Latin American governments are coming around to the fact that innovation is the name of the game ahead, as the time for commodity-driven exports and growth will be limited.
Some countries in the region, as well as some forward-looking companies, have turned to universities to cultivate this innovation.
The 1970s saw the creation of a very innovative company in the US - business analytics firm SAS - based on work and research from North Carolina State University. BNamericas caught up with the company's founder and CEO, Jim Goodnight, to learn more about that history, the company's business in Latin America and where he believes the future is heading.
BNamericas: SAS is known as the biggest privately run business analytics firm in the world. How and why did you decide to go down that road, and was there a temptation to go public at any point?
Goodnight: We started private because we could fund the project ourselves; we had profits in the first years since we had most of the product ready to go to market already. We have been profitable every year for 35 years, and we've had revenue growth for 35 years. Last year we were up 5.2%, and this year we're targeting double-digit growth. We've had two years where we had single-digit growth and we hope to get back to double digits this year as the economy begins to pick up around the world. The weak dollar also helps a lot because everyone's currency converts into more dollars.
Only about 35% of our revenues are US based; the rest comes from around the world with over 400 offices in 52 countries.
Back in the dot-com boom, a number of my senior managers really wanted to go public so that they could become millionaires overnight like everybody else in the country was doing. We talked a little bit about it, but by the time we finished talking, the bubble burst and everybody decided that they didn't want to be public.
Even in recent surveys, 85% of the people at SAS say "please don't go public." I remember at the end of 2008, it was clear that the economy was sinking fast because in November and December of that year, people just stopped buying. I've talked to a number of other CEOs, and they say the exact same thing. At the end of October, year-to-date sales were up 12.5%, and we finished the year up 5%. I remember SAP announced 3,000 would be laid off around the world. There were layoff announcements all over the place.
At the beginning of 2009, I did a worldwide broadcast to assure that no one that worked at SAS would be laid off during the year. We had sufficient money in the bank and didn't have to worry about being profitable because we're a private company and would do what it takes to pay everybody's salary. I did ask everyone to cut back on expenses, which they did, and we ended up with 2.2% revenue growth that year. We didn't lay off anybody and actually made a profit. It worked out well for everybody - people stopped talking about and worrying what would happen if they got laid off and went right back to work. It was a productive year, but we can do that because we're private - I don't have to worry about getting fired if profits go down.
BNamericas: So from the look of things, it's business as usual, no changes there?
Goodnight: Yes, we're continuing to grow. We're the number one analytics company, and I think we will remain so for many years to come. You can ask the analysts about that.
BNamericas: SAS started as a research project at North Carolina State University to help analyze agricultural research data. At what point did this turn from an educational endeavor into an actual company?
Goodnight: It really starts in 1972, when all of the funding for our project was discontinued and we had to raise the money ourselves if we wanted to continue working on this project. So we went to a number of the other universities that had similar analytical groups to ours, and we got all of them to chip in some money. And they wanted us to begin to sell it to outside companies to ease their burden. So we started doing that.
We went on with this mode of paying our own salaries for four years, and by 1976 we had about 120 customers. We went through our very first user group - all our users got together to have a meeting - and after coming back from that we were convinced we had to start our own company. We came to arrangements with the university and moved across the street to set up the business.
We've been growing ever since. There were five of us originally, and now we're close to 12,000.
BNamericas: Jumping from that time to the present day, how would you say the SAS experience has been in the last 35 years?
Goodnight: It's been evolutionary, with changes from year to year to ensure that products stay updated, that we're taking advantage of the new technologies because there've been so many since we first started.
We started just as a mainframe system, and then there was the advent of the Unix [time sharing] system, as well as other things like digital equipment. They had these minicomputers that all of our customers wanted us to run on. Then the PC came in the 1980s, and customers wanted us to run on that. That's when we made the decision to rewrite the entire system in a portable language. And we chose C [programming language] for that, and it was a good choice because C is very prominent right now. So we rewrote most of the language in C, so that we could move it from machine to machine with minimal effort. That portability to run on many different architectures allowed us to grow and expand as the industry did.
BNamericas: Revenues for 2010 were US$2.43bn. What percentage of that would you say comes from Latin America?
Goodnight: It's about 5%. But Latin America grew last year at 20% - it's one of the fastest growing regions, along with the Asia-Pacific region. But this year Asia-Pacific isn't growing because our number one market there was Japan, and it's flat in Japan. No one is thinking of buying a lot of software there right now. I hope that will turn around next year.
BNamericas: What were some of the verticals and/or countries that stand out in the region?
Goodnight: Brazil is big, but Argentina and Chile have done very well. Our number one sector in the region is still banking and finance/insurance - that's about 42% of the business. Bankers and insurers have been using modeling the longest of any of our customers.
Even back to 1990, American Express had over 2,000 models, so they needed a solution to make decisions with. Banks need models to determine who they'll lend money to - what the credit score is, computing value at risk, looking at default and all the loans they have outstanding.
For marketing campaigns, banks need to decide who to target. You don't want to be sending someone the same thing over and over again; you need to be very careful and judicious in who you're targeting. There is also the issue of budgetary constraint in most marketing campaigns, so we also help optimize each individual campaign - whether you want to use digital marketing and email, or the call center, print material or TV ads.
BNamericas: So there's a mix between business analytics and CRM when it comes to banks talking to their customers?
Goodnight: For us, CRM is about 90% business analytics in choosing and optimizing the decisions. The mechanical part in running the campaign is the other 10%.
BNamericas: One of the big up and coming things that consultancies have been talking about for a year or two is cloud computing. You have a US$70mn cloud computing center at headquarters. How would you say that business has developed in Latin America, and where is it going?
Goodnight: Well, let me say one thing about cloud computing - it's not really that new. The concept of renting time on somebody else's machine is called time sharing, and that really started back in the 1960s. And the idea of using a virtual machine was from the 1970s, when IBM had its VM/CMS operating system. So the concept of time sharing - using somebody else's hardware to run jobs on it - is nothing new. I think the only thing new about it this time around is that you're accessing it over the web, so you're using IP to talk to it. The cloud is something new to younger people, not to older people like me, so I don't get really excited about it.
But we do find the virtualization very useful internally, so I think you're going to find a lot more internal clouds than you will external, especially for corporations that have very sensitive data. They don't want to have it on someone else's machine. They want to keep tight control of the data resources.
We do have a number of customers running on our cloud facility, and there is a desire for that as companies don't have to invest in hardware, the infrastructure or the maintenance of the software, and it's an easy decision. Quite often it can be made at the departmental level, rather than waiting on your IT department to get something running.
BNamericas: What percentage of contracts would you say are being offered as Software as a Service?
Goodnight: Right now it's only about 5%.
BNamericas: And do you expect that to expand?
Goodnight: It probably will, a little bit. It's hard to say because we're growing in so many different places. I hope it grows a lot.
BNamericas: What are some of the business analytics trends you see right now, and where do you see it going in the future?
Goodnight: The real growth I see right now is in the high-performance computing area, which is something that we're pretty much doing by ourselves. We're ahead of everyone else on it.
An example of that is a bank that we're doing this [business analytics] for, was previously taking 18 hours, and now they're doing it in 15 minutes. Going from the single processor and executing programs sequentially, we now take a whole array of blade computers - maybe several hundred - and scatter the load over the whole set of them. The blade computers now have multiple cores, so we can get 12 cores, 24 cores that we can actually use on each blade. If you've got 100 times 24, that's 2,400 computer cores that you can turn on all at the same time. We've been working on software to do that in about the last 7-8 years, and we've found another breakthrough that will even speed that up. Our high-performance work is one of the most exciting things to me right now.
We're taking [US retail store] Macy's for example. They have 850 stores, and they have a total of 270mn store-level SKUs. They want to be able to go all the way down to the store level to decide which items to mark down each week. Macy's is a seasonal store, so every Sunday night we begin our markdown optimization run. Our original contract called for us to do it in 30 hours or less. We only take about an hour and a half because we spread it out over these blades. And blades are very inexpensive. Macy's had planned to spend US$3mn to buy hardware to run this thing in 30 hours, and we gave them about a US$500,000 piece of hardware that they can run in 1.5 hours. They liked that deal.
We have a retailer here in Chile that is using markdown optimization in their department stores.
Goodnight holds a doctorate in statistics from North Carolina State University, where he was a faculty member from 1972-76 and where SAS was founded, with Goodnight implementing the features that sat on top of the architecture and expanded the system's capabilities.
Goodnight has been an active speaker and participant at the World Economic Forum. In 2004, Harvard Business School named him one of the "20th Century's Great American Business Leaders" for his three decades of leading a business that has changed the way Americans have lived, worked and interacted.
About the company
SAS once stood for "Statistical Analysis System," having its roots at North Carolina State University as a project to analyze agricultural research. But university funding dried up and demand for such software grew, so the company SAS was founded in 1976 to help customers ranging from pharmaceutical companies and banks to academic and governmental entities.
Today, 35 years later, the company employs nearly 12,000 people worldwide. In 2010 it billed US$2.43bn.
In the US, SAS has been named one of Fortune magazine's "100 Best Companies to Work For" every year since 1998, holding the number one spot on the list in 2010 and 2011.