His name evokes images of an evil genius in a James Bond flick. His carriage and demeanor is that of a gentle, absent-minded academic. Among the world’s leading CEOs James Goodnight is a class unto himself. The 6 ft 5′ former professor of statistics at the North Carolina State University today heads the largest privately held software company in the world with annual revenues of US$ 1.5 billion. With personal assets valued at $3.2 billion, the SAS Institute CEO is on the Forbes list of the world’s richest people. Goodnight spoke with Joaquim P. Menezes, editor of ITWorldCanada.com during the 30th SAS User Group Conference in Philadelphia. In this exclusive interview, the lanky and normally laconic SAS CEO held forth on everything from software innovation to his personal management style. He revealed interesting and little-known facts about his company and himself – why he considers meetings a waste of time, for instance.
Major enterprise software vendors – SAP, Oracle and Siebel (in the CRM space) – say they have built analytics into their applications. Why then is there need for standalone BI tools?
What those vendors are really talking about, perhaps, is executing on a scoring model – where you throw a record at a model, with a bunch of table rows – and it comes up with a new column, a new predicted value, a score. The models themselves are really created offline. So their claims about doing online analytics is really nothing more than a lot of smoke and mirrors. I think people who spend a lot of their time in meetings are just wasting time. They think they are busy, but they’re not really accomplishing any work. Jim Goodnight>Text
As for Siebel, it is a relatively smaller software company. We are around four times larger, with considerably higher software revenue. [Siebel’s] software revenues for this year are estimated at $336 million. Ours will be close to $1.5 billion. We did $1.3 billion in software [sales] last year.
What about Oracle’s new BI initiatives? It has announced a Database with ETL, embedded OLAP, and data mining capabilities.
Didn’t [Oracle CEO] Larry Ellison himself recently say that there isn’t any innovation left in the enterprise software space? When you really consider it, ERP systems are just plumbing. You’ve got to have them to run the place…send out invoices, collect money, pay bills, write and pay cheques, order stuff and so on. Frankly that’s the only reason why these systems are around. But how much innovation can you really get with plumbing. The layer above them is where we belong. We’re not in the online interactive space where transactions have to occur quickly, and you don’t read a lot of data, but just one record. We’re in the business of gobbling up really huge amounts of data and ripping through it…and there are two totally different stories and philosophies behind the two processes. We store data in very large blocks, so in a single disc read, we can process a thousand records. That sort of thing just doesn’t work as well for indexed, relational type database storage.
Predictive analysis that provides foresight, not just hindsight, is a big focus at this SUGI conference. In what verticals has SAS seen some of its greatest successes with this approach?
Telecom is one. Retail is certainly another.
In the retail space we’re working very hard right now on pricing optimization and size optimization. Many large clothing manufacturers ship standard-sized items – two large sweaters, two medium, and two small – to every single store in the country. It turns out that many stores are frequented mainly by small-sized people, and vise-versa. If you can get the right sized sweater to the right store in the right volumes, it means less inventory. So we go in and look at clearance history and use that to cluster stores. Instead of a single view of all stores, we break it up into a dozen different views. For each one of these, we come up with different combinations of items that need to be sent there. We also use a lot of operational research, when we’re programming to determine pricing for sales and clearance.
Have SAS product sales and revenues increased in the compliance space as a result of the stringent requirements of Sarbanes-Oxley and other legislation?
In the long run [the new compliance environment] will certainly help profitability of our products. But that will happen once these products are [rolled out]. In the case of Sarbanes-Oxley, companies are still struggling to understand [how its requirements affect] the process they are using. Just about every one of them is bringing in consulting firms such as Ernst & Young, and Deloitte and Touche to help them document and understand all the processes. SAS Corporate Compliance is used as a depository for all the definitions and procedures used in the organization. We store the sign-offs with digital signatures. SAS is also doing a lot of work in areas such as anti-money laundering and fraud detection. Several Canadian banks have started using our anti-money laundering product.
Around four years ago there was distinct talk about SAS going public. What has happened since then that made you change your mind?
Our accounting firm said [in case we went public] they didn’t want to certify our financial statements unless we had a financial system they were familiar with – such as Oracle. So we had to get Oracle up and running and that was a three-year, extremely difficult process. By the time we finally got Oracle to work (for the most part, at least) the market had cooled down, the people at SAS cooled down to the idea of going public. Also, now with the increasing compliance regulations going public is certainly more trying and difficult.
How would you describe your own management style?
Some have called it management by walking around, I call it management by loitering. I’m personally involved in what goes on at SAS and for that you must have movement. I try to avoid set meetings. A lot of people feel they are working very hard if they are in meetings for every hour of the day. I don’t adhere to that belief at all. I think people who spend a lot of their time in meetings are just wasting time. They think they are busy, but they’re not really accomplishing any work.
I’m personally involved with our development processes at SAS – including software quality control. I make it very clear we will not certify any software until the bugs are out; or at least until our testing groups say they are satisfied. When a release finally goes up, we have a list of known non-critical issues. But no issue that has anything to do with data integrity [remains unresolved]. We typically start reporting bugs in a system when it’s in its final testing phase and fix the bugs in an average of two weeks after they are found. If we have a data integrity issue, everything stops until it gets fixed.
Today it can take as much as $1 billion and 10 years to launch a single new drug. How would SAS Drug Development speed up the process, while reducing the cost?
SAS Drug Development provides a methodology for organizing all the details – data and analysis – surrounding pre-clinical and clinical trials of a drug compound. It also makes it easier for the company to sell the compound to another pharmaceutical [firm]. (The average pharmaceutical compound is sold at least twice before it gets to market). It became clear to some of our marketing folk that this tool would be great for Sarbanes-Oxley as well. So a bunch of [SAS] programmers took it over and added their domain expertise to it, adapting that to Sarbanes-Oxley. Its impact on our revenue will be significant, but not on par with our marketing automation application – which has a much larger audience (telcos, retailers, banks and more). As compared with that Drug Development is used only by pharmaceuticals.