When Jim Goodnight, founder and CEO of SAS Institute isn’t running a multi-national, $1 billion-plus software company, he’s likely out somewhere being caustic, bashing Microsoft or Oracle or Cognos — or suffering through another press
Yet he’s extremely accessible, makes numerous appearances before customers, including the SUGI international user group meeting for SAS users in Seattle two weeks ago, where he talked about a new release of the company’s business intelligence software.
Earlier this week, Goodnight spoke to IT Business Group editorial director Martin Slofstra. Find out about his latest pet project and how he likes to spend his Saturday mornings.
ITBUSINESS: Some would say the real challenge is to bring business intelligence reporting tools to the masses. How are you planning to do that?
Jim Goodnight: What most vendors are calling business intelligence are simply SQL queries to a database where they allow you to create a query very easily. You do that, it goes to the database and the results come back from the query. That’s what companies like Cognos and Business Objects are calling business intelligence. It’s really very simple, very low-end, very easy-to-use tools for large numbers of users.
We have always concentrated on the more heavy-duty analytical capabilities appealing to the large number of technical people in an organization. But the mass of low level or business user that didn’t really need all that power was going somewhere else for ease of use. That’s one of the things we are correcting. The 9.1 release will be giving much better query and reporting tools at the low level. (Ed. Note: SAS 9.1 begins shipping to customers in a phased release beginning Aug. 31).
ITB: So there will be a two-pronged approach, one for technical users and one for executives and business users?
JG: That’s exactly what we are talking about. We never paid much attention to the low end of the market and companies like Cognos and Business Objects have. We realize we’ve given a lot of market share there and we are going to try to take some of it back.
ITB: What about Microsoft? They’ve announced their intention to release a beta test version of its SQL server business intelligence platform with reporting capabilities later this year.
JG: Microsoft, at this point, is fairly low on our competitor list.
ITB: Well it’s an intention, they don’t have a product yet.
JG: Yes, and they’ve also announced their intention to take over all of the four to five million small businesses in the country (U.S.) as well. They’ve announced a lot of stuff they are going to take over.
ITB: Still, if BI is to get bigger, and if you want to do enterprise-wide reporting on a real-time basis, there are a lot of underlying architecture and integration issues that you also need to address.
JG: The only difficulty in doing any of this is really getting the data together in all the disparate sources it is located. That’s most companies’ major problem. When we go to some of the banks and financial institutions, it takes a long time to convince them the hardest part of the problem is going to be getting the data out of your different silos. That’s always one of the biggest problems, the inability to pull the data together.
ITB: You’re looking at some new areas as well, risk management for example?
JG: We’re doing a lot of work in market risk. We also have a couple of installations where we are working on operational risk, one in Belgium and one in Poland, and we are working against the clock for Basel II. (Ed. Note: Legislation planned for 2007 that will force banks and financial institutions to meet more stringent reporting requirements for exposure to risk especially when it comes to loans and extending credit.) That is an area where we are beefing up on staff. One of the things that seems to be in more demand is credit risk, loans that a bank may have, we need to do some simulation and see what really is at risk there.
ITB: There’s a lot of number crunching and back-end processing involved.
JG: SAS is the world’s best number-cruncher. That’s what we do. We’re heavy into analytic capabilities and crunching numbers. Take anti-money laundering (Ed. Note: SAS unveiled anti-money laundering software in October 2002). Working with one of the bank’s data, we were processing about 20 million records in about a minute and a half. This is the kind of performance we are getting.
ITB: About a year ago, you also released a text mining product for analyzing e-mails and documents. How’s that been going?
JG: It’s an area that a lot of people believe you point this engine at a bunch of text and it comes back and tells you all sorts of things about it. That’s really not what it’s about. By and large, we are taking all the “”key word”” frequencies and doing a mapping so we can forecast. Normally a person would have to go through and score a document on how important is, maybe on a scale of one to a hundred. They go through and score maybe several hundred documents, and then we build a model so that we can accurately predict that score. Before, this was all being done manually.
ITB: Can you give an example?
JG: We had one situation of a company getting large numbers of complaint letters. They began scanning these in and then they would score it, so they would be able to know, out of hundreds of thousands of incoming documents, what needed to be looked into. Or, if you are looking at a news wire, and you have a model that can score it, this stuff is important. We’re using frequency counts of words to help us become the predictor of outcome.
ITB: Being able to predict is driving almost everything you do?
JG: Yes, we’ve got some new modules called high-performance forecasting that we are able to look at the time trends over the past and do extremely accurate forecasts. Take a store that’s got a couple hundred thousands of SKUs (stock-keeping units) and they want to know or forecast the sales volume of each SKU at hundreds of different locations. We can produce forecasts in a matter of minutes. We’re working with a number of companies to be able to forecast demand for the various products they produce and then move to supply chain optimization, that is, move all that (forecast) information into the supply chain.
ITB: You want to be able to analyse against everything, whether it be ERP, supply chain or CRM. So there isn’t any type of data you are not looking at?
JG: That’s exactly right.
ITB: Except that the data is not clean, that data is all over the place, and while every CIO may want an enterprise view, isn’t it an almost impossible goal to achieve?
JG: Clearly one of the things you want to do is work on the quality of the data. We purchased a company several years ago that cleans up data warehouses and data tables. For example, a person’s name is represented across the entire application one way, so we can take all the occurrences and variants of a name and an address, and map them into a single name and address for a particular person. This turns out to be very important when we do anti-money laundering, to be able to make sure we’re not dealing with some of the bad guys that are on our insurer’s watch list.
ITB: Still, the industry success rate is not very good. Look at the latest, for example, one research firm (Cutter Consortium) reported lately that only 27 per cent of organizations are happy with their data warehouses.
JG: A lot of people try to implement data warehouses with a competitive product like Oracle and no wonder it failed. I think SAS has been very successful in the data warehouse business. We tend to suggest building smaller data marts and gradually grow up to an enterprise-wide warehouse. People who start with the idea to build some huge, enormous enterprise wide warehouse most often fail. It basically just changes from underneath them while they are trying to do it.
ITB: The industry is going through a real tough time. What’s your take on when things will come back?
JG Well, I hope it comes back soon, our worldwide revenues are only up 18 per cent….
ITB: So how were you able to buck the trend?
JG: I really think companies like SAS that are in the business intelligence space are going to be the first group of companies to come out of the slump, and that’s because we provide software that helps companies keep their customers. If they need to identify which customers they are likely to lose next, we can do that. If they are doing marketing and need to identify which prospects they ought to include in a campaign, we can do that. It just goes on and on. The type of software we produce can be used for and produce a good return on investment and help that bottom line. This is not the time when people say let’s start a $25 million project and install a new ERP system. That’s not going to happen for quite a while.
ITB: Have you also been able to maintain a similar level of profitability?
JG: We’re down just a little bit but that’s because we chose not to lay off anybody and we chose to continue to hire. We increased our staff last year by eight and a half per cent. Being a private company, we can take a long-term view and we don’t have to worry about quarterly profits. Two years ago, I said, ‘The market is clearly slowing down, there’s a lot of people being laid off, now’s the time to go out and hire good people. There are good sales people out there, there are good R&D people out there, they are the people we are hiring.
ITB: Are you glad that IPO you once planned never went through?
JG: Oh, yeah. We never even got that close. We were still working on internal financial systems to make our auditors think they could audit our books better.
ITB: Do you think that the stigma that SAS is more for the technical user still exists?
JG: Our competitors always paint us in the corner – that we are software for a small group of techies. I’ve heard the CEO of Business Objects stand right in front of me and paint that picture. I got up afterwards, and said, ‘Well, that small group is generating about four times as much revenue as you.’
ITB: So it’s only your competitors who are behind this kind of talk?
ITB: Are you still involved with R&D at SAS?
JG: Not as much as I used to be. The last time I actively programmed was a month ago working on a Saturday morning doing some programming. But I very much miss it.
ITB: What were you working on?
JG: I was working on anti-money laundering.
ITB: So, will that show up in a product?
JG: Yes, it will eventually. You know, it’s almost like working on a puzzle — being alone and using your mind and concentrating on one particular thought process. It’s a great relief and it’s always fun. Like some people like to read a good book, I like to write a good program.
Comment: [email protected]