PC WEEK: Data warehousing has been a hot market and is expected to continue growing very quickly. How does NCR view the industry these days?
NYBERG: CRM and data warehousing are our real growth area because the market is exploding. We really saw a pick-up in market activity in the fourth quarter, which I think indicated what we are going to face going forward in the year 2000 -- basically, an explosion in this market space. This market is coming our way.
When I started here four and a half years ago, 1 terabyte was a huge system. Today, 1 terabyte is not a huge system, although it is still a huge system for our competitors because most of them choke at 1 terabyte. Our customers are making plans to start at say 400GB and move to 1 terabyte within six or eight months. We will see tens of terabyte systems this year already. Our biggest customer is already over 100 terabytes.
I saw in your magazine and some others estimates of the sizes of warehouses going forward, and it reminds me of the cell phone business in Europe. Every year the telecoms missed the projected growth rate by a factor of two. They grew two times faster than anyone expected. And I think that with data warehousing it will be the same.
PC WEEK: If the market is coming your way, with explosive growth in data warehouse sizes into many terabytes and perhaps even petabytes, what is NCR doing to maximize that business opportunity?
NYBERG: We have increased our resources in this space. We have organized the Teradata Solutions Group to run this business ... and we have this platform called Teradata. We can argue about the superiority of the Teradata architecture, but we have enough proof with more than a thousand data warehouses out there that this is the best, the most scalable platform.
But we decided two and a half years ago that that's not enough, that you have to provide applications on top of that platform. We have invested a lot of money and I think we have some really cool applications that we have introduced, like Relationship Optimizer, an event-driven application for a bank or an airline or insurance company -- any industry that serves consumers.
The fundamental issue when we talk about relationship technologies is, if you're a bank, it's very difficult to differentiate yourself with a product. If you invent a fantastic savings product, you can bet your competitors will have the same product within 48 hours. But there's something you have as a bank that can't be copied, and that's knowledge about your consumer. You've got to create relationships with your consumers and use data warehouse technology to do it in a way that you can make money at it.
What we have done is have the strongest, most scalable platform and have now added applications like Relationship Optimizer and Value Analyzer to be able to understand profitability by individual customers. We've been very successful in the telecommunications industry, where I think we have seven of the nine biggest telecoms using Teradata.
So platform, applications and the last one, which is sometimes underestimated, and that is professional services. We have literally thousands of people in the field who have done this many times before. I read in some magazines that some banks maybe haven't been so happy with data warehouses and they blamed it on inexperienced people. We've been at this business now for eight to 10 years. Those three components are absolutely crucial to success.
Our biggest issue is there aren't enough people knowing about us. In a sense we are to blame for that, and that's why this year we're going to step up our marketing, advertising and other stuff quite dramatically compared to earlier years.
PC WEEK: How has it happened that NCR is not as well known as it could be or should be?
NYBERG: I think it's based on the history of the company. With the announcement we made in October, we focused the company back on the three main businesses. Data warehousing is the most important for the future of NCR, the other ones are ATMs and store automation. We are going to be in businesses where we are either number one, two or three or can become one, two or three very quickly. In ATMs, we are number one in the world, in store automation we are number two, and in data warehousing we are number three. The question is, where are you going to be in 2002, because if you have an explosion [in warehousing], staying at three is OK for me as long as you stay at three in the coming four years.
PC WEEK: One high-growth area in the data warehousing market is so-called clickstream data analysis, what some people call Webhousing. We've heard from Microsoft, Oracle and others on that. What about NCR?
NYBERG: The market is coming our way. The customer who has a 100 terabyte system from us is storing only transactions. If you're an e-commerce company, you're going to store all the clickstreams. So the amount of that data is quadrupling or 10 times higher. That's my argument why the market is coming our way, and certainly the e-commerce boys, if they want to do warehousing in a serious way, absolutely need a platform that scales the way Teradata scales.
PC WEEK: Are we going to hear from NCR on clickstream data analysis in the near future?
HURD: I think you'll see a couple of things. You'll see us with Web analytic capabilities that will be integrated into Teradata. We look it as two opportunities -- one, there's an opportunity within our existing customer base and the traditional industries that we think about, and secondly it opens up a whole new target market at the same time. The e-target market, the e-plumbers and the infrastructure people, are now beginning to realize the power of the data that they have in their hands. As you know, there are billions of clicks going on every day that are left unanalyzed. We view this as a tremendous market opportunity for us.
PC WEEK: At the same time, time may be wasting right here. Your competitors are out with offerings and marketing messages already. How swiftly do you need to turn this around to maximize your potential there?
HURD: We have a company doing an analysis of the data warehouse market within the e-commerce market and found basically no installations of significance have occurred. But I think over the next 12 months, the market is going to go crazy. We are all over it right now. We're trying to call on every single Web company that we can, we're beefing up our offering, and you'll see both those coming together in Q1. You should have seen us announce in December an outsourcing capability, and really that has tremendous leverage in the e-space. You have a lot of companies that are undercapitalized and don't have the opportunity to install an in-house system.
PC WEEK: What are you doing for the mid- and small market? You've ported Teradata to NT -- how has that been received, and do you plan to go to any of the Unix flavors other than MP-RAS?
HURD: We're on NT and NT has gone well. We have installed not only many systems with NT, but have been able to scale it. We announced in October our first multi-node capability with NT. We're up to a four-node environment; you should expect to see an eight-node en-vironment in Q1. We're working with Microsoft to try to show scalability of NT.
The market you call small and mid- is growing real fast, and what we thought of as small and mid- two years ago is half a terabyte today. So it's really important for us to get NT scaling, and we're working real hard with Microsoft on that. There's no question we'd like more support from Microsoft as it relates to NT scalability and the perception of NT as an enterprise thing.
We are moving to Solaris 8 and that will be available at 64-bit time. We're working closely with Sun on Solaris on Intel, and that works proceeds. There is no other plan of record for any other Unix flavor at this time. What we want to be on is the market share leader. From an operating system perspective, we think we've chosen those leaders with Solaris and NT.
PC WEEK: With the port to NT, has your target market customer changed?
NYBERG: I think it's changed a bit, but, more importantly, where the ... focus on warehouses has the main driver of growth is, I think, the bigger issue. We are going down lower in the size of companies in the industries we are addressing. I think the combination of having NT and the market coming our way makes it easy for us to do that.
PC WEEK: Lars, some time ago you sounded something of an alarm regarding privacy of consumer data and announced an initiative to have certain capabilities in Teradata that would enable users to take steps to protect the privacy of the data. How has that message been greeted by your customers?
NYBERG: I think it's been received very well. Some customers were as aware of it as we were, and I think they've received it very well. There's a debate going on in society around this issue. I raised this issue because I felt this will be an issue in the future. Anyone who uses data warehouses has to be absolutely proactive in this space and also very clear about what the rules are in their company for the usage of data. There were some customers that hadn't thought about it. Some were surprised, and I think once they got the message and understood it, they were very appreciative of us taking the initiative.
PC WEEK: I assume you follow the activities of your competitors at the high end -- Oracle and IBM and, down a bit, Microsoft. What have you made of their marketing messages in the past six months?
NYBERG: I think their conclusion is the same as ours -- data warehousing (and CRM, which is only a part of the warehouse space) is one of the most attractive segments of the IT industry. I think some of my competitors are scrambling to create an image that they can deliver solutions in this new space, because most of them, if not all of them, are having scalability challenges.
It's an issue of architecture -- Teradata was developed for the sole purpose of being a warehouse database. That's why we have an MPP [multiple parallel processing] architecture. If you have an SMP [symmetric multiprocessing] architecture, you're going to have problems. It's not only the size of the warehouse, it's also a question of what kind of queries are you making, how many users do you have, how complex are the queries.
PC WEEK: When do you think the world is going to see a petabyte-size warehouse?
NYBERG: Well, in 1996, we showed in Tokyo up and running 11 terabytes. It was the biggest warehouse ever shown. Some journalist wrote, "Who in the heck needs 11 terabytes?" As I said, our biggest customer is now running more than 100 terabytes. If I say six years from now, it is going to be three -- it comes back to the cell phone estimate. I'm very honest about it, I just don't think we have a clue how quickly it is going to grow -- when you add clickstream, when you add multimedia, 50 terabytes is going to look tiny in a couple of years.
PC WEEK: What kind of ROI are customers deriving from CRM implementations? Is it worth their while, or are the dividends perhaps overhyped?
NYBERG: Fundamentally, it's easier to quantify ROI on applications other than CRM. Let's take "churn," for instance. The telecommunications company can tell you exactly how much churn costs them, and if you have churn in the neighborhood of 25 to 30 percent, it costs you an arm and a leg. The way we have been successful in that industry is to say, what if you could reduced churn by 5 percentage points, how many millions is that worth? And that is counted in tens or hundreds of millions of dollars. In relationship to that savings, the investment in the platform and professional services is a very good investment.
When you come to CRM, you say, if I can create the relationship, can I do cross-selling, can I increase loyalty -- you have to make a few more assumptions before you can determine what the income side of the equation is, and that's why I think you can get suspicious that some people might have hyped it.
The customers that I have talked to doing CRM, in banking and other industries, in my opinion haven't hyped it at all.
