IBM's pricing model (was RE: Model 400 Replacement ?) -- MIDRANGE-L

-----Original Message-----
From: midrange-l-admin@midrange.com
[mailto:midrange-l-admin@midrange.com]On Behalf Of jt
Sent: Sunday, December 16, 2001 3:07 PM
To: midrange-l@midrange.com
Subject: RE: Model 400 Replacement ?

<snip>

I was going to write a flame on IBM's pricing, because your clients being
able to run a B20, up until recently, and there continuing on with the
platform..  Well.. it's proof-positive that the bottom-line depends more on
satisyfing customers, rather than churning hardware sales.  It wasn't really
going to be a flame though, as I think there is an extremely simple pricing
strategy that provides the best of both worlds, as the server market
continues to become a commodity market, according to the CW...  However,
that's NOT the case in ALL companies, as this customer demonstrates...!

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I'm sorry, but I don't recall who stated (more-or-less) the fact that the
interactive penalty is based on this premise:  There are customers who are
used to paying top-dollar, and they generally still use 5250 displays.  And
there is a segment of computer purchasers who view the server market as a
commodity.

IMHO, those who view servers as a commodity end up paying the same, or more,
than an iSeries.  I believe the TCO white-papers that substantiate this.
Only makes common sense, to me.  I think this is one of those cases where
you get what you pay for.


The interactive "tax", however is NOT one of those cases.


The difference is clear to see, in the threads about the Fast400 tool.
People may be willing to pay out the nose, but more and more aren't...
Those that are willing, are NOT HAPPY CAMPERS...  They were told, by Mr.
Gerstner amongst others, that there was NO SUCHA THING as a GOVERNOR in the
iSeries...  The facts contradict that.  (Don't care what you call it.  Use
any terminology you want.. it's still a governor.  My impression is that
someone probably slurred the terminology, to convince Gerstner there wasn't
one.  But, of course, I'd have no way of knowing and ICBW.)

Now, I have no idea if this group I'm involved with is gonna bring a product
to market...  A product that doesn't require any bit-twiddling to end the
need for the interactive penalty.  Hard to predict THAT future, but I don't
need any crystal ball to state SOMEBODY'S gonna DO IT.  Probably within the
next year...

If you START with that POV, and you don't want to see the iSeries Division
go out of business (and I *DO NOT*), then this is going to present a problem
to IBM.  So I've been giving it a fair bit of thought.


I wrote IBM, a fair bit back, that I thought they could sell the hardware at
commodity prices, and make up the difference in service and support.

This may smell like the "Open" Source movement, but is subtly different.
The hardware/software/support is coming from a single vendor, which has been
delivering same for decades.  Plus, that satisfies the two diametrically
opposed needs of the customer-base:
a) Some prefer to pay a known amount, and are willing to pay a premium for
getting good service from a known vendor.
b) Some prefer to "go it on the cheap" and by all accounts pay through the
nose in hidden personnel costs, over an extended time.  (I'm convinced IBM
can make a killing in the Linux market, by pursuing these customers
aggressively.  JMHO.)

However, there was one facet of this idea that I didn't like.  It's unknown
that IBM /*can*/ provide good service and support to the potentially massive
number of small clients, that may decide to go for option a) above.  I've
seen progress, in that regard, where IBM appears to be cutting deals to get
BPs to service these customers.  But while I have no questions about IBM
service, I've had serious questions about whether they CAN provide
documentation and support, to justify those additional costs.  These costs
for service/support would have to rachet up A LOT, to cover the margin IBM
would lose if they sold ALL 400s at commodity prices.  Not to mention that
the margin isn't anywhere near as good, in service/support, to begin with.
Very labor intensive, compared to the interactive "tax".

So this discussion today, about a customer running a B20 for the past 15 to
20 years, prompted me to think of an ADDITIONAL way for IBM to meet the
diverse needs of the 400 customer-base.  This ties in with IBM's offering
COD (Capacity on Demand) where they ship extra processors which sit, unused,
until you need them.  I think this also ties into the ridiculous sums of R&D
money it takes to design and build a chip and it's OS.  Finally I see a tie
to how they build software emulators for computer circuitry (as described in
the building of the MV8000, decade or two ago, in _Soul of a New Machine_ by
Tracy Kidder).

I've thought for a while to see the best way to explain how these tie in...
This is the hard part, for me...  It's getting late, though, and I'm not up
to the task.  It boils down to the idea that I don't recall a HUGE pushback
when IBM implemented the software subscription service.  Most folks
understood it as a way to even out the budget using a fixed amount, and that
it was necessary for IBM to get funds to develop future OS releases.  IBM
sure hasn't disappointed the customers with the rapid enhancements to the OS
and languages (other than a few major glaring problems, which I'll not go
into here.)  So I'll just throw the idea out, and see if there's any
responses:

0)  Interactive "feature":  fogeddaboudit...!
1)  Charge more for services/support, to those that want these features.
2)  Sell ALL 400s at commodity prices, because the iSeries hardware CAN
compete against NT server farms and *nix...
3)  Unbundle the OS, as a separate billable, so it can be compared to other
platform's system software costs
4)  Continue to fold "univeral" functionality into the base OS (and I've
already suggested Sametime as one of these).
5)  Find the "sweet spot" for COD (they may have already done this, haven't
heard how widely-used it is).
6)  Take advantage of OS/400s inherent ability to make better use of
multi-processor systems (compared to NT, don't know about **ix).
7)  Use chip software-emulators to allow new OS to run on OLD processors (to
amortize R&D over longer time).
8)  Throw additional older-model processors into low-end systems (could
vastly improve low-end price/performance).
9)  (Get accurate bills out to customers.  Take Al Mac's suggestion to
contract the thing out, if necessary.)
10) The iNation will never reach it's potential, as long as there is a
perceived "fairness issue".  So eliminate this issue and use it to promote
the iNation.  (Maybe even make some or all of these changes exclusive to
iCitizens...?!?)


Last, but not least (I hope) is the thing that prompted this post:  Charging
customers two different prices, is (IMV, and many others) COMPLETELY UNFAIR
as long as it's based on which programming technique you use to draw
screens.  Makes little sense to base price on this, and this alone.

Far better, IMNSHO, to charge them two prices based on their needs.  As I've
stated, I believe there are plenty of customers who CURRENTLY place a
premium on the VALUE proposition of the 38/400/i platform.  That, IMV, is a
billable item.  So the trick is to identify those things that enhance the
value proposition, and charge for them rather than punishing people for
using a programming technique (regardless of whether it's a superior or
inferior technique).

One of the easiest items of the value proposition to identify is the ability
to write software once, and run it forever.  Being able to run it on all the
platforms is OK, but the value in having software NOT GO OBSOLETE has never
been calculated.  The money spent on Y2K should put it into perspective,
however.  That has always been a feature of CPF, so (IMHO only) should not
be considered a billable item.

But being able to run a piece of hardware is no longer a given, in today's
industry.  That point is driven home by the fact that IBM has decided they
can no longer spare the expense of allowing a given hardware model to run
for 10, 15, or 20 years.  This is completely contrary to what the rest of
the industry is doing, and doesn't play well into the idea the IBM is going
to HAVE to compete in a commodity market.  (BTW, IBM, it's competitors, and
ALL customers in the industry determine the speed with which the server
market becomes a commodity..  Not just IBM, nor just the iSeries Division.)

I think this would be a GREAT billable option.  Those folks (like those that
lease cars) that want the latest and greatest technology every year or so,
and can afford to "throw the old out" (like the PC commodity markets) will
be glad to churn hardware sales for IBM.  Those folks (like those that
purchase cars and drive them into the ground) that want to be able to milk a
computer for 5 or 10 years, should be able to pay for that privilege.  At
the same time, I wouldn't think it would be a WHOLE LOT extra, because IBM
gets to amortize the R&D on the chips over a more reasonable time frame (and
they get upgrade and service/support revenues, as well).

The reason I like this last idea is that it provides an additional,
fundamental approach to deriving revenues, besides the idea of increasing
service/support prices.  I never did like the idea of putting all the eggs
in one basket.  So IMNSHO, this second fundamental approach is extremely
useful because it can be balanced against that one.


As always, I welcome any and all comments, but expect none...



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Now, I *really* don't know why I think I have ideas that would benefit BOTH
the iSeries Community and IBM...  It just seems like at least some of these
would.  But, IMHO, regardless of whether IBM ever uses this idea, or not, I
think they'll probably get wind of it.

And I think IBM should compensate me for it.

Of course, as IBM legal has pointed out..  IBM has NEVER solicited any of
these ideas, so it makes little sense to them to pay for any of this.  I
tend to agree.

...Well.. I agree completely...  I don't believe in the business model where
some guy paints your address number on the street curb, and THEN tells you
he's collecting money for doing it.

That's pretty much the philosophy of "Open" source:  Most of the principles
have laughed all the way to the bank, based **primarily** on the blood,
sweat, and tears of folks that make less than the poorest migrant worker.
(I'm not at all sure why labor laws don't apply in this situation.)  Sure,
it's a "non-commercial" activity, because the software is given away for
free (heheh).  But, **just by coincidence**, Linus, ESR (and I imagine Bruce
Perens and more than a few others) have made some pretty good paychecks, in
their present careers in big business.  Probably just a coincidence,
though...


So, NO...  I don't intend to apply that same principle, and haven't.
(Granted, may have been corn-fused a time or two, about the issue...;-)

I'm just gonna ask a little favor:  which is I'd like permission to write
Mr. Gerstner a letter, on a non-iSeries related issue or two.  I understand
that's presumming a lot...  Makes no sense to me, even now, why I've written
them in the past, or these posts to this list either, for that matter.  When
I think about it too much, I think:  why should I write anything, at
all...?!?  (Well... I point the finger of blame DIRECTLY AT Chuck Lewis...!
Fool invited me to join the AAG, just because I wrote a Letter to the Editor
which got published.  Everything is, therefore, Chuck Lewis' fault, right...
ROFLMAO...!)

Anyhoo, just throwing the question out, just like I threw the ideas out...
HOPE it's true:  it never hurts to ask...

G'nite...

jt
SiliCow Valley
(Columbus, OH, USA)


"Have a GREAT day...!  And a BETTER ONE TOMORROW~~~:-)" (sm)