time to make decisions based on data and not on gut feel?

Satyen Sangani: (25:11) It’s funny because as I see it, we’re solving some incredible problems on the compute and processing layer. The speed, and the volume, and the interactivity — those problems seem like they’re really orders of magnitude we’re better off now than, of course, we were before. And then on the flip side, a lot of the human-cognition issues: I mean, we’ve seen at least five or seven companies who talk about building a metrics layer, which sounds an awful lot like the semantic layer that you were referring to.

Satyen Sangani: (25:49) And it’s funny, I read this one blog and somebody came out and said, “Well, at Airbnb, we had a single place where we defined all the dimensions and all the metrics, and that was a huge innovation for us.” And I thought to myself, “Wow, the fact that, that’s in 2022 being pitched as the most innovative idea is pretty interesting.” How do we solve that problem? Because that’s certainly the problem that I feel like I’m working on and it’s this, “How do I teach people to talk to databases?” Now, in your lens as a venture capitalist, how do you see innovation occurring in that domain?

Bernard Liautaud: (26:25) I think as this industry develops, you realize that it segments itself. It’s very hard to have one solution for everybody, because you have this concept of also data scientists. You have also business analysts, which didn’t really exist that much at the time. So, because of the culture of data developing in companies, there are a lot more people who are more at ease with using data.

Bernard Liautaud: (26:56) So, in the 90s, people working in businesses were data ignorant, meaning they had no concept of how to deal with data, they didn’t know databases, didn’t know SQL, and certainly didn’t know coding. You have to work with data now, if you want to be good at your job. But some people will work with data in a very structured way and they would just want to have access to data and lots of dynamic capabilities into the data.

Bernard Liautaud: (27:23) Some will want to do exploration. So, you have products that are specialized at “How do you explore data?” Meaning you have no idea what the end goal is. So, that’s a completely different kind of product. And then you have products that are specialized in massive distribution of information, and you have data again that are more like for data scientists. Twenty years ago, that didn’t exist at all, because the culture of data wasn’t quite there. So, I think as we grow, the business grows, but it’s also segments more and more. And I think it’s going to continue to be this way for a while.

Satyen Sangani: (27:59) Yeah, I would agree. I mean, it’s funny because I think a lot of people there’s this new movement around, for example, a modern data stack and some really incredible tools that are coming out. But you can also see in five years there’s going to be a more modern data stack. Because on some level, analytics is really the exercise of accelerating human cognition and sort of abetting human cognition.

Satyen Sangani: (28:25) And I just think that the fundamentally cool thing about human beings, especially educated ones, is that they just have this ability to think about problems in such different totally orthogonal ways and to the extent software can help people do that differently, better against different problems in different types of data. The returns to that are always going to be extraordinary.

Bernard Liautaud: (28:47) Yeah. I think we’ve seen the fact that people are very educated around the data. The data is available in many more ways because the capacity of the data bases to crunch enormous amount of data of at first time is and speedy time is 100x, 1,000x more than what it was at the time. You can also mix and match data, inside data and outside data, which was sort of unthinkable in the 90s.

Bernard Liautaud: (29:17) So, you could only access the data that was there in your corporate servers. Now, you can mix and match data. And then you can use more unstructured information as well. So, I think it’s about to become a lot more sophisticated. I think the industry still has — in my view — still a solid 25 years of great growth and new innovation to get to the holy grail of people being able to use information very easily and take all advantage of it with a mix of AI and exploration and dynamic information. There’s a ton of things that still need to be done.

Satyen Sangani: (29:59) So, maybe switching gears, this is a podcast about data culture, and a lot of the people who are listening, are listening both to understand the dynamics of the software space because they’re consuming a ton of it. And it affects how they think and act organizationally. But you obviously have this unique vantage point because you’ve seen companies sort of get transformed by data. And then you yourself built a company that was a data company, but also that had its own culture.


The Culture of BusinessObjects

Satyen Sangani: (30:26) Tell us about the culture of BusinessObjects. What was that culture? How did it differ from other software companies at the time? Did you even think about culture? Because everybody seems like they’re talking about culture now, but is that something that you considered at the time?

Bernard Liautaud: (30:40) I think we had a good balance between sales and marketing and product. Sales and marketing was a very, very important function at BusinessObjects. A little bit like Oracle, meaning that there was a lot of the company dynamics that was around a culture of great salespeople, high performance in sales and marketing, great branding. And at the same time, the company has been built on a series of innovation, the semantic layer, the great user interface, the beginning of that category.

Bernard Liautaud: (31:19) So, these technologies were really important. And a lot of us came from Oracle. I came from Oracle, my co-founder came from Oracle. So, we had a bit of an Oracle culture to begin with, but I think we gave it more of a human sign. And one thing that really made us different, in my opinion, was the international nature of BusinessObjects. And we were a French company, but we recognized very early on that France is not necessarily — certainly, at that time was not the heart of the software industry.

Bernard Liautaud: (31:53) There were very few software companies coming from France. So, we felt like in order to be a global champion, we had to bring in components from many different places. So, there was never a dominant geographical culture within BusinessObjects. So, there were great ideas coming from France, great ideas coming from the U.S., but also anybody, like there were some elements of innovation that came from Italy that we replicated. Some from the Netherlands, some from Germany, or Japan, or anywhere. And we had businesses in many countries. Our revenue was distributed across many different places.

Bernard Liautaud: (32:34) Very early on, we had a great development center in India, but through Crystal, we have like 1,200 people in Vancouver. We had people in San Jose, a lot of people in Paris, our Italian office, our Germany office. They were vibrant cultures and everybody was bringing their own local culture into the fabric of the company. That to me was something that I’m still very proud of. A lot of people, still today, believe that the best moment of their career was at BusinessObjects. We had a culture that I think a lot of people recognize themselves in, which is a mix of high performance, great innovation, international culture, and a lot of passionate people.

Satyen Sangani: (33:17) You obviously counsel a ton of entrepreneurs. What do you tell them about building culture?

Bernard Liautaud: (33:23) I think I would say, first of all, they need to think about it a lot. They need to decide what are the important elements they want as part of the company culture. And they need to practice it, and practice them all the time and communicate heavily about it. But probably the most important thing is the founder — CEO, leader of the company — needs to exemplify these values in this culture. I, personally, believe that companies’ cultures are modeled after the behavior of the CEO.

Bernard Liautaud: (34:04) The CEO, and therefore the executive team, behave in a particular way that basically creates the environments or the type of behavior that everybody feels that they should sort of mimic or adopt. So, as a CEO, you can’t say, “Well, I’m doing this because I’m the CEO, but you cannot do it because you’re not the CEO.” If the CEO is sort of really rough with his executive team, then the executive will be rough with the level underneath.

Bernard Liautaud: (34:40) And the roughness will be part of the culture. If the CEO is very human and is a listener, and takes input from many different people, that will be part of the culture. So, to me, the number-one element of how you create culture is by setting the example. You have to set the example, you’re going to be modeled after unconsciously or consciously.

Satyen Sangani: (35:13) Is that true at any level of scale? Is there a level of scale where that becomes not true?

Bernard Liautaud: (35:19) When I think about the cultures of the various companies that I have encountered across many years, the past 30 years, I find the Oracle culture is quite modeled after Larry [Ellison]. I find the PeopleSoft culture was very much modeled after its founder. And they’re each very different, but I think at the end, I think it stays there based on what the founder/CEO models.

Satyen Sangani: (35:51) Yeah. And I think, if you are trying to change culture within an organization, it’s a bottoms-up effort, but then there’s also a top-down effort to all of that. So, I guess maybe as we end the conversation, there are two things that I’d ask for you. I think the first would be, we are at that phase — I think where we’re sort of now growing from 700/800 employees to the next level — would love to get your advice to me personally, on how we do that.

Satyen Sangani: (36:20) And I guess, maybe secondarily, would also just love to get your advice to listeners who are trying to do this work: these people who are data professionals, what would you tell them about how to broaden the use of data within their organizations? What to focus on, what to think about, how to build these cultures that they’re trying to proselytize/evangelize/change?

Bernard Liautaud: (36:43) It’s a big question.

Satyen Sangani: (36:44) Well, start with users because they’re the ones who are listening.

Bernard Liautaud: (36:51) I think in the end, people do things when it brings them benefit. And therefore you don’t want to do things just because there’s like a higher purpose. You do things because it facilitates your job. I think it’s all about getting quick wins and it’s not about necessarily implementing a grand vision over time. It’s about the day-to-day benefits. “I need something and I can get that, and it’s easy, and it’s right there when I need it.” The worst is like, “I need to learn this thing and it’s complicated and I still have no benefit from it.

Bernard Liautaud: (37:31) “I’m going to need to go through training and this and that and lots of authorization, and IT, blah, blah, in order to get access to my first piece of relevant insight.” So, everybody wants data, but at the end, it sticks if you can apply it to a business situation.

Satyen Sangani: (37:52) And I guess I’ll just selfishly ask for the advice for me. Not that our listeners — well, they might be curious as flies on the wall. What would you recommend now I guess, 10 years in on the journey?

Bernard Liautaud: (38:04) To you specifically

Satyen Sangani: (38:06) To me specifically. [Laughs.] Yeah, absolutely. It can’t all be for the listeners.

Bernard Liautaud: (38:13) To me, great software companies managed to create alignment. And that alignment is basically communicated by you. And I see if you are able to say, “Okay, this is what we’re doing.” So, you’re creating that magnetic north. Everybody knows where the company is going. You yourself, you set the culture. So, you lead by example.

Bernard Liautaud: (38:39) And you create an environment that enables people to achieve extraordinary things. Then your job becomes a lot easier because you don’t have to tell people what they need to do, you just tell them, just look up. And when you look up, you look at the magnetic north and then you know where you’re going. Most people are going to make decisions without your input, but they just need to know where their company’s going.

Bernard Liautaud: (39:10) And if they know what their job is, where the company’s going, they will make the right decisions. With everybody pushing in one direction, companies are doing amazing things. Most of the companies that struggle is because — it’s not that people are bad. It’s just that they’re going in wrong or diverse directions. So, if you have created that clarity, you make people autonomous enough under that clarity of vision, and you provide the example. I think you’ll continue to achieve extraordinary ambitions.

Satyen Sangani: (39:46) Well, I clearly have my homework cut out for me. Thank you for that.

Bernard Liautaud: (39:54) Easier said than done. But you’re the one who has to do it.

Satyen Sangani: (39:56) That’s right.

Bernard Liautaud: (39:56) I’m done.

Satyen Sangani: (39:59) That’s right. Well, thank you for the assignment, Bernard. Thank you obviously for your time today, but more critically, thank you for the contributions to the space. I think we all stand on the shoulders of those who came here for us. And that is certainly you. So, deep appreciation for everything you’ve done to make our journeys possible.

Bernard Liautaud: (40:20) Well, thank you so much for your kind words. It’s been a pleasure.

Satyen Sangani: (40:27) Mark Twain once said, “History doesn’t repeat itself, but it often rhymes.” In 2012, 22 years after Bernard left Oracle to found BusinessObjects, I left Oracle to found Alation. My observation at the time was that people really couldn’t find, understand, or use all of the different data that was available to them. At the time, it felt like I was doing something unprecedented. Yet over the last 10 years, as I traveled through the path of building a company, I often met people who started their career as at BusinessObjects.

Satyen Sangani: (40:59) People who today are executives at today’s most well-known data companies. We talk about how Alation creates a map for disparate data systems in the same way that BusinessObjects created a semantic layer for databases. BusinessObjects calls that semantic layer “the universe.” Of course today, that universe of data has expanded massively. And the result was that we needed to create a new category of software called data cataloging, which now is one of the pillars of a broader category called data intelligence.

Satyen Sangani: (41:30) So, if history does indeed rhyme, it’s only because we’re able to learn from the patterns of the past and as much as Silicon Valley and the broader world of technology are all about the new, new thing, we can learn a ton by looking back. This is Satyen Sangani, CEO and co-founder of Alation. Thank you for listening.

Producer: (41:51) Alation gives enterprises the tools to make data driven decisions and grow a data culture. Our data catalog can minimize the time workers spend searching for and worrying about the data they need to do their jobs: turning months of frustration into minutes of action. Visit Alation — that’s Alation.com — today.