Good thoughts.
"May I ask whether you have much experience as a
sole proprietor or solo founder?"
Some!
"One reason a technology startup is significantly
more stressful than a standard business is that, by
definition, you are in uncharted territory."
Yes, there's some "stress". My view is that making
things work "in uncharted territory" is just what
has to be an expected part of much of work in being
successful.
Generally, for everything we have that did not exist
40,000 years ago, there was someone who did it for
the first time and was "in uncharted territory".
"Uncharted territory" does not have to be just a
shot in the dark or wildly risky; instead, in a good
case can have some good evidence quite early on that
are on a promising path. If not, then pick a new
approach to a solution of the problem or pick a new
problem. Problem selection is important.
You are partly right about a guy with, say, an auto
repair shop but in information technology, for an
important problem, market, etc., having something
original, much better than anything else and
difficult to duplicate or equal, can be an
advantage.
"With most tech startups, you have to invent a new
business model, create new technology from scratch,
and convince customers that your never-before-seen
product or service is something they should spend
their hard-earned money on."
Look at it this way: Start with a problem it is
clear the market, users, customers, etc. really have
and really care about, and then give them the first
good or a much better solution.
So, all are trying to "sell" the customer on is your
solution for the problem he already cares about.
"Though there are a few general patterns available
for web startups to use (e.g. freemium,
ad-supported), every startup has to figure out how
to become profitable within its particular niche."
For Web sites, sell ads. From the Meeker data at
KPCB, the average 'charge per 1000 ads displayed'
(CPM) of mobile and desktop is $2.125. Then if can
send, on average 24 x 7, 2 Web pages a second with 4
ads per page, should be able to get monthly revenue
of
2.125 * 4 * 2 * 3600 * 24 * 30 / 1000 =
44,064
dollars.
Then, for a solo founder, $44,064 a month is a nice
start.
The key is to have the site deliver a 'solution' to
a problem where the solution is good enough to have
users request the 2 pages a second. For that, need
to have something some people want.
How many people? Well, at least how many pages per
month? That's
2 * 3600 * 24 * 30 = 5,184,000
So, if each monthly 'unique' user sees 10 pages,
that is 5,184 monthly unique users. There are maybe
100 million people in the US active on the Internet
and 1 billion in the world.
For sending 2 pages a second, the 'operating costs'
of the business might be just one or two servers
worth less than $2000 each and an Internet
connection that, with TV and phone, costs about $150
a month, that is, next to trivial.
If more users like the Web site, then the revenue
can grow. So, be sure have built a 'scalable' site,
and then add servers and Internet upload bandwidth.
"Writing code and designing user interfaces is
nothing like repairing cars. There's no doubt that
there are a lot of intelligent, knowledgeable
mechanics (even some who are also software
developers), but there's a very large difference
between understanding an existing system, and
creating a new system from nothing but keystrokes."
I don't know! As a teenager, I repaired some cars.
And I've written software, e.g., created "a new
system from nothing but keystrokes". For me, in
most respects, the keystrokes are easier.
When I've created "a new system from nothing but
keystrokes". the challenging part was not the
keystrokes and not the programming but figuring out
what to program. In my case, that's been mostly
applied math.
E.g., once I was working for a software house near
DC. Our customer was a major Navy lab. I had a
week with nothing else to do, and a project was in
trouble, so I was sent to the project for the week.
In about an hour I saw that there was nothing I
could do for the project in a week.
But we were also in some competitive bidding on some
software for that lab. I saw a problem with what
the lab had requested and decided to spend the week
on something risky: I got 'smart' on the statistics
of the measurement of power spectra, wrote some
illustrative software, and at the end of the week
showed a guy from the Navy. The Navy was pleased,
and our company got essentially 'sole source' and
won the competitive bid. If my work that week had
failed, then I might have been fired. I got a
solution; I thought that it was fun.
"Most startup-worthy software projects will involve
intense creative effort at every step of
development, which is far more mentally taxing."
A lot of work in school involves "intense creative
effort at every step"; e.g., one exercise was show
that there are no countably infinite sigma algebras,
and the solution takes some "intense creative effort
at every step". I was the only one in the class who
got it, but it was fun work.
"A tech startup has to convince customers that they
need to spend money on an entirely new category of
product."
Well, can take my approach, "convince"
users/customers that, for a problem they already
have and really know they have but are not really
solving now, they need to use "an entirely new
category of" solution.
"only those who are determined beyond reason (or
already rich) should dream of going it alone".
"Rich"? Need routine cost of living. But for the
software, I paid Microsoft for Windows 2000 and then
for the upgrade from Windows 2000 to Windows XP, but
all the rest of the Microsoft I've been using --
Visual Basic .NET, C/C++, .NET Framework, SQL Server
Express, IIS, etc. -- has been for free. The 4000
pages of documentation from MSDN, also free.
When I go live I will need serious copies of Windows
Server and SQL Server, but those are available via
the Microsoft BizSpark initially for free. Now can
plug together one heck of a computer for $2000, and
can get some really good Internet access, with TV
and phone, for about $150 a month. Don't have to be
"rich" to fund that.
"Alone"? I'm failing to see that involving others
would speed the work or lower the costs. For
speeding the work, nearly all the time has gone to
my learning enough about the relevant Microsoft
software and working around various issues, and as
the CEO I should know that material. The work
really unique to my project has been fast, fun, and
easy.
In a sense, either the solo founder really knows
what the heck he is doing and is on a good path, or
he doesn't and isn't. In the first case, involving
others will likely dilute the quality of the work
and increase the costs, and otherwise not much can
save the project.
I don't know why you are eager to hire others
instead of just doing the work.