Conference ulysse::rdb_vms_competition

    Just so no one gets confused...
                      
    "Are the top...databases on VAX VMS" could be deceiving. While they
    are certainly the best databases in that environment, to be correct,
    we must say that top as in "market leader" is questionable.
                                                           
    According to Computer Intelligence, Oracle still is the number one
    relational database on both VAX VMS and VAX Ultrix/Unix systems in the
    U.S. The margin of leadership is shrinking. However, I would hate
    to have someone tell a customer that we are "number one on VAX VMS"
    as long as Oracle could show the CI numbers that indicate the contrary.
                                                           
    Worldwide, Rdb may have a lead, but it is impossible to prove.
                                                           
    Again, this reply is only inserted to clarify the comment in the
    base note.                                             
                                                           
    ---- Michael Booth                                     
                                                           

re: -1

Jamey Nordby (DBS, Colorado) sent along these Dataquest numbers in response to
the Computer Intelligence Corp. report They show the same general trends but
with markedly more robust share of market for Rdb. I think they support the
original statement.

Dataquest 
Share of market numbers

Market Share by 1987 revenues:
Rdb 27%
Oracle 27%
DBMS 10%
Ingres 16%

Market Share by total Licenses
Rdb 33%
Oracle 24%
DBMS 17%
Ingres 9%

(this is after Calandar 1987)

    Something is wrong with these numbers.
    
    According to this survey, Oracle sold almost 75% as many licenses
    as Rdb last year. But the Oracle revenue was equal to Rdb. Since
    Oracle prices are more than twice Rdb's, and Oracle has no run-time
    license, Oracle's revenue should have been drastically higher than
    ours. 
    
    Further, it is generally accepted that Oracle is very effective
    at the high-end sales, while Rdb has until very recently tended
    to the low-end. That should further increase the Oracle revenue
    advantage.
    
    It is because of these discrepencies that I question the IDC survey
    numbers. They conflict with numbers put out by Computer Intelligence,
    and frankly with every other marketing survey numbers I have seen.
    
    So all we can say is that there is not a unanimous decision by the
    market surveys. It is, however, worth noting that IDC is the only
    survey that shows Rdb leading the VAX/VMS market. If there were
    any corroboration, I would welcome the IDC information, and be one
    of the first to trumpet the "we're number one" theme. Until and
    unless that corroboration occurs, I will accept the information
    from the vast majority of market surveys that show Oracle as number
    one.
    
    ---- Michael Booth

        <Please feel free to reply to this note if you wish to comment>

ref: .3

It would seem to me that by selling more licenses at a lower price (Rdb/VMS)
would be close to selling less licenses at a higher price (ORACLE). This is
what the Dataquest revenue numbers apparently show. Rdb/VMS has more licenses
than ORACLE but the revenues were approximately equal. I say approximately
since I don't believe ANY survey to be exact. 

It is a known fact that at the last ORACLE Users Meeting, ORACLE announced
that they had sold 7500 licenses and we know that Digital already had 8000+
licenses at that time. This appears to me to put us in front of them which
is in line with the Dataquest Total Licenses sold numbers.

Those of us in the Database Systems Consulting Group have always, and will
continue to, put forth as positive as possible message about Rdb/VMS.

You sir, may continue to be negative if you wish.

Sincerely,

Larry M. Carpenter

    Re: .4
    
    >It would seem to me that by selling more licenses at a lower price
    >(Rdb/VMS) would be close to selling less licenses at a higher price
    >(ORACLE).
    
    Depends on the differential between prices and numbers of licenses.
    Given his price estimate (ORACLE more than twice Rdb/VMS), he's
    right; they're mathematically inconsistent.
    
    Assume, just to make the math easy, that the number of licenses
    in the market is 100.  So Rdb/VMS has 33 licenses and ORACLE has
    24 licenses.  Now Rdb/VMS costs $1 and ORACLE costs $2.  So Rdb/VMS
    revenues are $33 and ORACLE revenues are $48.  ORACLE's revenues
    are about 150% of Rdb/VMS revenues.
    
    The only way I can think for the numbers to be valid (barring a
    false assumption of price ratio) is if ORACLE sold lots of licenses
    for smaller machines while Digital sold more Rdb/VMS licenses for
    bigger machines (since license price is based on CPU, right?).

	We now know where the study comes from so we can now choose to quote 
	from the study or not.  

	We can discuss what the study means all we want, but without knowing
	they looked at and how they looked at it, it is just an interesting
	conversation.

	There was a quote in the Columbus Dispatch that I read a few years ago
	that applies here.

	"Statistics are like a street light, some use them for support, 
	 others use them for illumination."

1) We charge different prices on different systems, and so does ORACLE.
So, how can one be sure what the actual licenses/dollars mix is?  Also,
isn't ORACLE more likely to discount off of their list prices to lock in
the business?  Finally, what is the situation (or was the situation)
with OEM pricing between the two vendors?

2) The 2:1 pricing ratio is absolute bullshit.  Maybe if you buy ALL of
ORACLE then they get up to the 2:1 mark, but they don't sell it that
way.  For example, the precompiler for each language is a separate line
item!  How many people buy more than 1 or 2 precompilers?

3) I don't understand why we would use the less positive of the surverys
for marketing purposes.  For engineering planning purposes we may want
to see ourselves in as negative light as possible to ensure that our
future plans are aggressive enough.  But for marketing purposes, we want
to shove the positive numbers down the throats of whomever we can!

    This IS growing old. ALL I was trying to do is to say that there
    is overwhelming evidence that Oracle is currently number one. I
    just got back from Euro. Despite the success of Rdb there, IDC shows
    Oracle leading in UK, France, Germany...the three largest European
    markets. I just got market numbers from IDC Australia, and they
    show Oracle with a sizable lead as well.
    
    On Pricing:
    
   I didn't spend weeks looking at the pricing and strategy of Oracle
    to be erooneous in my conclusions. Yes, they discount. But I ask
    you, did we make $140,000,000 on Rdb last year as Oracle did in
    the VAX market?
    
    ON Marketing:
    
    Certainly we can trumpet the IDC claim to customers. But be prepared
    that Oracle has surveys "out the wazoo" that dispute it. My argument
    isn't with the numbers. Externally, I would always go with the positive
    numbers. But I do not think it is wise to kid ourselves internally,
    thereby becoming complacent in the database market.
    
    ---- Michael Booth

>    you, did we make $140,000,000 on Rdb last year as Oracle did in

The real question is, did Oracle make that on its database system or
other software as well?  What happens when we include Datatrieve,
Teamdata, Rally, TDMS, and CDD in our revenue totals?  How about adding
in the services generated by these products (which already show up in
the Oracle numbers)?  I'll bet we blow away the $140 million in a true
apples-to-apples comparison, if such a thing exists.

Oracle is likely the real number one, however we should use whatever
data we can to point out that their position is not overwhelming and is,
indeed, tenuous.  This will at least blunt their claims to be the clear
leader of VAX, and falsehood we should not let them exploit.

Several references have been made here regarding the Computer Intelligance 
Corp. database market share numbers.  A summary of the actual numbers is as 
follows:  (Note that the survey covers all VAX/VMS database products 
including CODASYL).



             CHANGES IN VAX/VMS DBMS PRODUCT PENETRATION
                 FROM END OF YEAR '87 TO OCTOBER '88
                   FOR VAX Rdb, ORACLE, AND INGRES
                                            
                 Source: Computer Intelligence Corp.
--------------------------------------------------------------------------------
# Licenses          % of All              #              % OF ALL        
                    LICENSES            SITES              SITES     
 -------------------------------------------------------------------------------

VAX Rdb/VMS:

Y-end 87
 1071                16.61                405              13.29      


Oct 88               
 1489                17.85                658              15.35


Delta
 +39%               +1.24%               +62%              +2.06%       
____________________________________________________________________________

Oracle:

Y-end 87
 1117                17.32                461              15.12


Oct 88               
 1496                17.94                713              16.63


Delta
 +34%                +0.62%               +55%            +1.51%
___________________________________________________________________________

RTI (INGRES)

Y-end 87
 859                 13.32               330               10.83


Oct 88               
 1087                13.03               514               11.99


Delta                
 +27%               -0.29%              +58%               +1.16%



What the summary shows is that:

1 - The Rdb installed base increased by 39% between mid-1987 and Oct. 1988
    vs. an increase of only 34% for Oracle.

2 - Rdb and Oracle are virtually equal in market share, even though Oracle 
    has been shipping since 1979, and Rdb only since 1984!!

Note: The same survey shows that the "formidable" competitor, Sybase, has an
insignificant .13% of the VAX/VMS database market vs. Rdb's 17.85%.

The only surveys with which I am familiar that track database market share on 
VMS are the CIC and Dataquest surveys.  They clearly show that Rdb is the 
fastest growing RDMS on VMS, mostly at the expense of Ingres, and that
    Rdb has the 
largest installed base.  I have seen NO numbers to refute that!

Attempting to validate these numbers by extrapolating from revenue figures 
seems to be a waste of time.  1) It is impossible due to discounting, 
different ways of reporting revenue, etc. and 2) and more importantly, a 
statement by two respected independent consultants that Rdb is the market 
share leader and fastest growing VMS database is the ultimate goal.  Let's 
not waste effort questioning or disputing it, let's get out and market it 
while it is current!  This is the message that I delivered at the
Federal Govt. Sales Training Symposium in November, at the Channels 
Marketing sponsored Executive Seminar earlier this week, and that I will
present at the National Database Symposium next week in Boston.  It is
the message that we will deliver to internal and external audiences at every
other forum in which we participate.  It will be a major component of the 
Digital Database executive overview currently being developed for use by 
sales.

Vickie Farrell,
Marketing manager, DBS

    my two cents worth.
    
    1) The market share analysis that I have been looking at is from
       Dataquest, not IDC.
    
    2) Oracle mad about $50million on the database product in 1987,
       not 140,000,000.
    
    3) There is at least 2:1 difference in prices between the oracle
       database kernel and the Rdb kernel. (List price on a 6240 for
       oracle is $120,000, Rdb: 57,000. Rdb includes precompilers and
       VAX/SQL. A look at maintenance charges is mind significantly
       INCREASES the gap when looking at true cost of ownership.
    
    4) Computer Intelligence is considered one of the most unreliable
       sources for VAX information. I had on-line access to their database
       for over a year in a previous life.
    
    5) Which of you are in marketing and which of you are the engineers.
       Don't marketeers present the best possible light to EVERYBODY,
       internally and externally. RTI is still claiming to be the #1
       database in their sales seminars, quoting a 2 year old study
       that favored them. Find whatever source you can that says we
       are the best, and USE it.
    
    6) In addition to the CI report, what other myriad of studies are
       there. please list them and their results.
    
    7) Oracle and Digital both discount heavily and each has a different
       hardware penetration. Having talked to the author of the Dataquest
       report many times, I am fairly confident that these are acurate
       numbers for what they are studying.
    
    Jamey

    Gartner Group Research Note # K-440-595.1 shows the following market
    share on VMS platforms:
    
            Licenses    12 month growth
    Rdb     7,550           3,500
    Oracle  7,000             500
    Sybase    300             230
    Ingres  3,500             400
    
    Estimates as of 10/88