Krita 2018 fundraiser Retrospective

[Boudewijn Rempt]

We've just completed the 2018 Krita fundraiser.  Here's a bit of an 
evaluation. Since there are at least two thing that I think are important for 
KDE, I'm also includeing the kde-community list.

We've just completed the 2018 Krita fundraiser. 

_Planning_

* During the fundraiser there was much pulling of hair and wondering what was 
going wrong, since we only seemed to be on track to make ~20k, while the last 
kickstarter made ~36k. In the end we got over 27k, if one includes the direct 
donation made by a larger sponsor. With the kickstarter, large donations 
usually were made at the beginning of the campaign, giving a boost at the 
start. I also approached existing sponsors, but only one answered (positively, 
but they haven't yet come up with money).

* We did not do a fundraiser in 2017: we did raise funds because of the tax 
situation. That broke down into ~20k from PIA and ~20k from the community. We 
didn't really work and create a campaign for that. This meant that we were out 
of practice this year. In 2016 we were like a machine: we had everything 
ready, a playbook for creating the campaign and a playbook during the 
campaign. Now it felt like we had to reinvent everything again.

--> Note for KDE: KDE e.V. doesn't want to do an End of Year fundraiser in 
2018 because it already received so much mony from Pineapple and Handshake. I 
would recommend not skipping the fundraiser, because I'm pretty sure the same 
will happen and the next fundraiser will be more difficult to organize.


_Kickstarter_

* This was the first fundraiser we did outside Kickstarter. Kickstarter itself 
was good for 4k in the 2016. We didn't want to use kickstarter because of the 
all or nothing thing is nerve-wracking, because kickstarter isn't fresh 
anymore, because of the expectation that you're a failure if you're not fully 
funded in a day and because of the cost and hassle of rewards.

* We did send out word of the campaign in kickstarter updates to previous 
campaigns, which did drive traffic a bit.

* Of the 2016 36k, at least 11k went to the cost of the rewards (artist's 
fees, production, shipping) and kickstarter fees. The net result is the same.

* We only had three rewards this time: voting, your name or nick in the about 
box and the Digital Atelier pack at 50 euros.

** The donation spread was very different from the kickstarter: we had a way 
higher proportion of 50+ donations. This suggests that we missed a lot of 
10..50 donations, so next time we probably should have a 10 and 25 euro 
attraction, too. Preferably digital, but postcards could be an option.


_Infrastructure_

* This was the first time we used Mollie for payments. Mollie is GREAT. We 
could accept all credit cards, lots of different kinds of internet banking 
transfers, SEPA, paypal, bitcoin. We had 11k through credit cards, 13k through 
paypal, 735 through bitcoin, and the rest through banks in the second half of 
the campaign

--> KDE really should consider using Mollie. It's friendly, well integrated 
with Wordpress and Civicrm, not at all expensive (total costs in were ~500).

* We were busy setting up civicrm for Krita. But that wasn't ready when we 
needed to go live, so we still sent out reward mails by hand, and did vote 
tallies by hand. That wasn't too much work, but led to mistakes. We've got the 
basis for civicrm setup now, and are working on creating the necessary custom 
data fields to import the csv files created by mollieforms, so we can approach 
our supporters again.

** Since civi wasn't up, we couldn't mail our supporters every time something 
new happened.

* We used the news-about-krita mailing list once, but forgot to use that for 
the 4.1.5 release. That might have helped.

* We had a hard time actually tracking what was going on on the site.

* We did not have a means to make our donors share that they had donated with 
their friends and network


_Traffic and Buzz_

* It was REALLY hard to drive traffic to the campaign. Much traffic seems to 
have come from Youtube, where we don't really have much of a presence 
ourselves. In fact, most traffic seemed to come from people watching old 
tutorial and review videos, going to the site and seeing the campaign. In 
other words, that traffic came through nothing we were doing.

* We had some events lined up for the campaign: a release, a preview release, 
some news articles, live streaming of bug fixing, a reddit AMA. Only the 
releases really showed a spike in donations, all the other things were pretty 
much ignored.

* It was not completely impossible to get the news out: Paul helped with 
getting us in Linux Journal, and we did some press outreach. I'm not sure how 
we could improve this, but without buzz, a campaign is hard going. Working 
with kde-promo was a great experience.


_Campaign_

* The campaign page was too long.

* The donation form at the bottom wasn't clear for some people: they never saw 
the voting option

* It was quite hard to get the idea of bug fixing and stability out: I thought 
we had a good story, but it might have been too technical still.

* Since there were no comments on the page (though we had provided for them), 
the campaign had no feel of buzz. We should try to imitate the Kickstarter 
comments tab next time.

* September/October might be a difficult moment for a campaign, with people 
coming back from holidays skint. Next campaign should be in May again.

* We tried to follow the lessons from the training in Vienna, but still messed 
up: our donation form was complicated, we didn't publish a target, we didn't 
engage our supporters enough.


_Conclusion_

The campaign _was_ a success. We have about seven to eight months funded. We 
did not incur a lot of costs (I paid for the Digital Atelier set myself and 
will start selling it in Krita's shop soon). But there are quite a few lessons 
to be learned from doing it all ourselves.

I think we really need professional help next time setting up a campaign: I 
think I would like to approach More Onion for an offer: Krita has one sixth of 
the downloads of a program like Blender, so we should be able to reach at 
least one sixth of the level of support Blender gets, which means we're not 
engaging our community enough.

-- 
https://www.krita.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kde-community/attachments/20181016/c1af351f/attachment.sig>