The article points to some problems, and there are
some. But the article also has some problems.
Here is an easier explanation:
First, the entrepreneurs are not bringing the
venture partners enough projects with good,
powerful, valuable innovation.
Second, the venture partners do not want to evaluate
innovation and, instead, want to evaluate
'traction'. They want to see traction significant
and growing rapidly. This concentration on traction
may come mostly from demands of the limited
partners. A failing of the article is that it
didn't explain the crucial role of traction.
Third, as in the article, there are some cliches
about 'innovation': It's very expensive in time and
money, very risky, done for its own sake just for
the fun of seeing amazing things, is from dedication
to change the world and not for ROI, is impossible
to evaluate, is largely separated from business and
reality, etc. To save space here, I omit 1-1
pairings of each of these cliches with parts of the
article. Exercise: Do the pairings. Hint: They
all exist.
So, it is easy for the venture partners to be afraid
of innovation, to fear that it is an interest and
direction in conflict with ROI and not an advantage.
With this fear, traction and innovation are worth at
most just the traction alone and otherwise maybe
significantly less.
Fourth, information technology (IT)
entrepreneurship, from both sides of the table, has
a problem with 'patterns': There aren't useful ones
easy to see just from the past 40 years. Why? The
examples of big successes are, say, Square,
PInterest, InstaGram, Twitter, Facebook, Google,
Yahoo, HotMail, Cisco, Oracle, Seagate, HP, Dell,
..., Microsoft, Intel. So, get such a big win maybe
less than once a year; in 10 years get only a few
big wins; then in the next 10 years technology has
changed so much that the empirical patterns from the
past 10 years don't work very well.
Fifth, the cliches in the article about technology
are too pessimistic. E.g., if innovation in
technology is so risky and unpredictable, then how
come it has done so much in the last 40 years? I
mean, we're talking a 3.0 GHz 8 core processor for
less than $200, main memory for $10/GB, a 3 TB hard
disk for $135, 1 GbE on each motherboard, 100 GbE on
one wavelength on a fiber with dozens of
wavelengths, etc. No, there are some reliable
'patterns' as illustrated by the technology
advances.
The pattern? Pick a pair of a problem and a
solution. The problem should be such that a good or
much better solution will get eager users/customers
(want no doubts about 'product/market fit'). For
the crucial solution, innovate, e.g., do some
research in applied mathematics, applied science, or
engineering. How to do such research? There is
education for that, long heavily funded mostly for
US national security. The education is available at
the top two dozen or so US research universities and
is called a Ph.D.
For more on this 'pattern', look at the progress in
IT of the last 40 years, say, from when Intel got
going, along with what the US DoD has done starting
early in WWII to the present.
Risky? Not really: Mostly can eliminate nearly all
the risk with just good technical work just on
paper. And before that paper is written, one of the
lessons usually learned in a good Ph.D. program is
how to pick a suitable problem where 'suitable'
means can get results that are "new, correct, and
significant" with the meager time and other
resources available to a Ph.D. student. For the
needed innovation just want "new, correct,
significant, powerful, and valuable" for the needed
solution within the time and money available.
Expensive? Not always. DoD and NASA projects tend
to be expensive due to some heavy hardware, but not
all innovation is expensive. E.g., RSA.
But generally the crucial work should be done and on
paper in good form before allocating a lot of
resources.
The solution? IT entrepreneurs need to do better
innovation with nearly all the risk removed by work
just on paper (or in PDF). The venture partners, if
they want to invest, need to be willing to evaluate
work on paper and invest at that 'stage'.
But there is now a threat to the venture partners:
Now the computing is so cheap that by the time a
project has the traction that venture partners want
to see, there is a good chance that the project is
already nicely profitable.
E.g., suppose an entrepreneur does some innovation
to get a powerful, valuable solution to a big
problem and brings up a Web site to deliver the
solution on the Internet.
Then look at the Meeker data at KPCB and average her
CPM figures for mobile and desktop and get $2.125.
Assume the Web site gets users enough to send 2 Web
pages a second with 4 ads per page. Then in one
month the revenue would be
2.125 * 4 * 2 * 3600 * 24 * 30 / 1000 = 44,064
dollars. If each Web page sends for 400,000 bits,
the upload bandwidth needed is commonly available in
US living rooms from a cable TV company.
It is likely that a computer with less than $2000 in
parts is sufficient as a server. And can consider
using cloud servers.
So, we're looking at a project with a $2000 chunk of
hardware, an Internet connection, with phone and TV,
for less than $200 a month, one or a few founders,
and $44,064 a month in revenue.
Now why should such a project, sending 2 Web pages a
second and getting $44,064 a month in revenue. call
a venture partner or return a venture partner's call?
Yes, this scenario has not yet resulted in a lot of
big wins, but the threat to venture capital and the
opportunity to entrepreneurs is just sitting there.
Likely one good example is the romantic matchmaking
Web site in Canada, Plenty of Fish, long just one
guy, two old Dell servers, ads just from Google, and
$10 million a year in revenue.
An established pattern? No. Obvious? Yes.
Note that I am saying that the opportunity is to (1)
get a Ph.D. in applied math, applied physical
science, or engineering from a good research
university, (2) pick a pair of an important problem
and a good or much better solution where the
solution can be generated from software running on
servers and delivered to users/customers over the
Internet, (3) for the solution do some innovation,
that is, some research in applied math, applied
physical science, or engineering, (4) write the
software, plug together the server(5), bring up the
Web site, go live, get publicity, users/customers,
ads, revenue, and earnings, and grow.
For the venture partners, they can (1) stay with
their current ROI, significantly less good than an
index fund, (2) look for more mobile, social, local,
sharing apps for teenage girls to gossip and look
for boyfriends, or (3) get serious about innovation
and encouraging and evaluating it and then funding
its exploitation. No, venture partners should not
invest in research, but they likely will need to get
quite serious once the research is solid on paper
and, in particular, drop their scorn, contempt,
ridicule, and phobia about research.
Funding based just on paper? That's how nearly all
large engineering projects, DoD projects, etc. get
done. Net, for what is crucial about innovation,
paper is fine.