Technology keirutsu
Sun acquires StorageTek, Lawson merges with Intentia …the M&A binge continues. I believe, underneath these transactions, the industry has been moving towards the way the Japanese manufacturing ecosystem is structured. GM, Ford in autos and Boeing, Airbus in aviation are similarly structured. Original Equipment Manufacturer – the OEM – with Tier 1 suppliers that supply major “systems” like nose gear assemblies, and Tier 2 and 3s which supply various parts and materials.
That is in the manufactured, product world. “Business process OEMs” like Citibank and GE and UPS are emerging and will increasingly compete with IBM and EDS and SAP to provide mortgage and health care and plant maintenance and countless other processes.
They will encourage vendor tiering as product OEMs have done. And since their business process OEM supply chain is more digital, this will hugely affect the IT/BPO industry. Major vendors (HP, Microsoft) will become Tier 1 vendors just as Delphi and Federal-Mogul are to autos. Specialist IT vendors will increasingly subcontract to them – as Tier 2 or 3 suppliers.
Here are some implications:
a) Tier 1 candidates include hardware/infrastructure heritage vendors (IBM, HP, Sun, Cisco), some from telecommunications (Verizon, BT), some from software (SAP, Microsoft, Oracle), others from IT services (EDS, Accenture, TCS/Infosys), with possibly some wild cards from content and BPO worlds. OEMs will expect more stringent quality, collaborative design, timeliness, security and other service levels of these providers (than have traditionally been expected or delivered by most IT vendors). The plumbing/infrastructure investments and the performance guarantees the OEMs will expect will limit the number of viable Tier 1s to no more than 50 around the world. Tier 1s will have to learn to not try to own or build everything themselves, but to leverage the best Tier 2 and 3s.
b) As we see in Detroit, Tier 2 providers are at times more profitable than the Tier 1s or the OEMs. Others that end up undifferentiated, though suffer through price pressure via reverse auctions. Tier 2 and 3’s will learn to align with Tier 1 vendors rather than try to break their necks and bank accounts selling direct to the OEMs. For all of Larry Ellison’s talk, he cannot buy up everything. For most offerings, he will need to partner with a bunch of Tier 2 and 3 contributors.
c) The OEM CIO becomes huge in this world. He/she will decide how much to buy v/s build, who to buy from, understand their economics (yes, even tier 2 and 3 economics) intimately, drive performance scorecards, collaborate more with vendors - and yes, at times “be like Jose Lopez at GM” - tear up vendor contracts and dictate what you will pay and ask for price reductions across your Tier 1,2 and 3 suppliers. In some cases, they will “encourage” Tier 1s do business with specific Tier 2s.
d) The Business Development function in vendors will become much more important as alliances, some long term, some tactical are formed and unwound.
Silicon Valley – digest your consolidations. Then start learning from Detroit about the realities of industry tiering.


You present an interesting scenario but the parrallels don't play out. The fundimental differences between the industries . In manufacturing, 50,000 widget pieces come together to make a widget, 42 times a minute, 24 hours a day. To manage widget design and complexity it is easier to get a Tier 1 that handles design, logistics, sourcing, etc of say 1,500 widget pieces and sends you a partial widget module. Production efficiency is the key to giving the customer what they want, a standard, low cost widget. Economies of scale and leveraging capital for cost reduction are key factors that make the multi-tier supply chain effective.
With business process delivery, customers are not looking for 1 size fits all widgets, and there are not 42 million customers that are going to buy a business process widget every 2 years. This does not lend itself to the same effective economies of scale at each tier of supplier, and an astute customer (customers at each tier, from customer to OEM to Tier 1 to Tier 2) would question the value of paying someone else for a service they could do as well as or better than their supplier.
Your model was tried before, in the VAR days where computers, software, and tools were sold to VARs who resold packaged solutions. Businesses realised that they were giving away profits to VARs and took over much of the day to day work of VARs. One huge disadvantage VARs have over manufacturers, VARs cannot leverage capital and economies of scale as well in their industry, so a VAR's customer can begin performing the VAR's service for themself and make all of the VAR's profit.
If it was cheaper for a large widget customer to buy the parts and build their own widgets it would be foolish for them to keep widget makers in business. The fundimental differences in large scale repetitive manufacturing and providing BP solutions make the multi tier supply model less feasible in BP than manufacturing.
Posted by: Devin Taylor | July 28, 2006 at 11:04 AM
Devin, I think it is already happening in many asepcts of tech. In a recent article on telcos, BusinessWeek said if Verizon and SBC had been more focused on R&D, Cisco and Nortel would not exist. They provide sohisticated network gear, switches etc. SAP is investing in a Netweaver ecosystem of smaller software companies. Chip manufacturers have outsourced manufacturing for a while now. Outsourcing firms depend on staffing firms to fill contingent requirements. So the tiering is already happening but not as pronounced as in mainstream manufacturing. Also in tech we seem ashamed on buying versus building. Tech firms typically do not have good vendor managers, procurement processes...
Posted by: viinnie mirchandani | July 28, 2006 at 06:05 PM