Choose Your Competition
The bubble is gone and
sanity reigns, but entrepreneurship continues to attract. People love to come up
with new business ideas. I often hear from someone who wants my opinion on
their business plan. I enjoy brainstorming new ideas and evaluating their
One of the first steps in evaluating a new business idea is to find out who
else is already doing it. Do some Google searches. Find out who your competition
will be. Find people who are doing something similar or related.
The next step is where a lot of would-be entrepreneurs make a wrong turn: The
strong tendency is to drop an idea as soon as we find somebody else is already
doing it. We somehow convince ourselves that we have to keep searching for ideas
until we find something completely new. Innovation is glamorous. We
love to hear stories about the guy who makes a fortune by inventing something
completely new. And besides, there's no sense starting with established
competition from day one, right?
To be fair, I'll concede that competition should not be taken lightly.
Entrenched competition can be tough to beat. Picking the wrong competitor as
your goliath can be suicide. However, avoiding competition altogether is usually
not a good strategy for getting a business going.
The big problem with avoiding competition is that you are also avoiding
customers. The existence of a competitor indicates the existence of paying
customers. If you can't find anyone who is making money with your idea, you
really need to wonder if there is any money to be made there at all.
As an example, suppose you have a cool new business idea. You want
to revolutionize parking and car rentals at airports. The idea is simple:
Airports are visited by two kinds of people: Some people bring their car and
leave by plane, so they pay to leave their car in a parking lot. Other people
arrive by plane and need a car, so they pay to rent one from a car rental
agency. Why not match these people up? Instead of paying to park their car,
travelers can actually make money by allowing it be rented while they're gone on
their trip. From the perspective of the renter, everything is the same as it is
for existing car rental agencies, except that we can charge really low rates
since we don't have to carry the capital costs of owning a fleet of vehicles.
As far as I can tell, nobody is doing this business model right now. If you
want to get into a new market which is wide open and free of competition, jump
on this exciting opportunity today. You don't even have to give me a piece of
the profits for contributing the idea. More specifically, I have no interest in
sharing a piece of the losses. :-)
There's a very good reason nobody is running this kind of business: Most
people do not want a stranger driving their car. Many people just don't
treat rental cars very nicely. The owner of the car is not likely to think
the risk and aggravation are worth the trouble for the money they'll get.
If we were still in the bubble, you could spray a coat of Internet on this
idea and find a VC
clueless enough to fund it. But good investors and good press are just not going
to be enough to make this idea work.
The lesson here is that "new" ideas aren't as valuable as people think. Most
of the time, when you find a market with no players, it's not really a
market. Money is made by beating competition, not by avoiding it. If you
want to start a new business, don't look for an idea which has never been tried.
Instead look for someone who is serving real customers but not doing it very
well. Find a way to do it better.
This approach seems scary, but that's merely because the risks of facing a
competitor are easier to see. If you're going to fail competing in an existing
market, you probably know how that failure is going to happen. In
contrast, the risks of creating a new market from scratch are far less obvious
and visible. But even though those risks are harder to see, they're actually
huge. An optimistic person can too easily convince himself that those
risks aren't really there. Believe me, creating a new market segment
is much harder than it looks.
Ask the folks at Segway. There is no
question that they have a completely new idea. Their product is without
competition. It's incredibly innovative. The press loves it. Their VC investors
love it. But reality has set
in , and people are now discovering that we spoiled Americans are not eager
to give up our cars.
It's important to clarify that we're talking about probabilities, not
guarantees. Business is hard, and opportunities for failure are easy to find.
Some people actually do succeed in creating a new market from scratch.
Lots of people do fail by confronting an established player. I still
think the odds of success are better when you're building something that clearly
has a market.
Your odds can be improved even more if you choose your competitor wisely.
How to Choose the Right
I'm a big fan of Jim Barksdale's philosophy for choosing competition.
Barksdale was the CEO of Netscape. He used to say that the best approach is to
find a competitor who is "big and dumb". Back when he was throwing this pithy
sound bite around, Netscape was saying that its competition was Lotus Notes,
which Barksdale claimed was clearly big and obviously dumb. Before coming to
Netscape, Barksdale ran FedEx, competing with the biggest and dumbest of them
all, the US Postal Service. :-)
Despite his obvious lack of tact, Barksdale's advice has merit. Targeting
so-called "dumb" competitors is obvious, since nobody really wants to compete
with smart people. Tackling a big competitor is the best way to make sure your
product or service has a real market to be pursued.
In one sense, we followed this philosophy when we decided to build SourceGear Vault. In this case, our
"big and dumb" competitor is SourceSafe. At first glance, this
may seem ridiculous, since SourceSafe is owned by Microsoft. Clearly, Microsoft
is quadrant three, big and smart, the worst possible combination to be found in
a competitor. However, Microsoft is big enough to be very multi-faceted.
For the products that are most important to them, Microsoft is very big and very
smart, but SourceSafe doesn't seem to be one of their top priorities.
SourceSafe's owner might be quadrant three, but this product effectively
functions in quadrant one. (In borrowing Barksdale's intentionally humorous
terminology I mean no offense to SourceSafe fans. But we're very familiar
with SourceSafe and its various problems. In fact, our SourceOffSite product continues
to sell very well as an add-on which addresses one of the big
problems with SourceSafe.)
While you're choosing your competitors, choose your partners too. If you can
get them, your best partners come from quadrant three. The "big and smart"
companies have a lot more resources than you do. The companies in the
"small and smart" quadrant can be good choices too.
The "small and dumb" quadrant is useless. Don't bother selecting these
companies as competitors or partners. Nobody survives very long in that quadrant
We chose a strategy which would make us SourceSafe's enemy but Microsoft's
friend. Vault is built entirely on the Microsoft .NET Framework, and it uses SQL
Server 2000 for all its repository storage. In fact, I would argue that
Vault deserves to be a case study for successful use of .NET by an
ISV. After completing a full release cycle we can confidently
say that the .NET Framework delivered on its promises. Using C#
instead of a lower level language saved us a lot of time. Visual
Studio .NET is the first IDE that I like better than emacs. Anyway,
Microsoft has been supportive of our efforts on Vault, sometimes even
enthusiastic. They evidently care about .NET and SQL a lot more
than they care about SourceSafe.
The resulting scenario is perfect for SourceGear. Our competitor is quadrant
one. Our primary partner is quadrant three. It is merely a distracting
coincidence that those two entities are technically one and the
For the sake of completeness, I should point out that this strategy
can have undesirable side effects. Every good market has more than one
competitor. When you select your primary competitor from quadrant one of
your market segment, you're going to accidentally end up with some competitors
from the other quadrants as well. That's okay as long as your market segment has
fragmentation and your product has differentiation. Those two
words have a grand total of ten syllables between them, but they're worth the
- Fragmentation: This means the market is shared by a
reasonably large number of players, all of whom are functioning
profitably. You typically don't want to fly kamikaze into a competitor
with 90% market share. For example, the market for desktop operating
systems is not fragmented. In contrast, the market for embedded
operating systems is highly fragmented.
- Differentiation: In at least one important way, be
different. Make sure that one particular niche of your market segment
has a very good reason to favor your product over the more established
competitor. Make the people in this niche love you. Until they do,
you can ignore the rest of your market segment.
Revisiting SourceGear's tactical situation, SourceSafe is quadrant one, but
there are lots of other players in this market segment. For example, Perforce looks like it is clearly in the
"small and smart" quadrant of the matrix. We definitely don't want to be
intentionally choosing our competitors from quadrant four. Luckily there
is enough fragmentation in the version control market that this won't be a
problem. The failure of Perforce is not in the critical path to our success. (In
fact, if Perforce failed that would bad for SourceGear. It would
indicate either a declining market or a scary predator that's coming after us
next.) There is plenty of room in this market segment for both of us,
especially since our two products are so different from each other. Fragmented
markets allow for that kind of thing.