-
Font Size:
-
Print
- TweetThis
Although I doubt SAP lost share in 2006, the share-gain claim is probably more of a preemptive strike against a sure-thing-in-the-Spring Oracle (ORCL) bombast (about taking share away from SAP) than it is crisp actionable market-share research by SAP. The announcement timing is also odd in that the claim was released almost two weeks before the year-end financial-analysts’ conference call, scheduled for Jan 24, and a couple of months before we can expect a 20-F up on Edgar. It looks like the market already punished SAP for the low-range bottom line results also announced on January 12 but let me pour some gasoline on the fire by raising some key questions about the implications of the top line results:
SAP says its worldwide share is based on a group of approximately 30 “Core Enterprise Applications” vendors that SAP projects will account for approximately $16.4 billion in software revenues in 2006. I cannot find such a list in the SAP website or SEC-filed material but I’ll keep looking. With many more than 30 players in the enterprise-applications market (even after six years of consolidation), picking the 30 can make a huge difference in any “projected growth rate” and resultant share. Does the core group include Cerner, which grew its equivalent of “software revenue” 13% in the first nine months of 2006, for example, as well as i2 whose equivalent of “software revenue” deflated in the same period? For that matter, what is an “enterprise application?” In my opinion, enterprise applications “software revenue” runs north of $30 billion a year (not that I think that software revenue is a good metric; see below). Who forecasted the growth of the 30 against which SAP is comparing itself. The press release reads as if SAP itself did the forecast based on industry analyst estimates. Most industry analysts that I am familiar with do not publicly forecast individual companies, just markets, so you could have some interesting averaging at work here. But more important, does it make sense to only look at what SAP calls software revenue (roughly, traditional licenses), particularly if revenue from companies such as salesforce.com’s (CRM) (which is not reported as license revenue) is not included in the “core 30” total? And even if such salesforce.com-like revenue is included in software revenue in SAP's methodology, it is more important to look at software and maintenance (what SAP calls product revenue) because maintenance revenue is essentially a license renewal And was NetWeaver revenue treated as application revenue in this analysis (if SAP kept up the pace in middleware that it announced in early 2006 for 2005 that could account for 1% of the share right there? Even when the 20-F comes out, we'll probably have to guess because SAP said in January 2006 that it was going to stop reporting its results by product line. Finally, it looks like SAP used $1.28 to the euro in 2006. What was that number in 2005 and how did it help the comparison?
I am not suggesting that you don't invest in SAP, but be careful if your primary incentive is the quoted applications market share gain. I will follow up and answer my own questions as much as possible after the Jan 24 conference call and 20-F filing (not to mention after the rest of the companies file before we know whether the key 4% number is accurate.)
Related Articles
|























This article has 1 comment:
As noted in my orignal posting above, I need more information from SAP, and almost everyone else in the market, before confirming or questioning SAP's enterprise-application market share gain claims made on January 12, 2007 for 2006.
But at least against Oracle, which I suspect is all SAP cares about, the claim looks valid.
If you make some assumptions about:
- exchange rates ($1.19 to the euro in 2005, $1.20 to the euro in 2006)
- what Peoplesoft and Siebel did standalone in January 2005 and 2006 respectively (zero)
- what Retek/Portal/iFlex/Ste... did in their last independent quarters (basically same as their final reported quarters)
then:
-- SAP grew both license and license/maintenance revenue (what SAP calls software and product revenue respectively) between 12% and 12.5%
-- After backcasting, Oracle grew the comparable numbers (what Oracle calls "new software licenses" and "software license updates and product support" respectively) around 9% and 4.5% respectively.
No matter what the size of the market--or how you define enterprise applications--it looks to me as if SAP gained share on Oracle in the enterprise applications market, not the other way around as claimed by Oracle in December 2006.
[Per my methodology, SAP's reported numbers are multiplied by .95 to adjust for NetWeaver before being adjusted up for exchange rate. I have made no adjustment to SAP for Frictionless Commerce, Virsa, and Khimetrics, which would lower the SAP growth rate a bit but I suspect SAP will end up using a higher euro/dollar exchange rate so the two factors will tend to even up. I use the same exchange rate (whatever SAP publishes in their 20-F) for all quarters.
[Also, per my methodology, the Oracle numbers are based on the 12 months ending November 30, 2006 and I am using the lower "...product support" numbers that Oracle began reporting in mid FY 2006 for 2005 rather than the originally reported "...product support" numbers. Oracle lowerd the 2005 numbers (which increases the growth rate) for accounting reasons related to their acquisitions.
[In general, I am using the same methodology as I used for many years at IDC but IDC numbers may differ if they have changed methodolgy since I left. In addition, I believe IDC may try to make some estimate for the Decembers to normalize Oracle to the calendar.]
The spreadsheet is available at my website if anyone is interested.