Subject: Re: Free *Network* Software Business?
From: "Michael R. Bernstein" <michael@fandomhome.com>
Date: Wed, 03 Dec 2008 09:11:38 -0700
Wed, 03 Dec 2008 09:11:38 -0700

On Tue, 2008-12-02 at 21:58 -0800, Ben Tilly wrote:
> On Tue, Dec 2, 2008 at 9:04 PM, Michael R. Bernstein
> <michael@fandomhome.com> wrote:
> >
> > On Tue, 2008-12-02 at 18:15 -0800, Ben Tilly wrote:
> >> On Tue, Dec 2, 2008 at 5:52 PM, Michael R. Bernstein
> >
> >>         Am I being *too* coy? I was hoping for some feedback by now...
> >>
> >> My reason for not responding is that I don't see how you'd monetize a
> >> free data format.
> >
> > Generally speaking, I expect the free availability of the format, and
> > the availability of free/open data (or content) in that format, to
> > actually create a market for proprietary/licensed data in the free
> > format, as well as services related to aggregating demand and supply
> > (free fonts and clip art help create demand for higher quality
> > offerings).
> 
> It sounds like you know your business model then. :-D

Well, I *think* so, but the subsequent discussion on business models
here has been pretty valuable.

> In this case your goal is to get your format widely adopted.  For
> which you might want to use the BSD license.  You should also aim to
> get your product incorporated into existing products - for instance
> make a Firefox plugin yourself.

That's good advice. To the extent that other software supports similar
existing services, I'll want to make sure my service is supported as
well (not by a FF extension, but there are relevant equivalents).

> > There may be some service and consulting revenue from companies that
> > want to deploy the service internally, but I am actually not sure how
> > soon any of that would materialize (could be a few years). [snip]
> Well with the model you're aiming for, you want to make deploying as
> fast and painless as possible.  Which indicates that you won't get
> much service and consulting revenue.

Hmm. True. Or at least not much beyond the 'personal consulting
business' level. Which could still be helpful in the short term.

> [...]
> >> Depending on what you're doing, other models may work.  For instance
> >> Sleepycat built a pretty good business around offering a free product
> >> which you had to pay to incorporate into proprietary products.  And
> >> for many years Aladdin made money from Ghostscript by having a
> >> proprietary product that they would open source old versions of.  Both
> >> of those models require that you maintain ownership of the copyright
> >> though.
> >
> > The Sleepycat model could work for the webservice, but the GhostScript
> > model won't. Perhaps for libraries intended to be used by desktop apps
> > though... Hmm... No, probably not.
> 
> If your goal is adoption, you may not want to put up artificial
> barriers to inclusion in proprietary products.  Or perhaps you do.  It
> is hard to say.

It *is* hard to say. In any case, there is a distinction to be made here
between adoption of the service software and adoption of the file
format.

Adobe created a pretty hefty revenue stream off of Type1 fonts by
licensing the format, but eventually their stranglehold drove Apple to
creating TrueType as a functional equivalent.

I don't think I could pull off what Adobe did in the first place. It
would be too easy to create a functional equivalent format that was more
open and easier to deal with existing tools, so I think the format I
devise is going to have to be as easy to deal with and adopt as
possible.

That said, there *may* be a sleepycat-like model in licensing libraries
specifically intended for the format.

> The dynamic you've got to think about is whether the potential market
> will grow too much faster than your ability to supply it.  If it does,
> you could create a huge market but never benefit much from it.  For
> instance if Microsoft likes your idea, then incorporates it into a
> hundred million desktops in a year, everyone is going to look to
> Microsoft for direction and you'll be forgotten.

That is a distinct possibility.

> If you haven't run a company before then you probably should find a
> good mentor.  Also read up on startups.  Paul Graham has written a
> number of potentially applicable essays.

A good mentor is hard to find. So is a co-founder. I've got some feelers
out. I haven't run a company before, but I've been reading up on
startups (including PG) for years.

This is the first time I've had an idea that feels this big and this
obvious, as opposed to "yeah, I could make some money doing that" ideas.

- Michael


["application/pgp-signature" not shown]