Subject: Donation models
From: "Benjamin J. Tilly " <>
Date: Wed, 03 Jul 2002 03:34:27 +0800

Given the discussion of how the recent K5 pledge drive succeeded, the
problems that the much larger Perl 6 drive got into are worth mentioning.
Both to show that just asking for donations is not always going to work,
and for getting an estimate of what donation drives can realistically

As many of you know, Perl is undergoing a major redesign/rewrite.  As
part of this, several key people were to be funded with grants to be
funded by community donations.  Quoting possibly incorrect figures from
memory, the goal was about $200,000.  They have so far managed a bit
over $100,000, and as a result they are running out of money.  Damian
Conway's grant has already been ended. Dan Sugalski's will end at the
end of this month (unless they get about $60,000, fast).  Larry Wall
will continue to be funded, and they are planning to spend some of the
money on a professional grant writer to see if they can raise more

Is this a success or failure?  It depends on your perspective.  In a bad
economy which has hit a lot of people very hard, they raised quite a bit
of money.  The people that they hired for that did absolutely amazing
work.  They didn't raise as much as they wanted, but they raised enough
that one full-time person, perhaps two, could have been supported for
the year.

Could they have reasonably done better, even in the current economy?
Contrary to the popular saying, I don't think that hindsight is 20/20.
Rather it is speculation with the comfort of knowing that events can't
prove you wrong.  Grasping at that straw of comfort, allow me to
speculate for a moment.

Two major points strike me.  First of all they spent a lot of energy
aiming for a big corporate sponsorship.  Unfortunately for reasons
discussed already on this list, open source business models tend to have
thin margins, and thin margins don't create great sources of corporate
sponsorship for worthy projects.  Secondly it seems to me that they
didn't properly capitalize (quite literally!) on the community
generosity they found.

A quick story may demonstrate the latter.  Imagine that you have donated
a fair amount already, you are sitting at a talk where you intend to
donate some more, you know they need more donations, but the speaker
looks guilty while saying to the audience, "I am not asking you for more
donations here.  The community support has been amazing, but you have
given more than you can afford already..."  I was in that audience, and
that night the $100 I planned to give didn't turn into the $200 in my
wallet, nor did anyone else donate anything.  Assuming that that event
wasn't isolated (and I have reason to believe it was not), they could
have definitely raised more.

I believe that pledge models to funding open source projects can work.
However I think that the following points are important.

1. The person/project that this is for must be seen as special.  Perl 6
   qualifies.  Rusty also proved himself.  But the common reaction to
   most people who say that they want donations so that they can do
   open source coding will be, "Why can't you find a job?"
2. While corporate sponsorship is nice, it is probably not going to be
   sufficient.  There are a lot of worthy causes chasing those dollars,
   and getting yours heard may take some serious work.  (I suspect that
   the professional grant writer for Perl 6 is a good move.)
3. A common feeling is, "I like the idea that I've earned the right to
   use the free software I've been using."  (Stolen from  If you are busy (who
   isn't?), then donating to a worthy cause is a really easy way to get
   there.  Charity should take full advantage of that feeling.
4. Make sure that everyone involved with the project understands the
   basics of soliciting donations.  For instance people only freely
   money that they can afford.  Respect that, and don't be shy about
   reminders that if they can afford still more, you can still use it.
5. I think it would be a good idea to study how universities solicit
   alumni donations, and see if there is anything that can be learned

All probably obvious except the last.  So what do I mean by it?  Well
one thing that struck me when I moved from Canada to the US is how much
more developed the US culture of alumni donations is.  How big a
difference is it?  Well let me compare the two I was at.  My undergrad,
UVic, has about 18,000 students at any given time.   Alumni donations
for 2000 were $5.56 million (Canadian) - about $3.7 million US.
Dartmouth College has about 5500 students.  The comparable donation
figure is $23 million.  That is about a factor of 20 difference per
student.  Granted, Dartmouth College has some very rich alumni.  But
they also put out more effort into creating and sustaining a culture
of generous alumni donations, and it shows.  It seems that Canadian
schools have been studying the US example, and donations seem to be
rising there as well.  (And if even those Canadians can learn...:-)

Personally I don't see why successful open source projects like Perl
should not eventually expect order of magnitude donation levels that are
similar to what universities have.  Probably not what US universities
set, but still.  Certainly I get similar opportunities from both.
Better opportunities for future employment?  Local networking
opportunities?  Promises of future improvements?  Good talks coming to
town?  Conferences?  Educational luxury cruises?  Ways to keep in
contact with old friends?  I get all of that from UVic, Dartmouth, and
Perl.  Frankly Perl's versions are more interesting to me, and I take
advantage of more of their offers.

Granted, for a lot of this I should say, "Thanks, Tim."  Except I don't
need to because I thanked him with a bookshelf full of animals
already. :-)

So what kind of figures are we talking about?  Well UVic manages to keep
track of about 57,000 alumni, and manages to average $65 (US) per person
per year in donations.  Judging from, say, the popularity of Perlmonks,
in a few years YAS could well be soliciting 10,000 active users.  If
they can get up to $50/person, less the costs of soliciting those users,
the result will be several full-time developers.  (Chosen, of course,
for already outstanding contributions.)

Perl is well-established and unusually well-organized, but still has a
good distance to go to that goal.  However I think that serves as a
reasonable benchmark of what is realistically plausible from trying to
fund development costs through a donation model.  Tim O'Reilly is
probably in the best position here to estimate what it takes in time and
money to get there.  And, of course, the community contributions are
several orders of magnitude larger than just what is done by people who
are hired by donations...


Download the free Opera browser at

Free OperaMail at

Powered by Outblaze