ggplot2 premium services survey report
Hadley Wickham
I've attached the results of last weeks survey on premium services,
with a bit more context about why I'm thinking about it. Please use
this email thread for any questions, discussion etc.
Thanks!
Hadley
--
Assistant Professor / Dobelman Family Junior Chair
Department of Statistics / Rice University
http://had.co.nz/
Jean-Olivier Irisson
>
> I've attached the results of last weeks survey on premium services,
> with a bit more context about why I'm thinking about it. Please use
> this email thread for any questions, discussion etc.
From your write up, I conclude that ggplot2 membership and premium support should be two separate things:
- the membership could resemble that of many other scientific societies:
. a yearly fee of $20-30, which people pay to signify their interest in the subject, to support ggplot. If you can reward donators with cheap/free memorabilia, that's even better. Ideas: a t-shirt, a mug, an acknowledgment on the website (for those who want it), the right to add "I am a ggplot supporter" message in their email signature, a badge on their website, a newsletter that you (and/or others) would edit with "the best of ggplot on the web" etc.
. additional fees for more expensive goods, such as a (signed) copy of the book (in other societies, this is typically journal subscriptions)
- the premium support system could be on a subscription or time-billing basis and would work as any other consultancy work would.
You would have to figure out a good business strategy to make this work of course, but something based on increasingly expensive plans such as :
. free : the ggplot mailing list !
. 50$ a year : 15 premium-support emails
. 100$ a year : 30 premium-support emails or skype calls, discount on 1 online course
etc.
is common with web services and easily grasped by most people.
I think these two items should be kept separate because they target different people and would be paid with different sources of money. The membership would probably interest most people on this mailing list and would be paid with personal money. The premium support targets very specific needs (some of which you highlight in the document you sent) and would probably be paid with company money.
That's just the opinion of a future ggplot membership subscriber of course ;)
Jean-Olivier Irisson
---
Observatoire Océanologique
Station Zoologique, B.P. 28, Chemin du Lazaret
06230 Villefranche-sur-Mer
Tel: +33 04 93 76 38 04
Mob: +33 06 21 05 19 90
http://jo.irisson.com/
baptiste auguie
I have some reservations on the premium support idea.
It may be that the way ggplot2's help and development work now --
organically driven by enthusiasts from this list, R-help, SO, github
-- are already close to optimal (a trade-off between how many hours
dedicated but busy people are willing to spend on it, and the quality
of the contributions).
Charging for some select answers would certainly alter that delicate
balance. For one thing, if one pays for an answer, one should expect
an authoritative one. But these can often only come from a handful of
people, and their professional consulting rate would be very high as
you point out.
I don't know what motivates us folks to answer currently; it's
probably a variable mix of reasons, from improving your online
reputation, learning, entertainment, killing time, showing off,
curiosity, urge to teach things, genuine help, etc. It could be argued
that raising money would never work as well as an incentive, and could
also have detrimental effects on the personal drive of some list
members.
Are there examples of open-source communities that have set up a
successful premium membership? Maybe we can learn from their story.
If money can speed up the development of ggplot2, I think the other
options (t-shirts, books, donations) are much less interfering; in
this regard would it not be an option to simply add a line to the
ggplot2 mailing list signature saying that donations are welcome?
Just my 2 cents
Cheers,
baptiste
> --
> You received this message because you are subscribed to the ggplot2 mailing list.
> Please provide a reproducible example: http://gist.github.com/270442
>
> To post: email ggp...@googlegroups.com
> To unsubscribe: email ggplot2+u...@googlegroups.com
> More options: http://groups.google.com/group/ggplot2
Jean-Olivier Irisson
>
> It may be that the way ggplot2's help and development work now --
> organically driven by enthusiasts from this list, R-help, SO, github
> -- are already close to optimal (a trade-off between how many hours
> dedicated but busy people are willing to spend on it, and the quality
> of the contributions).
> Charging for some select answers would certainly alter that delicate
> balance. For one thing, if one pays for an answer, one should expect
> an authoritative one. But these can often only come from a handful of
> people, and their professional consulting rate would be very high as
> you point out.
>
> I don't know what motivates us folks to answer currently; it's
> probably a variable mix of reasons, from improving your online
> reputation, learning, entertainment, killing time, showing off,
> curiosity, urge to teach things, genuine help, etc. It could be argued
> that raising money would never work as well as an incentive, and could
> also have detrimental effects on the personal drive of some list
> members.
I agree that the paid support could discourage those who basically do it for free right now. I personally consider the time spend answering questions as a way to "pay" for this otherwise free bounty that is ggplot (and I don't do it as often as I should given this standard) but it's difficult to know the motivation of everyone indeed.
> Are there examples of open-source communities that have set up a
> successful premium membership? Maybe we can learn from their story.
Ubuntu and Canonical spring to mind, but the size of the communities are not the same. In addition, I think that most canonical employees are not involved in the Ubuntu support forums, so there is already a clear separation between the free, user-powered support and the paid one. This wasn't true for ggplot (and I would bet that a great part of its initial momentum came from the excellent support of Hadley on the mailing lists).
I can't find another example.
Jean-Olivier Irisson
>
> options (t-shirts, books, donations) are much less interfering; in
> this regard would it not be an option to simply add a line to the
> ggplot2 mailing list signature saying that donations are welcome?
I thought of another solution which might warrant examination: advertisement. How many subscribers does the ggplot mailing list have? If the number is significant, Hadley might be in a position to attract targeted advertisers: computational software, commercial stats packages with a specific niche that R does not cover yet, but also, more generally, companies that provide gear/goods/software for research.
With a new survey similar to the one conducted a while ago, Hadley could get an idea of the demography of ggplot subscribers and that would be very valuable info for the marketing departments of some companies. For example, the last survey showed that a large majority of ggplot2 subscribers were biologist so you could go to pharmaceutical companies that sell lab kits or lab gear.
I guess nobody likes ads, but if it helps to support ggplot, I wouldn't mind having a weekly, topical ad sent on the ggplot mailing list (one business model to copy could be that of Daring Fireball for example, for those who know it).
Camp, Mary
# Fig. 1
# correlation matrix for 5 column data.frame
# use theme for faceted plot!
# create data
v1 <- rnorm(100)
v2 <- 10 + .5*v1 + rnorm(100)/2
v3 <- -3 + -7*v1 + 2*rnorm(100)
v4 <- 1 + .5*v1 + rnorm(100)/2
d4 <- data.frame(v1, v2, v3, v4)
library(ggplot2)
theme2 <- theme_update(
panel.grid.minor = theme_blank(),
panel.grid.major = theme_blank(),
panel.background = theme_blank(),
panel.border = theme_rect(linetype="solid"),
axis.line = theme_segment(),
axis.text.x = theme_text (colour="black", hjust = 0.5, vjust = 1),
axis.text.y = theme_text (colour="black", hjust = 1, vjust = 0.5)
)
plotmatrix(d4) + geom_smooth(method="lm")
Mary J. Camp
Biometrician
Biometrical Consulting Service
Beltsville Agricultural Research Center
U.S. Dept of Agriculture
This electronic message contains information generated by the USDA solely for the intended recipients. Any unauthorized interception of this message or the use or disclosure of the information it contains may violate the law and subject the violator to civil or criminal penalties. If you believe you have received this message in error, please notify the sender and delete the email immediately.
Art
kw
http://www.kickstarter.com/projects/1834879952/goats-book-iv-inhuman-resources
My cousin is using this to be able publish another volume in a series
he is writing. Note that he got four times more pledges than he asked
for.
I don't have any statistics on the percentage of people that fulfill
their pledges. I don't know how big his audience is compared to the
ggplot2 audience.
I am also not shilling for my cousin or kickstarter. Of course I want
my cousin to do well! On the other hand I have little or no knowledge
of how kickstarter works and I don't own any percentage in either of
the endeavors.
I hope no-one is offended by this posting.
Take care,
KW
On Mar 20, 5:19 am, Jean-Olivier Irisson <iris...@normalesup.org>
wrote:
Allen S. Rout
> Here is a potential way for you to generate revenue:
> http://www.kickstarter.com/projects/1834879952/goats-book-iv-inhuman-resources
>
> My cousin is using this to be able publish another volume in a series
> he is writing. Note that he got four times more pledges than he asked
> for.
>
> I don't have any statistics on the percentage of people that fulfill
> their pledges.
At pledge time they commit to have Amazon bill them. I think they're
pretty much committed. As long as they make their funding goal.
- Allen S. Rout
Allen S. Rout
> I agree that the paid support could discourage those who basically
> do it for free right now. I personally consider the time spend
> answering questions as a way to "pay" for this otherwise free bounty
> that is ggplot (and I don't do it as often as I should given this
> standard) but it's difficult to know the motivation of everyone
> indeed.
There's significant research on this topic. I recommend 'Predictably
Irrational', by Dan Ariely, as a readable sketch, aimed at the general
public but reasonably detailed.
The bit most relevant to this conversation is "The Cost of Social
Norms -- Why we are happy to do things, but not when we are paid to do
them." The key points might be phrased something like this:
+ It's easy to move an individual from viewing behavior as "social" to
seeing it as participation in a "market".
+ It's difficult to move the other way, from market-encourage behavior
to socially-encouraged; it's also difficult to maintain both
motivations simultaneously.
+ Market behavior outcompetes social behavior in a community.
To quote:
"When a social norm collides with a market norm, the social norm goes
away for a long time. In other words, social relationships are not
easy to reestablish. Once a social norm is trumped by a market norm,
it will rarely return."
I'm not (quite) speaking against paid support; I'm more pointing out
the mental pitfalls to which people are heir, once they start
accepting pay for behavior they were once proud to do for free as a
social duty.
Hadley's manner towards his community is IMO nearly a textbook for
managing such behavior; he pushes credit out to the community. Other
humans who hope to get some regular market reinforcement from ggplot2
ought to follow his example and keep an eye on their attitudes; The
market tendencies are written deep in our genes in many ways.
- Allen S. Rout
Jean-Olivier Irisson
>
> I'm not (quite) speaking against paid support; I'm more pointing out
> the mental pitfalls to which people are heir, once they start
> accepting pay for behavior they were once proud to do for free as a
> social duty.
I'm not sure I fully understand everything in your email but the above paragraphs summarizes what I did understand: people who helped on the mailing list might not continue as joyfully if they are paid to do it. That might be very true but that was not my point. My point was that paying a few people to ensure the premium support might discourage the others which provide this service for free on the regular mailing list.
igreg
<I'm not (quite) speaking against paid support; I'm more pointing out
<the mental pitfalls to which people are heir, once they start
<accepting pay for behavior they were once proud to do for free as a
<social duty.
to contribute to a social community is more valuable than to get paid
for helping others.
I think that a lot of users would like to contribute something in
monetary terms to the development of ggplot.
So why not just give the opportunity to pay a voluntary amount of
money to support the developers? (without return for a premium-
support)?
- greg
Ista Zahn
I totally agree. I would encourage Hadley to put up a paypal or
similar mechanism for donating to fund ggplot2 development. I know
I'll be happy to donate, and I suspect many others will as well.
Best,
Ista
>
> - greg
Raphael Saldanha
I'm very new with ggplot2 and this list, but on those some months I had seen some amount of messages of users who just want to create some nice-looking graphs, and had just landed on the R and ggplot world, having basic primary questions about ggplot2 and R itself.
The paid system could aim this kind of user, where the time to see the graph working is important enough to pay for, and the study and comprehension of ggplot and R can come later.
With this, the workflow will be answer first the pending "paid" questions, offering a complete full solution covering ggplot2 and R code, and after keep helping on the "open" list.
Complete full solution on the "open" list will not disappear, will just take some more time to shows up, because the "priority" of the paid questions.
And a suggestions: the paid solutions could be later accessible to everyone in the form of tutorials or something like. What do you think?
I think that the NPR model would work well for ggplot.- tell us what you're up against and how development of ggplot would benefit from our support. i assume most people have no idea, i dont.- have pledge drives!- have double your money challenges.- use schwag is a powerful incentive. I'll pay $20 for a $7 t shirt and then I'll raise it $20 while I'm paying for the shirt.-raffles. get community members to donate things that can be raffled.I would not take on any more services until trying out these things. ggplot is already an awesome service.James
--
-
******************************************************************************
James McCreight
cell: (831) 261-5149
VoIP (to cell): (720) 897-7546
--
You received this message because you are subscribed to the ggplot2 mailing list.
Please provide a reproducible example: http://gist.github.com/270442
To post: email ggp...@googlegroups.com
To unsubscribe: email ggplot2+u...@googlegroups.com
More options: http://groups.google.com/group/ggplot2
--
Atenciosamente,
Raphael Saldanha
saldanha...@gmail.com
James McCreight
Bruce Raup
Wikipedia model would be better than the NPR one, but a cool ggplot2
gear store, and a simple way to donate on the website would probably
go a long way toward raising the revenue you're needing. --Bruce Raup
Christian Jansson
Hadley Wickham
more flexible scatterplot matrix implementation.
Hadley