Friday, April 28, 2006

Microsoft beefs up business intelligence with bid for Proclarity

I missed this one earlier this month. Microsoft is building out its offerings for business intelligence tools with an acquisition of niche BI vendor Proclarity.

The move makes a lot of sense. Business intelligence continues to be a hot market, as companies of all sizes are putting more and more information into databases but are finding that its not as easy to pull it out and make sense of it. Secondly, the market for business intelligence software is highly fragmented, with many small players, such as ProClarity. Although there has been some vendor consolidation already (e.g. Business Objects acquisition of Crystal Decisions, and Hyperion's merger with Brio, both in 2003), there is much room for more. Although Proclarity only has about 150 employees, it is a well-respected name and Microsoft is smart to snap it up now before someone else tries.

From a technology perspective, the deal also makes sense. ProClarity's offerings already interoperate well with Microsoft's Excel, SQL Server, and SharePoint. Microsoft says that it plans to integrate ProClarity into its Server Platform business, which means buyers of Microsoft SQL Server will be less likely to third-party solutions to fill their business intelligence and reporting needs.

Builder AU has more on the deal, including other steps that Microsoft is taking to strengthen its position in business intelligence.

Thursday, April 27, 2006

Manugistics to be acquired by JDA

Earlier this week, Manugistics agreed to be acquired by JDA, one of the leading software vendors in the retail industry, ending years of speculation about Manugistics' future.

Manugistics, one of the leading supply chain management vendors, had been struggling Manugistics has had a rough for the past few years. Revenue was at a peak of $310 million in 2002, but dropped to $272M in 2003, $243M in 2004, and $193m in 2005.

JDA is one of the leading enterprise system vendors in the retail industry, the other being Oracle with its recent acquisition of Retek, and SAP, which is building its own solutions largely from scratch.

JDA has been around since the 1970s, and has been acquiring a number of smaller firms in the retail space over the past seven years, including Timera (workforce management), Engage and Zapotec (advertising, marketing, and promotions), Vista (collaborative commerce), J-Commerce (point-of-sale), E3 (inventory optimization), Neovista (data mining), Intactix (retail space management), and Arthur Retail (advanced planning and allocation).

JDA's bid for Manugistics gives it an even wider portfolio of products to offer its retail clients. The deal also gives JDA access to Manugistics' enviable list of top retail clients, such as Hershey Foods, Federated Stores, Kraft, ConAgra, Unilever, Campbell Soup, Avon, Limited Brands, Staples, Sears, Black and Decker, and Goodyear Tire.

However, one must question what JDA intends to do with those Manugistics products that are not related to retail, such as Manugistics extensive offerings in aerospace and defense, which it acquired from WDS a few years ago. Will JDA continue to support those products, or will be there be a series of divestitures from JDA to buyers interested in Manugistics' cats and dogs, as so to speak?

Supply Chain Digest has more on the JDA/Manugistics deal.

A Spectator reader also points out that Manugistics' competitor, i2, has wasted no time jumping on the PR horse, offering Manugistics' clients a program to "upgrade" to i2's products.

Update, Apr. 28. The word to me from inside Manugistics is that most employees are feeling good about the deal. The rumors and speculation about Manugistics' future were getting tiring and this deal resolves questions about its financial viability. My source, who has been through several acquisitions, says you can tell alot about a company by how its executives present themselves, and the folks from JDA have been particularly upbeat and appear to "have it together." Manugistics has lost some good leadership in the past few years and will benefit from the management depth that JDA brings to the deal.

My source points to the strong combination of consumer-packaged goods (Manugistics core strength) with retail (JDA core strength). He is less concerned about the future of other products in Manugistics' portfolio: those targeting government, travel, transportation, and hospitality sectors. Even though these comprise a small part of Manugistics' business, they are profitable and they often lead to innovations that can be introduced into Manugistics other products.

Bottom line? The deal looks good from all angles.

Related posts
SAP walks away from Retek deal
Manugistics prepping itself for the auction block?

Wednesday, April 26, 2006

Oracle upgrades Siebel's CRM software-as-a-service

Oracle last week released a new version of the CRM on-demand product that it picked up with its acquisition of Siebel. The new version (release 10), has been rebranded as Oracle CRM On Demand, and it includes improved customization capability, new features for sales and service, and enhancements for the life sciences and financial services industries.

What's most interesting to me, though, is what this says about Oracle's commitment to software-as-a-service (SaaS). Oracle has picked up many new products in the past two years, including a huge portfolio of applications from Siebel. But one of Oracle's first moves is to upgrade and rebrand this on-demand product from Siebel.

Oracle offers its own E-Business Suite as a hosted service, including its own front-office products which used to compete with Siebel's. But Oracle's offerings are all built under a single-tenant architecture: a separate installation for each client. If you visit Oracle's on-demand data center in Austin, what you'll see is racks and racks of Dell servers running Linux, each one supporting a different client.

In contrast, Siebel's offering--which it originally acquired from Upshot to compete with Salesforce.com--is a multi-tenant architecture. In other words, it can host multiple clients on the same system instance--a far more efficient and scalable approach than the single tenant architecture. At least that's the theory: Salesforce.com, the poster-child for the multi-tenant approach has suffered service outages that highlight the potential for a single failure to impact many customers.

Oracle CRM On-Demand is a tiny part of its business today, but it gives Oracle's engineers and support professionals valuable experience in the multi-tenant approach. It will be interesting to see how Oracle positions its new CRM On-Demand relative to its other CRM offerings, and whether it can avoid some of the missteps of Salesforce.com in maintaining service levels.

Related posts
Software on demand: attacking the cost structure of business systems
Big eyes, big stomach: Oracle buying Siebel
Salesforce.com's credibility suffering from service outages
On demand computing: the rebirth of service bureaus

Siebel loses $59M and responds by going on a shopping spree

Tuesday, April 25, 2006

Oracle to extend support for JDE on IBM iSeries

At its Collaborate 2006 user conference in Nashville this week, Oracle announced that it will continue support for existing J.D. Edwards Enterprise-One and World products beyond 2013, the date to which it had previously committed support for IBM's iSeries hardware (formerly AS/400).

Oracle's decision basically means it will continue to provide support for applications running over IBM's DB2 database product, which competes with Oracle's flagship database.

As far as I can tell, this week's decision does not say anything about how or whether Oracle will support IBM's DB2 as part of its Fusion strategy to merge Oracle's various applications into a single code base.

I've speculated in the past whether Oracle would try to move existing IBM database users to Oracle and when support ends whether it would try to sell off the remaining users to someone like SSA Global, which makes a business out of supporting IBM-based applications. This week's decision probably says that the answer is, for the second half of that speculation, no. Oracle probably is finding that the maintenance business for these users is profitable and better kept than sold.

Computerworld has more on Oracle's decision to continue support for JDE on IBM.

Related posts
JDE users want Oracle's Fusion to support IBM technology
Oracle mulls support for competitor databases
Oracle, IBM, Microsoft battle for technology infrastructure of PeopleSoft customers
IBM is a loser in Oracle/PeopleSoft deal
Oracle: no plan to spin off JDE product lines

Possible outcomes of Oracle's takeover bid for PeopleSoft
Green screen refuses to die

Friday, April 21, 2006

Wal-Mart sticks to RFID plans despite CIO switch

Just two weeks after I wrote a long post about Wal-Mart's IT systems under CIO Linda Dillman, Wal-Mart switches CIOs.

Dillman is still with Walmart, however. She is moving to head of risk management and benefits administration. The new CIO, Rollin Ford, comes from a position as head of the retailers supply chain and logistics function.

Dillman was the driving force behind Wal-Mart's push to RFID. For Walmart suppliers hoping that the CIO switch will mean a slow-down in Wal-Mart's RFID program, forget it. According to Line56,
One of the first actions taken by incoming CIO Rollin Ford was to publicly reaffirm Wal-Mart's RFID project. "There will be no slowing down," Ford stated.
It's not that Ford will need much of a learning curve. In his previous position, he was a member of the Wal-Mart RFID executive steering committee for the past three years. long learning curve.

Ford added that Walmart will be retiring the RFID EPC standard Gen 1, replacing it with Gen 2, effective June 30.

Computerworld has more on Wal-Mart's CIO switch and implications for its RFID program.

Related posts
An inside peek at Wal-Mart's IT systems
Wal-Mart launches RFID pilot, but will privacy concerns stall adoption?
Details on Wal-Mart's RFID specifications

Tuesday, April 18, 2006

New action in the engineer-to-order ERP space

Following the acquisition of Encompix by Made2Manage, another deal in the engineer-to-order ERP market is taking place. Now Intuitive, a small manufacturing systems provider, is buying out Relevant, a long-standing provider of project-based manufacturing systems.

Intuitive has been quiet for many years, but since being bought in November 2005 by Marlin Equity Partners, a private equity firm, it has started an acquisition program, first picking up SupplyWorks, a supply chain management provider, in March, and now acquiring Relevant.

Relevant is best known for its use at the legendary Lockheed Martin Skunk Works, where its project-orientation is particularly well-suited.

However, there is an indication that Intuitive may not be interested so much in Relevant's product but rather is buying its customer base and expertise in the ETO sector. The press release says, "Intuitive's long-term product plan is to adapt Relevant's project-driven processes and functionality to the Intuitive ERP product. Intuitive will offer continued support to Relevant's customers and deliver several enhancements under development."

There's more on the Relevant deal in a press release on the Intuitive website.

Related posts
Making money in software with a niche-industry strategy
Made2Manage acquiring ETO vendor Encompix

Thursday, April 13, 2006

Making money in software with a niche-industry strategy

Tom Milay, Vice President of Industry Solutions at Made2Manage, contacted me concerning my article last week on his firm's recent acquisition of Encompix. Generally, he felt that what I wrote was accurate, but he noted my questioning of how the firm could make money by running each of its acquisition as a separate business unit. So, I took him up on his offer to do a phone interview, which now forms the basis for this post.

Background
Back in 2003, Made2Manage was a small publicly-held Tier III ERP focused on small industrial manufacturers. I had evaluated the vendor's products once or twice and was impressed with its "small footprint" and web-based training that seemed to be appropriate for the small-company market that it was going after.

Shortly thereafter, Made2Manage was taken private by Battery Ventures, a venture capital firm, for $30 million. The word I heard from inside and outside the company was that the new management team was scaling back on product development in order to focus on selling into the installed base: sort of a "back to basics" strategy.

Since then, I noticed that Made2Manage was quietly accumulating several niche vendors in the small business ERP space, several of which I had evaluated in the past. Its acquisitions included DTR Software (ERP for plastics manufacturers), ADS (a former M2M reseller and developer of M2M extensions), Cimnet Systems (ERP for PCB manufacturers), and AXIS (ERP for manufacturers of "rolled products," such as cable and sheet metal). Then, last month Made2Manage acquired Encompix, a tiny vendor focused on the engineer-to-order segment.

So, what's behind this series of acquisitions and how does Made2Manage rationalize them? If the targets were struggling prior to being acquired, what does Made2Manage plan to do differently? Milay came to Made2Manage with the Battery Ventures deal, so he was in a position to clarify the firm's strategy.

The benefits of an industry focus
I would sum up the strategy as one of intense focus on very narrow manufacturing niches. This would be in contrast to the major Tier I vendors such as SAP and Oracle, which build systems that apply horizontally across a broad range of industries--manufacturing and non-manufacturing--with industry-specific functionality that can be switched on or off according to the client's needs.

Made2Manage is taking the opposite approach. Instead of configuring a general-purpose package to serve a specific industry, M2M offers a package that simply serves that industry, or sub-industry, and nothing more. If SAP is a Swiss army knife, M2M's portfolio of packages is a draw of knives, each for a specific use: one might be a meat cleaver knife, another a filleting knife, and another a paring knife. So, with this approach, M2M does not want to rewrite its acquired systems to a common code-base. In other words, it does not want to turn them into a single Swiss army knife. It wants to keep them as separate knives.

"The fit of a product to a specific manufacturing sector requires separate products," Milay told me. "The market is shifting away from generalized applications, such as SAP, which requires extensive configuration, to packages that fit the customer's industry sector out of the box."

A little side note, here, to say that I'm partial toward industry-specific solutions. I've seen the advantage of a niche approach in many sales demonstrations that I've viewed in the past 15 years. A tiny vendor such as DTR (plastics) or Encompix (engineer-to-order) can beat SAP or Oracle in a sales demonstration simply by bringing in presales personnel that speak the language of the prospect.

For example, an Oracle or SAP rep might be assigned to the process manufacturing sector. But even that might not be enough of a focus. The rep might be selling to a pharmaceutical manufacturer on Monday, and a winery on Tuesday. Wednesday he makes a sales call at an injection-molding plant. What does he know about injection-molding? Not much. He knows a lot about Oracle's Fusion middleware and its vision for next-generation web services but not much about finite scheduling, which is critical to an injection molding company.

On Thursday, the DTR sales guy calls on the same injection-molding plant. He's never been in a drug plant or a winery. All week he has been talking to other plastics manufacturers.

Who connects better with the prospect?

Now, DTR may not win the deal. The prospect might be more comfortable going with a name-brand, such as Oracle or SAP. But in most cases, the prospect will be far more comfortable with the DTR sales guy. I have seen scenario many times, not just with DTR but with other niche vendors as well.

Four strategic elements
With that background, here are key points of M2M's strategy, according to Milay.
  • First, shift from a software focus to a solution focus. The firm's deep industry expertise allows it to make money by providing post-implementation assessments and general business consulting, beyond the initial sale and installation of the software. This deepens the relationship with the client and brings additional revenue from the client base.
  • Second, automate customer service and support. One of the first things that M2M does when integrating a new acquisition is to document the most frequently encountered customer service problems and post them on its Web portal for customer support. It also deploys Web-based training for customers, if it does not already exist. As a result, 55% of customer incidents can be resolved by the customer using the Web, requiring little if any intervention by a customer support representative. According to Milay, automation of the support function is key to making the operating units profitable.
  • Third, centralize some corporate functions. M2M does operate each software package in its portfolio as a separate operating unit, but it is centralizing corporate functions such as marketing, administration, finance, and senior management. Although this does not cut as much cost as combining the software development organizations, it does provide some economies of scale.
  • Fourth, deliver all sales and service activities directly. M2M employs a separate direct sales force and services group for each software package in its portfolio. This flies in the face of conventional wisdom that says the best way to serve the small business market is through resellers and value-added resellers. M2M, in contrast, maintains a small selling team for new sales of each software product, plus a single salesperson to maintain contact with the installed base. On the service side, there is a small team of consultants for each software product.
I have my doubts about that last point. Over the past ten years, I've short-listed some of the products that are now in the M2M portfolio, and I've had to explain to clients why the vendor was flying in sales people from the other side of the country. It's not a plus, and it's even more of a problem when the client realizes that if he buys the software he'll be paying major travel expenses for the implementation consultants. Furthermore, I would be concerned that if M2M is successful in ramping up sales of its software products, it will find it difficult to scale up implementation services to assist those new customers. Milay agrees that there are some shortcomings with this model, but feels that the benefits outweigh the costs.

Concerning the need to maintain four separate software organizations, Milay said that over time, they do expect to realize some economies of scale in software development, by transitioning all their products to a service-oriented architecture, using Microsoft's .NET framework. Even though the packages will remain separate, there may be opportunities to share software components, especially where they provide functionality outside of the core niche manufacturing processes that make each package unique.

Milay realizes that continuing to operate each acquisition as a separate business unit may appear to be the most cost-effective approach. But it's important to see the whole picture. "We believe that our industry focus leads to greater market share in each industry niche, which leads to overall revenue for the company," he said. "This more than offsets the less-than-optimal organizational structure of operating each software package as a separate business unit."

Whether Made2Manage will be successful in the long run with its strategy is an open question. It is sticking to a traditional software license model at a time when only a few vendors are able to make money with it. Most of the buzz these days is around offering software as a service and leveraging open source. But, conventional wisdom is often wrong,

If early results are any indication, Made2Manage may have found a different path to success. The firm does not reveal financial results, but Milay assured me that ever since the company started on this road, it has been highly profitable.

Related posts
Made2Manage acquiring ETO vendor Encompix
Made2Manage sees bright future in plastics
Made2Manage going private

Monday, April 10, 2006

Open source consolidation: Red Hat bids for JBoss

The software vendor consolidation trend is spreading to the open source sector, with Linux distribution and services provider Red Hat announcing its intent to acquire the popular open source application server vendor JBoss for at least $350 million. The deal, which is subject to regulatory approval, is expected to close by the end of May.

The deal allows Red Hat, which sits squarely at the OS-level of the technology stack, the opportunity to move up into the middleware layer. It also gives Red Hat expertise and offerings for customers that are want to deploy a service-oriented architecture (SOA).

On JBoss's side, the deal gives the company access to Red Hat's worldwide distribution network. JBoss so far has been primarily selling into North America and European markets.

There are some interesting complications, however, in what the deal means for Microsoft. Although both providers are squarely in the open source market, it hasn't prevented JBoss from cooperating with Microsoft in allowing its middleware to run on top of Microsoft's Windows Server operating system. Combine this with Microsoft's recent announcement that it will support Linux running as a guest operating system under its Virtual Server product, which Microsoft is now offering as a free download.

The bedfellows keep getting stranger.

What does this mean for enterprise system buyers? It means more choices and options for deploying service oriented architectures over a low cost platform. SAP, Oracle, Lawson, and other major vendors are in the midst of a technology change to transition their systems to SOA, which allows easier mixing and matching of software components and the ability to build composite applications using pieces of existing applications. The Red Hat/JBoss combination provides a low cost OS and middleware platform that most if not all of these application vendors will want to support.

Computerworld has more on the Red Hat/JBoss deal.

Related posts
Microsoft to support Linux, virtually
Software buyers turn cheap

Sunday, April 09, 2006

Enterprise software: four horsemen and four fortresses

At Forrester's IT Forum 2006 last week, Forrester executives Andy Bartels and John Rymer gave a joint keynote on "The Future of Enterprise Software." The presentation was an excellent overview of the competing forces that are changing and will change the enterprise software market in the coming years and what that means for buyers and vendors.

Bartels and Rymer said that overall the market will not return to the double-digit growth of the late 1990s--not because there is a lack of demand from customers for new functionality but because the large company market is saturated. The primary opportunity for new sales is in the small and mid-size businesses, which have a lower price point but sometimes cost more to sell to and service. Forrester research also points to increasing frustration on the part of customers to high software maintenance fees, further throttling vendor ability to milk revenues from the installed base.

Factors that drive vendor cost and revenue
The enterprise software market today is under siege by what the speakers call "the four horsemen of commoditization:"
  • Open source software
  • Service oriented architecture (SOA)
  • Software as a service (SaaS)
  • Offshore outsourcing
These four forces are driving tremendous change in enterprise software--for both buyers and sellers.

In the case of open source, the impact so far is primarily in server operating systems and Java middleware, and each major software vendor (IBM, Microsoft, Sun, SAP, Oracle, etc.) has a different approach to incorporation or resistance to open source. Many buyers, however, are already using open source or very interested in doing so.

Service-oriented architecture (SOA) has broad strategic acceptance among vendors, and it represents a platform change not unlike the transition from client-server to Web-based architectures that started a decade ago. Forrester predicts that SOA upgrades will drive software vendor revenues higher for the next four years, but after that the ability for buyers to mix and match software components using open SOA standards will lead to increased competition that will drive software prices lower. In other words, by adopting SOA, vendors are building the seeds of their own destruction. If you are a software buyer, however, this is good news.

Software-as-a-service (Saas), or software on-demand, is being used by companies of all sizes--it is not primarily a small company market, as many observers first thought. The major vendors are all adopting SaaS in some fashion, often for limited purposes. Buyers are using it selectively.

Offshoring is a mixed trend. Contrary to popular opinion, most IT shops are not doing a lot of offshore outsourcing, although many software vendors are using it to lower their own costs of development and maintenance. Although software vendors based offshore (e.g. India) are having a hard time breaking into the major markets in North America and Europe, they could become a factor to contend with after 2012, especially those that deliver high quality software at a low cost.

What's interesting, though, is that each major software vendor is incorporating or fighting these four horsemen differently. For example, Sun is moving to 100% open source for its products, while Microsoft for the most part is fighting it. Microsoft and Salesforce.com (of course) are jumping on the bandwagon for SaaS, which Oracle and SAP appear to be forced into adopting it.

Fortresses against change
Standing up against these four horsemen, according to Rymer and Bartels, are four factors they call "fortresses of stability." These are the factors that serve to constrain the changes from open source, SOA, SaaS, and offshoring. The four fortresses are:
  • Market Concentration: the dominance of a few major vendors, such as Microsoft, especially at the operating system level.
  • Intellectual Property Rights: the vigorous enforcement of patents and other IP rights, especially to counter the threat to vendors from open source
  • Installed Bases: the tremendous effort required for a software buyer to migrate from an incumbent vendor, especially at the applications level.
  • Brand Loyalty: the comfort factor that accrues to vendors that are well-established in the software marketplace
According to Forrester, although these four factors will mitigate the forces of change in the intermediate term, over the long run they cannot maintain revenue and profit levels for the major vendors today.

For new and emerging software vendors, these trends are good news. Contrary to conventional wisdom that software market share will eventually consolidate around a few players, the large vendors today only have one third of the software market, and they are NOT growing that much faster than the market as a whole. This means there are still opportunities for new vendors to ride to success on one or more of the four horses.

For software buyers, most of this is good news. Prices for software over the long term have to fall, in line with other elements of IT spending. Buyers will be best served by standardizing on one of the "ecosystems" (i.e. SAP, Oracle, IBM, Microsoft) as a basis for choosing new software components, while at the same time making use of open source, SOA, and SaaS to drive productive change.

Related posts:
Software vendor growth not in software
Software on demand: attacking the cost structure of business systems
Software on demand: small companies still don't "get it"
Buzzword alert: "open source"
Open source: turning software sales and marketing upside down
Key advantage of open source is NOT cost savings
Offshoring leaves software firm not so jolly
Risks of offshore outsourcing

Friday, April 07, 2006

An inside peek at Wal-Mart's IT systems

Wal-Mart's CIO, Linda Dillman, gave a keynote in Las Vegas at the Forrester IT Forum 2006 earlier this week, and I found it to be one of the highlights of the conference. Dillman wasn't the typical keynote speaker, talking about "my vision for the future of IT." Rather, she presented practical insights into the IT function for one of the world's largest and most successful businesses.

IT strategy
Dillman outlined three principles for IT at Walmart. These reflect Wal-Mart's strategic "IT maxims" (although she didn't use that term), which guide development and deployment of all IT systems and capabilities.
  • Merchants first. IT personnel at Wal-Mart should be merchants first, and technologists second. IT talks to customers (users) not in terms of technology but in terms of the business. "Our users don't know much about our technology platforms or tools. They shouldn't care whether we are running mainframes or UNIX," she said.
  • Run common systems and common platforms. Wal-Mart runs a single system with a single set of code worldwide. "The first thing we do when beginning operations in a new country is to migrate store operations to our code," she said. "The cost advantages are huge: Our IT budget is well less than one-half a percent of sales. We can do this because we do not have to invest in multiple systems."

    In addition to the cost benefits, the single-system approach allows Wal-Mart to leverage best practices, which are embedded in the system, across regions. When up-and-coming executives transfer to a different part of the world, they have the same system and processes that they have been used to in their old job. This supports Wal-Mart's leveraging of human resources worldwide.
  • Centralized information systems. Wal-Mart runs all worldwide information systems out of its headquarters in Arkansas, with a second data center providing backup and failover. "A point-of-sale transaction entered in China comes back to Arkansas for credit card authorization and then returns to China to complete the sale," she said. "The whole process takes place in less than half a second."

    One benefit of this approach is that most of Wal-Mart's developers are in one place, allowing them to more easily collaborate. The second benefit is that the developers are located in the heart of the business, eating lunch with buyers and talking about issues in retailing. This keeps the developers tuned in with the real concerns and needs of the business.
Regarding best practices, Wal-Mart takes a middle ground. On the one hand, it does not force all regions worldwide to do things exactly the same way. Nor does it let each region do its own thing. The middle ground is to define the best practices specifically for each region.

Wal-Mart believes that some core functions apply to all regions. For example, "every day low price" can and should be applied in all markets, regardless of what the local managers think from their past experience. They do not need to do special price promotions. "When local managers change over to every day low pricing, they find that it does work and they never look back," she said. "This is non-negotiable."

Other practices can vary by market, and when Wal-Mart finds a new best practice in a local market, the developers program it into the core system. "Localization is handled by turning things on or off. In some cases, we turn off large blocks of functionality for small markets, because it would just hurt their productivity," she said.

Wal-Mart's supplier network
Dillman spoke about the growth of Walmart's digital network, now known as RetailLink, which was initiated in 1991 as a data warehouse providing daily sales data. According to Dillman, at the time, if they had done an ROI analysis on this initiative, they would have never approved it. "We just did it on gut instinct," she said.

The development of Retail Link, by which suppliers today have access to sales, shipment, orders, returns and other data on their products in Wal-Mart stores, flies in the face of retail mentality. Traditionally, because knowledge is power, retailers and suppliers do not share information. But Retail Link has shown the value to both parties of making information available.

Wal-Mart's RFID initiative has also shown the benefits of information sharing. Gillette, for example was able to tell from RFID data which stores did not get product out to the selling floor in time for a new product launch date and was able to discount such stores from their sales analysis. A smaller supplier that provides Christmas seasonal merchandise was able to track pallets through Wal-Mart's distribution chain. They saw that a group of pallets went into a DC but were not moving out to stores. They alerted the DC to the problem, which was able to expedite delivery to stores in time for the holiday season, saving the supplier from having to suffer lost sales and mark-downs.

Wal-Mart's development practices
Dillman was asked about Wal-Mart's view of buying versus building its applications and whether it was making use of service-oriented architectures (SOA) in development. She indicated that Wal-Mart does use some packaged applications for some functions, but for the "core" system, it is all in-house developed code.

Regarding SOA, she took a pragmatic view. She indicated that SOA as a technology will not by itself lead to faster and more flexible software development. She attributed Wal-Mart's success in developing and extending its core systems to the fact that they write all their own code and do so in a highly modular approach. In recent years, as Wal-Mart's IT group has gotten much larger, it has had to formalize its best practices so that they can be promoted among all staff members. This, in her view, is more important than the technology of SOA.

Dillman's presentation gave me a different perspective on Wal-Mart. Wal-Mart has long been known as a company that pushes its suppliers to do business electronically, and much of what is written about Wal-Mart in the technology press is from the perspective of the supplier that has to comply with Wal-Mart's mandates. But Dillman's presentation provides a different perspective--from inside Wal-Mart--and it shows how one very large organization uses IT to a competitive advantage, while spending much less on IT than most of its competitors.

Related posts
Wal-Mart launches RFID pilot, but will privacy concerns stall adoption?
Outsourcing: what would Wal-Mart do?
Wal-mart suppliers face October deadline for Internet-based EDI

Monday, April 03, 2006

Microsoft to support Linux, virtually

I'm at the Forrester IT Forum 2006 in Las Vegas this week, blogging the conference. The first keynote this afternoon was by Andy Lees, a Microsoft VP in charge of server and tools marketing. Speaking at an incredibly fast pace, Lees is talking about what Microsoft is doing to increase IT productivity. One of his main points is to reduce the complexity of IT, which is of course a noble objective.

But there was one an interesting announcement that he made: Microsoft is going to start giving away its server virtualization technology (its Virtual Server 2005 R2 product), which allows one physical computer to run multiple instances of a “guest” operating system. The real kicker, though, is that Microsoft will not limit guest operating systems to its own products. It is also offering, at no-charge, virtual machine “add-ins” to run select Linux distributions, along with technical support to help customers consolidate Linux-based applications under Microsoft's Virtual Server 2005 R2.

Microsoft made this announcement here at Forrester's conference and, more significantly, at LinuxWorld in Boston today.

Some Linux fans will want to attribute some nefarious motive to Microsoft's support for Linux. But I think it's rather a sign that Linux is a fact of life in many data centers and that Microsoft is better off being part of the solution to Linux support rather than pretend it doesn't exist. Whether system administrators will want to let Linux run on top of Microsoft’s Virtual Server is another issue, of course.

There are more details in a Q&A document on the Microsoft website.

Coincidentally, I’m currently running a Computer Economics Quick Poll on the virtues of Linux and Microsoft server operating systems. If you’d like to participate and receive a free copy of the analysis resulting from this survey, you can take the survey now.