IBM Gets Down to Business with Its Business Partners
It’s no secret that IBM’s Business Partners (BPs)—the companies that either sell AS/400 applications or sell AS/400 hardware or both—are one of the secrets of the AS/400’s success for the past decade. From the get-go, thousands of companies, some small, some very large, have helped IBM to sell AS/400s to existing midrange customers and also to find new markets for the machines. Like most relationships, the ones that exist between Big Blue and its thousands of BPs are complex. These companies, which now account for more than half of IBM’s AS/400 sales, often have a love-hate relationship with IBM. While they agree that the AS/400 is a great machine, they are rightly concerned that IBM’s priorities are in the wrong order as it tries to position the AS/400 against UNIX at the high end and NT Server at the low end. For years, BPs have been clamoring for IBM to change the way it markets AS/400s, and now, after a year of reorganization of its server development, marketing, and sales teams, IBM is giving BPs change. It is perhaps more change than they had bargained for and, in some cases, more than they can handle.
IBM likes to thump its corporate chest and throw around numbers when it talks about its AS/400 BPs—8,000 BPs selling equipment and over 25,000 applications—but IBM rarely further qualifies those numbers. Part of the reason for IBM’s vagueness when it talks about BPs is that, for competitive reasons, IBM is being inscrutable. It doesn’t want to make life any easier for Hewlett-Packard, Compaq, Sun Microsystems, or Microsoft. The other reason IBM doesn’t get terribly specific about AS/400 BPs is that it doesn’t actually know what many BPs really do for a living and what their plans are for the coming years. The numbers shown in the charts accompanying this story are estimates made by IBM; they are not hard numbers, but rather best guesses.
IBM isn’t necessarily being slack about keeping in touch with these relatively inactive BPs (“inactive” means they don’t have a lot of direct contact with IBM). It’s just
that many of the larger BPs in the AS/400 market, whether they are selling hardware or software, provide technical support and sales services to dozens of their own smaller business partners. These myriad smaller business partners—which might be a one-to-five- person company that might sell a few hundred thousand to a million dollars worth of hardware, software, or services a year—don’t show up on IBM’s radar by design. IBM has set up the AS/400 distribution channel and the application development channel in such a way as to minimize the amount of money and time it needs to devote to sales and applications and to maximize the amount of autonomy and ingenuity in the AS/400 marketplace. This strategy might have been a good idea in the early days of the AS/400 market, but it does not necessarily make a lot of sense 10 years later. That’s one reason IBM is starting to pull back on the reins.
The Nitty-gritty So who exactly are these BPs that IBM relies on so heavily, and what are they up to? According to IBM, there are over 8,000 BPs in the AS/400 marketplace. That number has remained fairly constant over the past few years, but IBM says that the widespread acceptance of the AS/400e models, which support e-commerce, groupware, and modern enterprise resource planning (ERP) suites, will drive that number up by 30 percent this year to over 10,000 BPs. That growth is based on the assumption that IBM’s recent requirements that each and every BP have at least two certified AS/400 experts on staff—one certified for sales, the other for technical support—doesn’t cause massive numbers of AS/400 BPs to seek their fortunes elsewhere. BPs, who were notified of this requirement last May, are supposed to have their people certified on IBM’s AS/400 tests by November 30, 1998. Many are not going to make it, which is why IBM is letting BPs have one person certified for both jobs. But the company has warned BPs that if they want to maintain their BP status—which allows them to sell equipment, get PTF software patches, get technical and sales support from IBM, and learn about IBM’s future plans for the AS/400—they should plan to have two people on their staff certified
during 1999. This kind of certification is the norm in the Novell, Netware, and Microsoft Windows NT Server markets. IBM has concrete reasons for requiring customers to come up to speed on OS/400 V4. A lot of things have changed in 10 years with the AS/400, and the people out there on the front lines of the AS/400 army need to have the weapons to fight against the UNIX and NT advocates. The AS/400 can make a case only as strong as the BP making the sale, and if that partner really knows only OS/400 V2 and RPG, it will not only show, it will hurt a sale.
This kind of thing happens all the time. We know of one instance where a BP sold a small AS/400e machine to a new customer but then configured it all wrong. To save a few bucks, the BP put two
8.5-gigabyte disks, which have the lowest cost per megabyte in the AS/400 disk line, into the server. Of course, online transaction performance is dependent in large part on the number of disk arms in an AS/400; four 4.2-gigabyte drives would have been a better choice. Also, with four drives, the customer could have chosen a disk controller card with RAID 5 data protection rather than have no protection at all. The customer was understandably angry with the BP and IBM for allowing this to happen.
Some BPs have made a big stink about having to be certified, but IBM is right: BPs have to know what they are doing. But it’s probably not the only reason IBM is requiring certification. A large number of AS/400 BPs are living off of their legacies, supporting applications originally written for the S/3, S/36, or S/38 and ported to successive OS/400 versions and releases with little modification. Small BPs may make a decent living supporting their S/3X applications on the AS/400, but they aren’t driving
AS/400 sales; they aren’t bringing home the bacon for IBM. And IBM will use the certification process to get rid of them. IBM is, of course, mum on this subject and doesn’t want to speculate on what effect certification will have on the number of BPs who stay with the AS/400. It is likely that between now and the turn of the century, several thousand BPs will leave the AS/400 fold. So maybe IBM won’t have 10,000 BPs for very long.
IBM says that of the 8,000 BPs it can identify, about a third of them are active, meaning that they take IBM classes or attend the annual BP conferences or rely on IBM for technical support. That makes about 2,600 active BPs and 5,400 inactive ones. About a third of the active AS/400 BPs sell hardware exclusively; another 15 percent sell both AS/400 hardware and applications or systems software. Only about half of these 2,600 partners rely on software sales and ongoing application support to make their living. Among the larger group of inactive AS/400 BPs, the numbers are very different. More than three-quarters of these BPs rely on software sales, and only 14 percent sell hardware; a mere 8 percent sell hardware and software. It is no wonder, then, that IBM is trying to locate and motivate those BPs who know how to sell hardware, which is the only way IBM makes money in the AS/400 game, either directly through a processor sale or indirectly through service contracts. (See Figure 1 for a breakdown of AS/400 BPs by type.)
Since client/server computing took off in the mid-1990s, many of these BPs have been retrofitting their applications to take advantage of this approach to computing. About a third of the AS/400 application portfolio has been reworked for client/server architectures. In the past 18 months, many of the same BPs have been scrambling to make their applications ready to run on the Web and in their customers’ intranets (which link their employees together) and extranets (which link companies to their own business partners). About a quarter of the AS/400 application portfolio has been transformed to use Web technologies; about 14 percent of the 25,000 applications have both client/server and Web support. But a whopping 58 percent of the 25,000 applications that run on the AS/400 are still traditional green-screen applications that were written in RPG and COBOL. (See Figure 2.)
Even so, at this point, even the new client/server and Web-enabled applications for the AS/400 are being deployed with relatively old AS/400 compiler technology. According to IBM, about a quarter of the active AS/400 independent software vendors (ISVs) have expressed an interest in moving to RPG IV and Java for server-side application development over the next few years. These both allow software developers to create programs that are more modular, which means that individual chunks of programs can be reused with little or no modification. While BPs have to learn a new way of programming, this approach has obvious benefits. Another 10 percent of these active ISVs plan to use object-oriented programming (OOP) tools to put a new face on their old RPG programs. But two-thirds of these active ISVs have no plans as yet to move their applications away from RPG OPM (the “Old” Program Model, as the monolithic RPG/400 is sometimes called). And among the 4,400 nonactive
ISVs, only about 5 percent have any plans to use advanced OOP technology to move their applications to the future. It’s no wonder that IBM’s current strategy is to focus on the efforts of its top 150 money-making ISVs until Java and the related San Francisco program template technology matures on the AS/400. Most AS/400 ISVs are too busy supporting existing customers to learn about OOP tools, and many of them—at least hundreds, perhaps thousands—will not make the transition that IBM wants them to. (See Figure 3.)
The Key to Success IBM isn’t just trying to nudge BPs ahead to prove that it is right about RPG IV or Java. There’s a lot of money and midrange market influence at stake in all of this, for both IBM and its BPs. IBM is convinced that the key to success in these emerging markets, as well as the means to expand the AS/400’s presence in North America and Europe, is the expansion of its BP channel, but it has to get that channel in
shipshape condition if its plan is to succeed. Big Blue has always had a relatively thin AS/400 direct sales team—before IBM rejiggered its midrange sales staff last November, only a little over a thousand IBM AS/400 sales reps were in the United States, and probably an equal number in Europe. BPs easily have five to 10 times as many people pushing product, and they do so in all of the toughest, stingiest accounts to boot. In 1994, BPs sold more than half of the AS/400s shipped that year, accounting for exactly half of IBM’s $5.85 billion in AS/400 system sales. Two years later, BPs accounted for about 56 percent of IBM’s $5.4 billion in AS/400 system revenue, and they still made more individual deals than IBM’s own sales force. But IBM’s AS/400 direct catalog and telephone aftermarket sales efforts stole some business back from BPs. According to IBMers in charge of AS/400 marketing, Big Blue wants to expand the AS/400 reseller channel by 1999 so that it will account for 65 percent or more of the $7.5 billion in system sales that we expect IBM to be able to pull in despite the economic woes in the Asia/Pacific region. During 1999, IBM’s own worldwide sales force will bring in only 20 percent of sales while IBM’s combined Web and telesales efforts will bring in perhaps another 15 percent. (See Figure 4.)
The AS/400 has been ahead of the midrange market in general when it comes to driving sales through BPs, but according to Robert Damron, a midrange analyst at Cleary Gull Reiland & McDevitt, Inc. in Milwaukee, Minnesota, the rest of the midrange is starting to catch on to IBM’s AS/400 strategy. Damron says that in 1994, about 20 percent of midrange sales went through the partner channel (the AS/400 was more than twice that level). Last year, says Damron, about 40 percent of the midrange market went through the channel, and he expects it to grow to about 70 percent by 2000. This will put Compaq, Sun Microsystems, HP, and other smaller midrange vendors on par with the AS/400 Division in terms of reliance on BPs. For IBM to get to that $7.5 billion in AS/400 system sales, it is going to have to beat every bush, shake every tree, and look under every rock in the civilized world, and its BPs—many of whom will go bust if they try to stay in the used AS/400 equipment or PC server businesses—are just the people who will help IBM to get there. IBM has a very simple tactic to motivate its BPs. IBM will explain to them that they will make more money selling AS/400s than they will selling PC or UNIX servers; the irony is that it will be the BPs who will have to do all the real fast talking to get customers to buy an AS/400e instead of a PC or UNIX server, but they will probably prove IBM absolutely right. AS/400s may have higher prices than PC or UNIX servers, but they have much fatter margins. After years of razor-thin margins, many server resellers would welcome a machine that gives them some real profits. The AS/400e (with the appropriate deep discounts from IBM and the backing of IBM service, and as the choice of several thousand excellent Web-enabled and client/server applications) may be their salvation as much as BPs will be IBM’s. It all comes down to the sales pitch.
Of course, IBM and its BPs will have to sell a lot more AS/400s for IBM to grow the AS/400 business to $7.5 billion a year. IBM
Server Division General Manager Bill Zeitler believed this time last year that it was possible to double revenue by 2000 to $10 billion, but with Asia and Eastern Europe, the two fastest-growing areas for AS/400 sales, in economic turmoil, this is not likely. Most likely, they need to push over 100,000 boxes plus 17,000 upgrades out the door in 1999 while simultaneously keeping the average selling price from dropping too far below $50,000 for a configured system (it was about $60,000 during 1997) to make that $7.5 billion. To keep that system price high despite aggressive price/performance improvements every 12 to 18 months, IBM and its BPs are going to have to convince customers to use their AS/400s for a whole lot more than bookkeeping.
Inner Ring: 2,600 Active Business Partners Outer Ring: 5,400 Inactive Business Partners Source: IBM estimates
Figure 1: A breakdown of AS/400 BPs by type
Traditional green-screen Client/server enabled Web enabled Client/server and Web enabled
Source: IBM estimates
Sells hardware only Sells software only Sells both hardware and apps
Figure 2: A breakdown of BP applications by type
Inner Ring: 1,150 Active Software BPs Outer Ring: 4,400 Inactive Software BPs Source: IBM estimates
Figure 3: Projections of future BP application development environments
Annual AS/400 Hardware, Software, and Services Sales Other salesWeb salesBPsales IBM direct sales IBM maint & services
Source: IBM, various industry analysts
Moving to ILE RPG or Java Using alternative OOP tools No plans beyond OPM RPG yet
1993 1994 1995 1999
Figure 4: A comparison of AS/400 sales sources over time