Tuesday,
4 March 2003 8:30 am
For someone like me who has been
tracking for years how much high tech companies spend on
R&D, the spreadsheet
recently published by Mike Tarsala at CBSMarketwatch
was fascinating. It collected in one place the 2001 and 2002
R&D spending as a percentage of revenue of the 100 largest
US technology companies, exhibiting both interesting patterns
and enormous variances between companies.
The winner for
the most R&D spending in 2002 as a percentage of revenue among
the 100 companies in the survey is Applied Micro Circuits, who
spent 103% of their revenue on R&D! On the other extreme, Dell takes the (perhaps dubious) honors
for spending a mere 1.5% of their revenue on R&D. No other company in the survey came close to these extremes.
Looking just at
the computer and computer peripheral makers, there appears to be a
distinct cluster of companies that believe 6-7% of revenue is the
right spending level for R&D. Another cluster believes
R&D spending should be around 14% of revenue. Then there are
two outliers (Dell and Storage Tech) who are going there own way,
obviously believing their competitors have it wrong.
The printer companies are remarkably similar in their
spending patterns, with Lexmark spending 5.6%, Xerox 5.8%, and H-P
5.8% (you would almost think they are copying each other…).
H-P, in addition to being a printer company is, of course, also a computer company. Its
computer company archrival
IBM coincidentally spends almost exactly the same percentage of
its revenue (5.9%) as H-P does.
On the other hand, we already noted that Dell has a
completely different spending pattern (1.5%), while competitors
Apple spend 7.8%, Silicon Graphics 13.2%, EMC 14.4%, Sun and
Network Appliances 14.7%, and Storage Tech 16.9%. No
computer company in the survey dared spend more than 16.9%.
The
communications equipment makers show a markedly different pattern.
Apparently, in that industry, nobody dares spend less than 8.9% of
revenue on R&D. Scientific Atlanta and Avaya spend the least
at 8.9% and 9.3% respectively.
Ciena spends the most: a full 66.5% of its revenue goes to
R&D! The rest are
fairly uniformly spread between 14% and 25%.
The
R&D spending by software companies ranges widely, with Siebel
spending the least at 11.9%, and BMC spending the most at 37% of
revenue. There is one clear
cluster of companies that spend in the 14.9% - 16% range, including
Novell (14.9%), Intuit (15%), Microsoft (15.2%), Symantec (15.3%),
Network Associates (15.8%).
Chipmakers spend
a significant amount of their revenue on R&D.
The lowest is Intel, at 15%, while the highest is Vitesse,
at 66.4%. The rest
are fairly evenly spread between the two extremes, with a cluster
evident around 30%-32% that includes AMD (30.3%), Agere (31.8%),
and Cypress (32.2%).
What conclusions
can we draw from this?
 |
In the
computer industry, innovation has moved upstream in the
value-add chain, meaning that new technology is injected by
the chipmakers, board makers, and software companies, rather
than the computer companies. The latter, as
most clearly illustrated by Dell, have become final assembly
and distribution
companies (see my book
for more on this), with relatively low R&D needs. |
|
In the
communications industry, there is less evidence of the degree
of vertical disintegration that we see in the computer
industry. However, the recent spin off of Lucent's
chip-making division (Agere), and the attempts by both Intel
and Microsoft to become the key
suppliers of chips and software for the cell phone industry,
illustrate that the industry is changing and may in the future
function more like the computer industry. |
|
Some
companies (e.g. Ciena, Applied Micro Circuits) are investing
heavily in R&D on the bet that it will drive significantly
higher revenue in the future, while most other companies are
adjusting their R&D spending downward in order to survive
with their current level of reduced revenue. |
|
No matter
where your spending falls in these ranges, you need to have a
clear theory as to why it makes more sense than what your
competitors are doing. Also, you need to know what evidence
would disprove your theory, and know what you will do if such
evidence presents itself. |
|
The companies
at the extremes stand to gain the most competitive advantage if their
strategy turns out to be right. |
Making Sense of It All