SSA cornering the market on AS/400 enterprise software. SSA just announced that it is acquiring Infinium (formerly Software 2000), a vendor of financial systems for the AS/400 (IBM iSeries). Infinium's 1,800 customers are located mostly in North America in process manufacturing, gaming, hospitality, and healthcare. Infinium products include modules for human resources, payroll, financial management, CRM, materials management, process manufacturing, and corporate performance management.

This latest announcement follows on SSA GT's acquisition earlier this year of Computer Associates Interbiz group, which was CA's collection of enterprise system acquisitions, such as PRMS, KBM, MK, MANMAN, Warehouse BOSS, CAS, Masterpiece, and Maxcim. Last year, SSA bought MAX International, a small company ERP system.

Most of the systems in SSA's portfolio are AS/400-based, with its original flagship product, BPCS, being one of the major AS/400 ERP systems through the 1990s. SSA's strategy appears to be to acquire as many of the remaining AS/400 packages as possible and continuing to receive the maintenance revenue stream from their installed base clients. (SSA is the only ERP vendor I've seen that actually has a page on its Web site aimed at parties interested in being acquired). A number of SSA's clients are large multi-national companies that SSA would love to keep in the fold. It remains to be seen whether SSA can keep existing clients satisfied to the extent that it is not worth the effort for them to switch.

HIPAA privacy compliance may be taking back seat to EDI. In 2003, medical providers will face HIPAA deadlines for privacy regulations as well as mandated use of EDI for submitting claims for payment. But for those that can't meet both deadlines, the lesser of two evils might be to let privacy compliance slip and be sure to get electronic transactions implemented in time. The risk analysis is simple, according to Jeremy Pierotti, consultant with Partners Healthcare Consulting. "The chances of being inspected by the Centers for Medicare and Medicaid for privacy violations are low, but the chances of not getting paid for submitting a non-standard claim are high," he says. Health Data Management has more.

Manhattan Associates emerging as leader in warehouse management systems

Manhattan Associates just announced its 2002 Q3 results, coming in with net income of $6M, a 30% increase over same quarter last year. More impressively, the profit increase didn't come just by cutting expenses. Revenue rose 8% to nearly $43M. In this market, those are excellent results, and Manhattan has been putting up strong numbers on the board since the late 1990s.

The standalone WMS market is highly fragmented, with hundreds of niche vendors. Manhattan's main competitor has been EXE Technologies, a well-established vendor with whom it was evenly matched a couple of years ago. But that has all changed, as Manhattan Associates has continued to grow through the recession, while EXE has shrunk. Other major WMS players include Catalyst, HK Systems, McHugh, and Optum. But none has been able to turn in results similar to Manhattan's.

Separately, Manhattan last week signed a letter of intent to buy Logistics.com, for about $20M. The addition of Logistics.com transportation management systems and services will extend Manhattan's core WMS products, allowing it to provide a more comprehensive supply chain planning and execution suite of offerings.

Buzzword alert: "rich client"

No, it doesn't mean a client with a lot of money. But first, some background. Until recently, software vendors spoke of client/server presentation schemes as falling into two main categories: "thin client" and "fat client." A thin client architecture is one where business rules and processing logic reside at a server level and the desktop client machine is used only for user presentation. Because a thin client architecture typically makes use of a Web browser, or other general purpose GUI presentation program (e.g. Citrix ICA client), no application program code need be installed or maintained on the client machine. Hence, the client is said to be "thin." Fat client, on the other hand, is an architecture where application programs are installed on each desktop machine.

The advantage of thin client is that it is simple to administer and maintain. But, as anyone who uses browser-based applications knows, thin client systems do have some drawbacks. They are typically slower, especially when deployed over a wide-area network, since even routine data validation requires processing at the server level. Furthermore, they can be cumbersome, since the browser interface (even using DHTML) does not have as many capabilities as a native Windows application. The fat client architecture overcomes these limitations by doing much processing locally, and by taking advantage of the GUI functionality of the desktop operating system (e.g. Windows). On the other hand, a traditional fat client system is difficult to administer because you have to maintain program code on each desktop.

Software developers are looking for an alternative. Hence, the term "rich client." Now it might be tempting to treat the term "rich client" as simply a synonym for fat client and leave it at that. And, I'm afraid that's what many application vendors are going to do--simply rename their old fat client systems as "rich client." But the real concept of a "rich client" goes farther. Rich client, as its proponents maintain, is an attempt to capture the benefits of a thin client architecture without the limitations. How? By using presentation technology that is more feature-rich than a Web browser but still allows the application logic to reside at the server level.

The main contenders for rich client development platforms include Macromedia Flash MX, which takes advantage of the fact that Flash is already installed on the majority of user desktops, Microsoft's .NET Framework, using features such as managed code and Windows Forms, and Sun Java WebStart.

Additional candidates for a rich client development environment include Droplets, an application platform for Java and C++ development, Sash Weblications, an IBM toolkit from Alphaworks, the Altio client applet and presentation server, the Spidertop Bali development tools for Java, and the Curl Surge Lab Integrated Development Environment (IDE).

For more details on the trend toward rich client, see Ted Neward's post in his weblog on the O'Reilly Network.

Update: For a follow up discussion on tools for rich client, see my post on Nov. 17.

Getting tough on ROI. InformationWeek has a good article that explores various methods that companies are using to evaluate the return on investment from IT projects. Among the companies with best practices for managing ROI:

1. Schneider National, Inc., which categories IT projects according to whether they reduce costs, increase revenue, or simplify business processes, allowing management to use a different set of criteria in justifying each category of project.
   
   
2. Schlumberger, which uses worst-case, likely case, and best-case scenarios, allowing management to focus on those factors that reduce the risk of realizing the worst-case.
   
   
3. Vanguard, where executives meet in "sunlight sessions" to debate and challenge the projected benefits of a proposed IT initiative, allowing overly optimistic assumptions to be caught before the project is approved.
   

The article also includes a case study of Citibank, which used a complex valuation simulation tool to trace the potential impact of an executive portal project on Citibank's stock price. Although such an approach has certain appeal, I believe that the complexity far outweighs the benefits for all but the largest companies. In my experience, most companies would benefit simply by doing a reasonable cost/benefit calculation at project initiation, verifying the assumptions, managing risks, and measuring the results after implementation.

Siebel makes strategic bet on Microsoft

CRM vendor Siebel is the latest major player to pick sides in the Web services platform battle between open J2EE standards and Microsoft's .NET Framework. Although Siebel will continue to support and interoperate with J2EE-built applications, it is putting the bulk of its development effort into .NET. In addition, Siebel will optimize its applications for Microsoft server operating systems and its SQL Server database. In return, Microsoft has agreed to make its Biztalk integration server compliant with Siebel's Universal Application Network (UAN). This will allow existing Siebel installs as well as other software that is UAN-compliant to interoperate via Biztalk. The new version of Biztalk is expected to ship in early 2003.

Siebel's move to get close to Microsoft is in contrast with other large enterprise systems vendors, such as SAP, which is a strong supporter of J2EE, and J.D. Edwards (JDE), which is in near total alliance with IBM, another strong supporter of J2EE. Siebel's move is not without risk. Microsoft is rolling out its own small/mid-tier CRM solution. Even though Siebel sells mainly to larger organization, at some point Siebel may find itself competing with Microsoft for CRM deals in mid-size companies.

Furthermore, by identifying with Microsoft, Siebel risks alienating those large company prospects that have standardized on J2EE as well as on non-Microsoft databases and operating systems. Reportedly, the bulk of Siebel customers (65%) run over Oracle databases, with 25% on MS SQL Server, and the remaining 10% on IBM's DB2. How will all those Oracle and DB2 shops feel about Siebel's favoring Microsoft technologies?

CRN has more analysis of the Siebel/Microsoft announcement.

Epicor picks up Clarus e-procurement products for a song

Last week, Clarus announced that it's selling its core products to Epicor for a mere $1 million in cash. Clarus is/was a best-of-breed vendor of indirect e-procurement systems such as applications for expense management, private trading exchange (PTX), reverse auctions, and electronic settlement (electronic bill presentation and payment). Clarus was in the same space as major e-procurement players, such as Ariba and Commerce One, and Clarus had a few big name accounts and partners, such as Microsoft.

Epicor, formed from the merger of Platinum and Dataworks a few years ago, has gotten a reputation on the street for carrying too many products in its portfolio. Most were acquired by Dataworks, prior to the merger with Platinum. These days, Epicor is focusing on actively developing and marketing only a few of them, specifically its "e by Epicor" product line (based on the Platinum products), its Avante, Vantage, and Vista manufacturing systems (based on the Dataworks offerings), and its well-regarded Clientele products for customer service and technical support. At first glance, it would appear that Epicor is continuing its tendency toward product proliferation with its acquisition of the Clarus products. Nevertheless, if Epicor can integrate the Clarus products horizontally across its other offerings, the acquisition will look like a smart move a few years from now. Especially at the fire-sale price it's paying. AMR provides a good analysis of the acquisition.

ERP does improve business performance, if implemented correctly

HBR Working Knowledge has a good interview with Harvard Business School's Mark Cotteleer regarding his survey on the effect of ERP on company performance. In the most recent survey, 86% of IT executives describe their ERP implementations as successful and over 60% report that the benefits exceeded expectations. On the other hand, 14% of implementations are "troubled or abandoned," and 40% either just met or fell short of expectations.

Recognizing, as Cotteleer says, that "successful" and "painless" are not the same thing, what makes some companies successful while others fail? The answer is in how ERP is implemented. Cotteleer goes on to describe some useful maxims for successful implementations, including: "stay the course" (i.e. give users time to get used to the new system before rushing off to resolve non-critical issues); "the devil is in the data" (e.g. "Over the years we have witnessed pitched battles erupt over, for example, how to define units of measure"); and "know the difference between understanding something and liking it."

On that last point, Cotteleer says, "Implementations get bogged down when ... project managers focus on finding a way to make everyone happy. Sometimes that way does not exist. Managers should recognize that and move on when needed."

SAP sales and earnings are ... up

Swimming against the tide of diminishing earnings from enterprise application vendors, SAP this week reported 2002 Q3 net income of $198 million, compared with $36 million for the same period in 2001. Interestingly, while product revenue overall was up 3% to $1.65 billion, revenue from CRM sales was up 19% while SCM was down 3%. SAP's press release has more details.

SAP is clearly benefiting in this weak market from its large installed base and dominant position worldwide to maintain license revenues at the expense of most other ERP vendors such as Peoplesoft, which just reported an 11% drop in Q3 net profit and a 20% drop in license revenue. SAP is also getting some uplift from sales of SCM and CRM products, at the expense of stand-alone supply chain and CRM vendors such as i2 and Siebel. For evidence just look at Siebel's quarterly results this week, where it reported a whopping net loss of $92 million on a 34% drop in license fees.

Buzzword alert: Part 11 compliance

Over the past few years, a number of software vendors selling into the pharmaceutical and medical device industries have been claiming that their systems are "Part 11 compliant." Here's some background. In 1997 the US Food and Drug Administration (FDA) issued its final rule on the use of electronic records and electronic signatures, publishing it in the Federal Register under 21 CFR Part 11. Hence, the term "Part 11." Essentially, Part 11 provides criteria by which companies regulated by FDA can use electronic records and electronic signatures as equivalent to paper records with handwritten signatures in meeting FDA regulations. Furthermore, FDA investigators have started to inspect companies' computer systems for compliance to Part 11. In some cases, companies may find it easier to replace legacy systems than to remediate them. This has created a market opportunity for software vendors serving FDA-regulated industries.

However, some software vendors, hoping to win a piece of this business, make claims about their systems that go too far. Here are a few examples, without naming the vendors: "[Package name] is fully compliant with 21 CFR Part 11." .... "[Vendor name] has developed proprietary software utilizing 128-bit encryption technology that fully complies with 21 CFR Part 11." .... "This solution is 21 CFR Part 11 compliant and will provide an immediate solution to using electronic signatures with minimum investment and minimal impact on legacy systems." .... "[Package name] is 100% compliant with the US Food and Drug Administration (FDA) final ruling on Electronic Records and Electronic Signatures referred to as 21 CFR Part 11." And my personal favorite: "Are you concerned about Title 21 CFR Part 11 FDA regulations governing electronic records and electronic signatures? Don't be. The FDA edition of [package name] is fully compliant."

The basic problem is that these claims imply that the packages themselves are "compliant," whereas FDA regulations and guidance make it clear that it is the end users and their companies that must be compliant. One package may be easier than another to implement in a compliant fashion. But compliance is much more than buying and implementing a certain package. In a recent meeting with one software vendor (the minutes of which are public record), FDA representatives made the following simple and clear comment: "During the meeting we discussed the appropriateness of representing software as 'part 11 compliant.' We explained that the term is a misnomer because people who are subject to part 11 are responsible for compliance with the rule and because achieving compliance involves implementing a collection of administrative, procedural, and technical controls. We suggested that where software has technical features that are required by part 11, it would be appropriate to map those features to particular part 11 controls and then let prospective customers determine for themselves the potential suitability of the software in their own circumstances."

Open source ERP

A couple weeks ago, I asked to hear from anyone who knew of a truly open source ERP system. I didn't get any responses, but I did come across a research note by Paul Hamerman of Giga Information Group on Compiere, which appears to fit the definition. Compiere's license agreement, which is modeled after that of Mozilla and Netscape, provides source code to users at no charge. However, the product is built over the Oracle database, and Compiere is an Oracle database reseller. So the company evidently is giving away the Compiere source code and pulling through Oracle license sales as well as training and implementation. It's an interesting approach, although one would imagine that companies looking for "free software" would be reluctant to turn around and buy Oracle database seats. I also wonder how many small and midsize firms, which would be the natural market for Compiere, would be willing to invest in the requisite effort to keep up on patches and fixes. The whole open source movement, to me, makes more sense for operating systems, tools, and utilities, which can leverage a much larger development community. It would seem that the higher you go up the technology stack toward complex business applications, such as ERP, the more difficult the open source model would be to sustain. I would like to be wrong on this one, but I'm still waiting for an "existence proof."

Wal-mart still pushing its suppliers to Internet EDI

There was more news this week on Walmart's Internet EDI initiative. Both IBM and Sterling Commerce announced that they have been selected by Wal-mart to provide integration and network services for 8,000 of its suppliers as they move to Internet EDI (EDI-INT AS2 standards). According to the IBM press release, IBM was chosen because of its global reach, its experience in the apparel, consumer products, and retail industries, and its implementation services. IBM will also provide VAN services as a backup contingency to EDIINT. In addition, IBM points out that it has built AS2 support into its WebSphere integration products, allowing Walmart suppliers who use Websphere an easy migration path to AS2.

According to the Sterling Commerce press release, Sterling will serve in a similar capacity, giving Walmart and its suppliers a choice of two such providers. Sterling, of course, is a major EDI provider, and one of the companies that helped develop the AS2 specification.

One of the early barriers to Internet-based EDI is that the Internet by itself does not provide the level of data security and reliability that is required for B2B commerce. This is why EDI traditionally uses value added networks (VANs) for transport. The AS2 standard eliminates this barrier by providing, among other things, a reliable and secure Internet messaging protocol, using public key encryption (PKI). When I reported on Wal-mart's EDI AS2 initiative on Sept. 17, the Spectator started to receive a large number of search engine hits for that post, probably from all those Walmart suppliers that are no doubt highly interested in what hoops they will need to jump through for this major customer. As I noted previously, the positive impact of Walmart's action for Internet commerce cannot be understated. It is the sort of "supplier mandate" that could greatly speed up general adoption of Internet-based EDI and B2B commerce in general.

More thoughts on Home Depot's priorities. My associate Lewis Marchand sent me some comments on my post regarding Home Depot's data warehouse/business intelligence project. Lewis specializes in business intelligence applications, so I was interested in his feedback. He says, "It definitely surprises me that a firm this size has not gone to BI before, given the complexity of their business. What they are proposing is huge and they are certainly going for broke." He also points out that Home Depot's plans to implement BI systems in three separate areas is aggressive. "I would think they would concentrate on areas with the greatest potential return first, which for them of course is their supply chain."

Who knows? Home Depot has over 250,000 employees in 1500 locations. Perhaps it has found that employee performance improvement, satisfaction, and retention is currently the key constraint to success, and therefore a priority for business intelligence. Or, more likely, Home Depot believes that developing BI applications in the HR area is an easier first step than doing so in the supply chain planning function. Either way, it will be an interesting case to watch.

Everyone agrees: it's a tech buyer's market. Earlier this week, Gartner CEO Michael Fleisher told 5,000 IT executives at the Gartner Symposium ITxpo that if they are planning to buy anything, now would be a good time to do it. "This is, quite simply, the best market ever for technology buyers," said Fleisher. While pointing out that the IT industry is suffering from overcapacity and the absence of any "killer application" on the horizon, he forecast that "there is essentially no chance for a tech recovery in 2003." He also predicted that 50% of all technology brands will disappear by 2004. CRN has more details on Fleisher's talk.

Along the same line, Dylan Tweney, writing for Business 2.0, announces the death of the million dollar software deal.

Home Depot is on a shopping spree for data warehouse and business intelligence tools. Home Depot announced last week that it is planning a huge roll out of a data warehouse capability that will cost tens of millions of dollars. Business intelligence applications will be rolled out in three phases: 1) HR applications to provide analytical dashboards and metrics to improve employee performance, satisfaction, and retention, 2) inventory planning applications to give material planners near-real-time access to point-of-sale (POS) transactions to better manage supply, demand, and store assortments, and 3) supplier access to the inventory and sales data, so that trading partners can better manage demand and logistics. The system will be built on IBM's DB2 database running on an IBM AIX box with sixty (60!) terabytes of storage.

This opportunity at Home Depot is an exception to the trend away from large complex software deals. Perhaps it is indicative of Home Depot's need to catch up with competitors such as Wal-mart (Walmart) that already have large analytic capabilities in place. It will be interesting to see what sort of creative deals vendors put together in order to play in this pond. Still on Home Depot's shopping list: extract, transformation, and load (ETL) tools as well as business intelligence (BI) tools for data analysis. Although, to my knowledge, no specific vendors have been mentioned as being on Home Depot's short list, expect all the major BI vendors as well as select supply chain vendors (e.g. i2, Manugistics) to want a piece of this deal. Since the data warehouse is being built on IBM's DB2 database, it is possible that IBM's Datawarehouse Manager would be considered, as well as one or more of the major ETL vendors, such as Informatica (PowerMart/Center), Ascential (Datastage XE), and SAS (Warehouse Administrator). Some leading vendors of analytic tools include Business Objects (BusinessObjects), Cognos (PowerPlay and Impromptu), Information Builders (WebFOCUS), MicroStrategy (MicroStrategy 7), Computer Associates (Eureka Suite), Sagent (Sagent Solution Platform) and Hummingbird (BI/Suite).

Buzzword alert: "open source"

Over the past month or so, I've noticed some ERP and supply chain software vendors refer to their products as "open source," when in truth all they are doing is making their proprietary source code available to clients, something that many software vendors have been doing for years. You buy a license for Package X, and the vendor provides some (or even all) of the source code so that you can modify it, but only for use within the licensing entity. That last condition is what makes proprietary code propietary. When you license proprietary software, you have no right to redistribute the product, even if you have the source code. Open source, on the other hand, is a specific licensing model whereby the author(s) of the software freely distribute the source code along with the rights to redistribute it. The details of various open source licenses, such as the GNU General Public License (GPL), are more complicated, but that's the gist. For a more information on the meaning of "open source," see the definition on the Open Source Initiative (OSI) web site. Examples of true open source software include the Linux operating system and the Apache web server. To my knowledge, no vendor of significant ERP or supply chain management application systems licenses their system on an open source model. If you know of an ERP or SCM system that is truly open source, please e-mail me.

First look at Microsoft's Axapta ERP system

Yesterday, we got a quick demo of the Axapta product, courtesy of mcaConnect, a new nationwide Axapta distributor. Microsoft obtained Axapta through its acquisition of Norwegian vendor Navision earlier this year. Axapta has not had very much market presence to this point in the US, with few people even having heard of it prior to 2001, when Navision obtained it through its acquisition of Damgaard in the Netherlands. However, Axapta has a fairly strong presence in Europe, where it is positioned toward mid-tier manufacturing firms. In the US, Microsoft Business Solutions is targeting Axapta at mid-tier manufacturing firms of $50-800M in annual sales. Based on what we saw, however, we think Microsoft would do best to aim at the lower end of that range, because the product is a mixed bag in terms of features and functions. It appears to have strong functionality in certain areas, such as a rule-based dimensional product configurator, a knowledge management module (including balanced scorecard), customer/supplier/employee survey instruments, an integrated project management module, and a Web self-service capability. On the other hand, the product is weak on multi-plant operations. For example, although it supports multiple warehouses for inventory control, it does not allow multiple facilities for planning and scheduling (i.e. multi-plant MPS and MRP). That last point alone could disqualify Axapta from many upper mid-market deals.

From a technology perspective, the Axapta product is built using its own development environment called Morphx, that generates a language called X++ (a blend of C++ and Java). I would speculate that, at some point, Microsoft may want to rewrite Axapta using its own development toolset (Visual Studio). But for the short term, Microsoft will probably be content just to run the X++ code through its .NET SDK so that it can operate within Microsoft's .NET framework. It is interesting to note that, because of their use of MS Visual Studio, other mid-market ERP vendors such as Frontstep (formerly Symix) and Made2Manage, are actually "more Microsoft than Microsoft" when compared to Axapta. And I would be surprised to see this change any time soon. It is not a trivial exercise to rewrite an entire ERP system in a new development tool set. Microsoft is more likely to devote new Axapta development efforts to address any gaps in functionality that will be required for it to compete in the US against more established players and to provide integration to Microsoft's new CRM solution. So, my advice to prospective buyers is this: if having Microsoft standing behind your ERP system is important to you, consider Axapta. But if a Microsoft-standard development environment is what you are looking for, look elsewhere for now.

Which brings us to the bottom line: the strongest point in favor of Axapta, of course, is that Microsoft is behind it. When considering the financial viability of many of the other mid-market ERP vendors, many buyers will find this a strong plus. Another strong point in Axapta's favor is its distributor network. There are reportedly 50 resellers in the US with rights to sell the product, of which about 20 are actively building Axapta sales and service groups. Many of these are former distributors of competing ERP Tier II or III products that have fallen on hard times. These resellers are experienced in selling and servicing the mid-market, and they are hungry. We have only seen Axapta in one deal so far in southern California (which it won), but we expect to see it more often as these distributors gain traction.

Large system implementations require organizational discipline

Drew Rob, writing for Datamation, has a good case study of a $22 million implementation of Peoplesoft ERP and procurement applications at the City of Los Angeles. Although there were significant issues and complaints about the new system, a post-implementation audit found that nearly all of the problems were with users and entire departments not adhering to City policies. The project team implemented the new system "by the book," but it appears that large numbers of users hadn't been reading the book to begin with. As Robb points out, "Those complaining … were actually moaning about their own processes. PeopleSoft [was implemented] in accordance with mandated city policies and processes. Yet many within the city were not in compliance."

I often complain about the poor state of software quality. However, this case study illustrates that large system implementations often fail not because of software deficiencies but because of lack of organizational discipline. Nevertheless, this story does have a happy ending. After additional post-implementation work to improve organizational disciplines, the City of LA was able to cut check processing staff in half, cut warehouse staff by 40 heads, reduce inventory from $50M to $15M, and give each City vendor a single point of contact. Furthermore, LA saved $5M a year in contract consolidation and significantly improved the number vendor discounts taken for timely payment.

Read about the case study in Datamation.