SURVEY ON SOLUTIONS TO PRIOR ART SEARCHING
Internet Patent News Service
June 30, 1999
During the week of June 21, 1999, I sent out a survey to readers of
the Internet Patent News Service, asking them what they thought of
software patents. Here is a summary of the results, with many thinking
that too many software patents are invalid, and that the Patent Office
needs help with its activities.
1) Are you a lawyer, software engineer/computer scientist, or other?
I received 45 responses, pretty much evenly split between
lawyers and soft/comp types. The answer to 1) didn't seem
to correlate much with answers to the other questions.
2) What percentage of software patents are invalid because they are either
not novel, decently obvious (up to a few prior art items being combined),
having poorly constructed claims (i.e. an Internet claim mentioning separate
clients and servers), and/or not enabled?
% Invalid Responses
--------- ---------
20-40% 11
50-70% 15
80-99% 10
This averages to about 60%. Having had to satisfy lawyers'
requests (i.e. for a fixed fee, I search until they are happy)
for over 100 invalidity searches, I would put the percentage
at at least 70%, especially given that 80% of issued software
patents effectively cite no non-patent prior art. But
whatever the number greater than 50%, with 20,000 software
patents issuing a year, lots of legal headaches are being
created - licenses to sue, not checks in the bank.
3) How much money is a reasonable amount to have spent (either by the PTO,
applicant, or government) for a decent patentability search for an
electronic patent application?
Search Fee Responses
------------- ---------
< $1000 12
$1000 - $2000 20
> $2000 13
Averaging about $1500, or ten percent of what most pay for
their US software patents to issue (typically ranging $10,000
to $20,000). Software patent reform then reduces to the
problem of allocating this cost to some party in the process.
4) Should the PTO create a corps of searchers with statutory authority
(freeing examiners to focus on application analysis)? (YES/NO)
YES - 37 NO - 8
Interestingly, PTO management doesn't seem very interested
in this suggestion. Europeans are apparently abandoning
this method as well.
5) Do established companies prefer quality over quantity as the basis for
their patent strategy? (QUAL/QUAN)
QUANTITY - 30 QUALITY - 15
What follows are comments from lawyers to question 5):
QUANTITY. This is true for everyone, not just established companies. If
the Saddam Hussein of the software patent world has WMDs (Weapons of Mass
Destruction, a/k/a software patents), so must everyone else, as many as
they can. (There are rumors that the Saddam Hussein of the software patent
world lives either in Redmond, WA or Armonk, NY, but these are only rumors.)
====
Generally, the larger the company, the more they value the Quantity of
patents over the quality of any particular one. Its the old portfolio
effect, and good for impressing licensees.
====
But you are missing the point. Most people want patents to hang on their
wall, show to investors/stockholders and use in marketing. They don't
want to litigate.
====
But really, I think it's a little more complex than a simple binary
decision. Many established tech companies would say they prefer quality,
and many probably think that is what they're doing, but in practice a
quantity emphasis seems to dominate. It's worth looking closely at what the
qual/quan tradeoffs typically come down to, and how decisions are made.
Also, there is something to be said for diversification as a "quality"
strategy because of uncertainty about the future -- just like with financial
investments -- even though it might superficially look like a "quantity over
quality" choice.
====
Most prefer quality but there is a significant minority who prefer quantity.
Of those who prefer quality, they often settle for quantity because of
resource issues.
====
But let me generalize a bit here. I think this preference for quantity over
quality is more a function of who runs the company than how "established"
it is. Most companies, new and established, are run by short-sighted
bean-counters (some call them "MBA's) who know nothing about patents or,
for that matter, anything else except bottom lines. All these hucksters
want to know is how much it cost to get the patents - and the more you get,
regardless of quality, the cheaper they are. So quantity predominates over
quality almost all of the time. The only attention quality gets in these
companies is lip service, and lots of it. Unfortunately it typically takes
at least five, and often at least ten, years before the shoddy work on a
patent is found out, and for about 99 % of patents it is never found out.
After five or ten years, it is likely that the hucksters running the
company when a patent was drafted are gone to run another one and the
patent professionals who were involved have moved on too. This lack of
accountability also reduces incentive to hold quality above quantity. In
the unfortunately rare companies, new or established, that are run by
imaginative, bright people who are smart enough to know something about
intellectual property and take a long-run view of things, quality matters.
The main reason it matters is because these people are smart enough to
count beans too - counting beans is about the most simple-minded thing
in the world to do - and they recognize that over the long-run quality
intellectual property pays many times more than shoddy intellectual
property even when there is a lot more of the shoddy.