To All: Revised FY95-96 revenue analysis
I've revised my numbers because I found another
constraint. I now believe revenue of $160/drive
and ~$11.30/disk.I also believe FY96 revenue for
zip will be ~$1.3B asuming ~5M drives sold in FY96.
The quarterly breakdown is as follows:
Q1 $162M
Q2 $267M
Q3 $380M
Q4 $507M
These are interesting numbers no?
-- Fernando
p.s. mgsam: great that you're going to go through the numbers.
Two heads are better than one! Sending the analysis to
Chiros or anyone is fine by me. This is a public forum
after all!
------------------------------------------------
First, I assume you have a copy of my first analysis
in front of you. Otherwise, the following discussion
won't make much sense.
Here's what's different in my revised analysis.
Although I was very pleased with myself after the
last analysis I did continue to look it over for
inconsistencies. This led me to two more pieces of
information which I used to pin down the numbers
even more precisely (I hope!): If you examine the FY95
revenue estimates from my first analysis, and you add up
my estimated total Zip revenue for FY95 you get
$133M. Now K. Edwards has told us that total FY95
revenues for Zip+Jaz were $174M. So I have to
conclude that Jaz revenues for FY95 must have been
$41M. Jaz was introduced in December so this is a
one month revenue figure! Hmmm...
Now, K. Edwards also told us that Zip+Jaz in Q1 FY96
was $185M and, as I argued in my first analysis, Jaz
sales are probably 10%-15% of this. 15% of $185M is
$27M. So, my analysis is telling me that the first
month of Jaz sales was $41M while the following three
months combined were AT MOST $27M. Clearly something's
out of kilter!
Well, I can use what I know about Jaz to further
constrain the shape of the cummulative unit zip drive
sales curve. Here's the way I use the Jaz information.
1) I require that the total FY95 Jaz revenue be about
1/3 of Q1 FY96 Jaz revenue. This is because there
were 3 months of Jaz sales in Q1 FY96 and there was
only 1 month of Jaz sales in Q4 of FY95. Now this
doesn't account for the fact that Jaz sales are
probably ramping up, nor does it account for any
seasonal effects in December sales. I'll examine whether
ignoring these effects is OK at the end of the analysis.
2) So now that I have another constraint, I again diddle
the sales numbers and get yet another smooth curve that
approximately satisfies all the constraints. (For
those of you who aren't familiar with this kind of
analysis, I'm performing what is called a nonlinear
optimization with the monthly cummilative sales
numbers as the free parameters...Rocket Science!)
Anyway, the results of my optimization are given below. The first
thing that happens is that I get a new relationship between
drive revenue and disk revenue, here's selected entries
from the new table:
Table 1: disk revenue vs. drive revenue
drive($) disk($)
200 7.19
180 9.25
160 11.30
125 14.90
110 16.44
100 17.46
90 18.49
80 19.52
I think you retail types might find these numbers a little more
palatable than my first set of numbers. I now believe $160/drive and
$11.30/disk are quite reasonable
guesses.I should mention pretty much any entry in the table
between 180 and 100 (drive price) produces the same answers (to within a
few percent) in the final results. Anyway, if I perform the optimization
(i.e. diddle the numbers in the first column in the table below, while
attempting to satisfy all the constraints) I get the following
revised results.
Table 2: Estimated unit sales and revenues of zip products
(All numbers in millions)
zips zips disks zips disks sales
cum. /mo. /mo. /Qtr /Qtr /Qtr
($)
0.000 0.000 0.00
0.000 0.000 0.00
0.036 0.006 0.04 0.006 0.04 1.45
0.045 0.009 0.07
0.065 0.020 0.16
0.091 0.026 0.23 0.055 0.46 14.05
0.131 0.040 0.35
0.191 0.060 0.52
0.276 0.085 0.75 0.185 1.62 47.95
0.376 0.100 0.92
0.501 0.125 1.16
0.656 0.155 1.45 0.380 3.53 100.74
0.826 0.170 1.65
1.021 0.195 1.91
1.256 0.235 2.29 0.600 5.84 162.00
1.544 0.288 2.79
1.877 0.333 3.28
2.242 0.365 3.67 0.986 9.74 267.87
2.655 0.413 4.18
3.114 0.459 4.70
3.624 0.510 5.26 1.382 14.14 380.89
4.177 0.553 5.79
4.780 0.603 6.37
5.440 0.660 7.02 1.816 19.19 507.38
This gives me $9.8M of Jaz sales in Q4 FY95 and $23M of Jaz sales
in Q1 FY96. The ratio is not quite 1/3, but I find the total quarterly
sales don't change very much from here even if I get down to 1/4.So the
the analysis is not very sensitive to seasonal effects or Jaz ramp-up
effects. Also, let me stress again, since I calibrated off Q1 FY96,
the results aren't very sensitive to the particular choice I make
concerning revenue per drive and revenue per disk, provided I take
the numbers somewhere from the center of Table 1.
Second point, my eye-ball optimization is sufficient to learn quite a
bit, about zip. If this were an "industrial strength" analysis I'd be using
real numerical optimization techniques. But hey, I do enough of that in
my day job, this is for fun (and profit!) |
