Oracle consolidation is ‘one-way bet’
Dangerous strategy, Bloor argues
It is entirely plausible that Oracle's attempt to buy PeopleSoft was an attempt to throw a spanner in the works of the latter's acquisition of J D Edwards, writes Phil Howard of Bloor Research.
However, it is equally possible that the conspiracy theorists are wrong. If we assume the latter, for the sake of argument, this means that Oracle is promoting consolidation in at least two (apparently) distinct areas of its business. It is worth comparing and contrasting these.
The first of these, as we all know, is with respect to Larry's idea that there should only be two mainstream ERP vendors. His contention is that neither Oracle on its own, nor PeopleSoft in conjunction with J D Edwards, can compete with SAP and that, unless something is done about it, then SAP will rule the roost. His view is that only a strong second force can compete effectively with SAP.
This is a plausible argument but I don't agree with it. If we compare it to the business intelligence market, for example, the situation is exactly analogous. SAS is the market leader and it is bigger than both of its new nearest competitors (Business Objects and Cognos) put together. Yet I do not see any suggestion that these two companies should combine.
If we stick with the BI space for a moment, a more likely scenario is that vendors such as Business Objects and Cognos grow by acquisition. This is precisely what they have been doing, with the former buying Acta last year and the latter purchasing Adaytum.
Going back to ERP: there are lots of smaller companies in this environment. Moreover, these typically specialise in the SME space, which all the larger vendors are trying (often not very successfully) to penetrate. So wouldn't it make sense to consolidate the market by mopping up some of these suppliers?
Let's move on to Oracle's other consolidation strategy. This is its approach to multiple databases. IBM and, to a lesser extent, Microsoft, advocate the use of federated databases; that is, leave your existing databases in place but tie them together with technology such as DB2 Information Integrator. Oracle's approach is that you should replace existing databases, where it is sensible to do so, and consolidate these into a single central database.
Clearly there are arguments for and against both consolidation and federation when it comes to databases, and I shall not rehearse them here. However, it is worth noting the major argument made against consolidation - that it is risky and potentially expensive. Ripping out existing databases and consolidating the data and applications into another system has much more scope for problems than federation, particularly if those databases are non-Oracle databases in the first place.
Moreover, this is a one-way strategy: if it doesn't work, going back to what you had before will also be very expensive. With federation, on the other hand, you still have the option of consolidation if federation doesn't work in the way that you had hoped it would.
These arguments equally apply to consolidation in terms of ERP. Oracle buying PeopleSoft is a risky strategy. Not for Oracle perhaps, since it will not only gain new customers but should be in a position to convert significant numbers away from competitive database products to boost usage of its own database.
No, I think it's a risky strategy for the market. It's another one-way bet. Strategies are at their best when they give you multiple ways to win. As far as I can see, the proposed acquisition of PeopleSoft has only one way to win and that's dangerous.