The surprise was not that i2 Technologies was bought, but by whom and for how much. i2 was on the market for a long time with few takers, though there was little comment about SAP, Oracle, IBM, or Microsoft as suitors. This is in itself interesting as it would seem to indicate little perceived incremental benefit to the gorillas in the market. Even though these companies, to varying degrees, are still acquiring companies, they are outside of the core supply chain management space. While i2 has recently emphasized services rather than software sales, which would reduce the market value, a large portion of its revenue still came from license sales and subscriptions, so the price is a surprise. There is still some chatter that other companies will enter into a bidding war for i2, but I doubt it. There is too little incremental value to SAP APO and Oracle applications, IBM is not in the application space, and there is too little Microsoft content in the i2 solutions for Microsoft to be interested.
Looking back, it is interesting to observe some patterns, which naturally are not that apparent at the time.
- SAP and Oracle have for years positioned that a single provider is less risky than adopting a best-of-breed approach, principally because of the integration effort required. Larry Ellison is famously quoted as saying “Best-of-breed is dead. Unless you are at a dog show”. Cute, but is it accurate? I think not. The reality is that both SAP and Oracle sell modules to satisfy specific business needs, and that the integration between the modules is no less complex than it is between a best-of-breed solution and an ERP. However, around 1998 Geoffrey Moore presented at an i2 user conference shortly after “Crossing the Chasm” had been published. The central theme of his talk was “who owns the data, owns the process”. No one at i2 seems to have paid attention to the message.
- The other very effective positioning strategy used against best-of-breed solutions, particularly by SAP, is that the business processes embedded in their ERP and APS solutions are best practice, so it is best to adopt these business processes rather than adapt the software to the existing business processes. Of course this is true of non-differentiated business processes, and also masks the cost of customizing SAP and Oracle, as well as the headaches created with upgrades.
It is interesting that JDA bought both Manugistics and i2, which were the darlings of the supply chain management space in the late 1990’s, both of which faltered after the .com bubble burst. Every indication is that the Manugistics purchase was successful for JDA, though the purchase has not provided a lot of growth to JDA beyond the original incremental revenue. JDA adopted a custodial approach focused principally on financial management and maintenance renewal, not on innovation. It is likely that JDA will adopt the same approach to the i2 purchase, though in this case they will use i2 solutions to strengthen their position in discrete manufacturing, whereas Manugistics was applied to retail and consumer products. In a maturing market this may be the right approach, but it is difficult to see how JDA will combat the SAP and Oracle dominance without innovation.
Given the dominance of SAP and Oracle, can any other solutions be classified as anything other than best-in-breed? JDA has maintained the best-of-breed approach. It will be interesting to see if they can be successful with it. Other best-of-breed players to watch are E2Open, GXS, GTNexus, Infor, Lawson, Logility, MCA Solutions, Sterling Commerce, and WeSupply, amongst others.
Best-of-breed is not dead. There is still space to challenge the planning paradigm developed by i2 and Manugistics in the early 1990’s, and adopted by SAP and Oracle. These solutions were developed to address some fundamental issues with MRP, namely that it does not consider demand, capacity, and supply simultaneously, and that it was awfully slow. However, at the time companies were much more locally based and fully integrated than they are today. I could be wrong, but I was not aware of the term “fabless semiconductor” in the early 1990’s. Outsourcing and off-shoring were still novel concepts. The supply chain planning solutions developed to satisfy that world are not best suited to a massively outsourced, global, and heterogeneous data source environment. The legacy planning solutions assume a world of centralized command and control in which a few planners in the dominant player in the supply chain determine the fate of all the other actors. This was possible when companies were fully integrated and had full knowledge of the capacity and inventories are all locations in the supply chain, and suppliers only provided commodity components.
This is no longer the case. The general business environment is a lot more dynamic than in the 1990’s with customer expectations of delivery have shortened tremendously. At the same time much of the manufacturing capability has been outsourced, meaning that companies have to be much more responsive even though they have less direct control over manufacturing. What is required in today’s environment is more collaboration than control; more coordination than optimization. At the core of the SAP and Oracle solutions is the concept that they own and control all the data. This is simply not true when a supply chain may consist of 3 or more tiers, each with their own ERP system. This brings us back to the issue of integration. No longer is it a issue of integration between modules, but of integration between companies each having their own ERP systems and other data sources. No longer is it about optimization of assets, but of effective coordination between trading partners. I think there is a lot of room for niche players to fill this space.
You can find additional commentary on the i2 acquisition on Ben Worthen’s Wall Street Journal’s Business Technology blog here (subscription required).