The future of Gittip
One small slice of the pie
The primary audience for this post are those already familiar with Gittip.com, but I’ll try to provide some context for newcomers.
What is gittip?
Gittip allows you to give weekly donations to any other person, ranging from 25¢ to $100. The purpose of this is to both say thank you for content they created that you enjoyed, and helping to give them financial freedom and flexibility to devote more time to the projects they love that provide so much value. Currently, gittip is focused around the open source software community, as that is where contributors to gittip have the most experience, but the platform works just as well for funding youtube artists, human rights workers, comic writers and so on. The idea is that people will be able to give up their jobs and focus full time on what was previously a hobby by necessity.
This piece by @izs, provides a good (if slightly pessimistic) view on the future of open source funding. In the more general gift-giving space, gittip differentiates itself in a number of ways. It is person based, unlike project focused sites such as Kickstarter and Indiegogo, or content focused sites such as Flattr or Centup. Unlike these sites, gittip also aims to be sustainable and continuous, small tips from large numbers of people reduce volatility. Also unlike an interesting site such as Patreon which allows users to tip for each piece of content created, gittip is ‘no strings attached’, it is a ‘genius grant’ and tipees are free to use the money for whatever purpose, which will generally go towards living to produce the awesome content they got tipped for, whether that is youtube videos, or a popular open source library.
Another, and perhaps the most important way gittip differs from these sites is that it is an open company. Gittip is funded on gittip. Development and all communication, from irc to hangouts happen in public. The gittip finances are completely open, and Chad, the founder, is a great guy who has a big vision, but is willing to listen and learn from the community.
Where is gittip now?
Gittip is still very small. A single project on kickstarter can receive 100 times the money gittip moves in a year. There is not exponential growth yet. However, there are reasons to be optimistic. John Resig’s initiative with Khan Academy of giving every developer the perk of having $5/week to give to the Open Source developers of their choice is very exciting. The reaction on twitter showed that this approach has mind share, and hopefully other companies will join in. There are already a number of corporate tippers such as MaxCDN, Balanced and Heroku. More are sure to join as mind share grows and we start to focus more on encouraging these companies.
Perhaps the best reason to be optimistic about gittip’s future is it’s growing community of contributors. Chad prides himself in “ building a team that’s building Gittip”. Gittip resonates strongly with some people’s personal values, and this is reflected in more active contributors. Chad still does the majority of the work, but reducing friction to contribution and cementing best practices, has started to yield results. Gittip is funded on Gittip. The gittip account is a special type of account called a ‘team’. A team account allows the owner to add users as members. These members can then assign themselves a proportion of the money received by that account. Gittip’s team is large, growing, and could one day be a sustainable model for multiple developers.
Polish, growth, vision, stability.
Essentially, the way forward seems to be to freeze the feature set gittip currently has, and focus on polishing to perfection and extending the reach to as many people and organisations as possible. Growth should really be our main focus right now. It is the surest path for gittip to be taken seriously and to increase the pool of contributors.
A major area the site itself needs to improve on is making it immediately obvious on first visit, what gittip is and why you would use it. We need to put the user story first. The community is great, but the site should be able to convey its purpose on its own. Work is being done to ensure there are solid landing pages to sell gittip to various communities, such as contributors, or companies, but our primary focus on the front page should be to hook and convert new users that are seeing the page for the first time. The best way to do that is to have concrete use cases, such as “allow your favourite youtube artist to have the financial freedom to work full time on their great content” or “Help an open source developer build and support the software you use “
The design of the site is functional, but is not as slick as other sites. Gittip needs a clear user flow with immediately obvious next steps. There are easy wins for our conversion rates with concisely worded and enticing calls to action. More polish is needed before you can simply give someone the link and expect them to get excited about it.
Growth is fundamental. The more money that flows through gittip, the more it will be seen as a legitimate source of funding. For content producers in other industries to seriously promote the platform to their fans, there needs to be more success stories.
Two main approaches will drive growth. A better conversion rate, and reaching more people.
Currently the funnel for new users is something like:
land on site->be sold on gittip vision->find actual people to tip->tip
We need to be better at selling the vision through concrete use cases, and far better at finding people that the user already has a strong desire to tip. Discoverability is very important and we need to take full advantage of social networks to do this. Showing twitter users that you already follow,or github users whose repositories you have starred will greatly increase the conversion rate. Show me my friends on gittip!
Long term, multiple currency support may help actually tipping, but I don’t believe that is currently a bottleneck.
We should also understand how social login on gittip is used differently by different groups of people.
For tippees, it is a way of confirming your identify to allow easy discovery by your existing fans. If we offer more accounts, then fans can use the username on the platform they are most familiar with, rather than having locate a potential tippee through a 3rd party like twitter. As we add more accounts, it makes sense to only show accounts that are connected.
Tippers use social login as a convenience to afford creating an account on yet another service. However, offering more accounts greatly enhances social discovery. We can use the information from these services, such as followers and friends to recommend potential people to tip, which will greatly help conversion rates. Connecting services such as youtube, will be greatly beneficial when we introduce gittip to communities that use those platforms.
Open Source community
The first area for more ‘aggressive’ marketing is the industry with most traction on the site, that the contributors are most familiar with, obviously being Open Source.
Two sides are clear.
We need to reach out to medium sized projects directly, and encourage them to advertise to their users to donate on gittip, or understand why they are unwilling to do so if that is the case. Even in these early stages, there are reasonable success stories, such as SQL alchemy or Requests.
Lets reach out to the top projects for each language and understand what they need from us to move forward. Sometimes a nudge is all they need, and we shouldn’t be afraid of giving that.
The other side is reaching out to companies.A good place to start would be to look at existing gittip users and identify prominent employers. It’s obviously a far easier sell to go though an already convinced developer. Khan Academy’s initiative may appeal.
Beyond Open Source.
The next step will be bringing on board industries that focus on producing content for a web audience. Bloggers, youtubers, artists, esports, comic writers, tutorial and guide writers, contributors on reddit or stack exchange. The strategy here will be to identify prominent personalities and encourage them to spread the word to their fans. This is the best way to reach a large audience and create virality.
Further out, once we have gained credibility and mindshare, we can move offline into charity workers, volunteers, human rights workers and scientists.
Chad is a great listener. He wants to please everyone, but I personally think gittip would benefit if he was more assertive. A focused vision will be stronger than being pulled in all directions.
An example where someone needs to assert a stronger direction is illustrated in the discussion about users’ profiles.
Who is gittip’s audience? Are we looking to find new fans to fund new projects like kickstarter, or get old fans to say thanks for previous awesome work. New projects or new fans imply that the profile page needs to read like a marketing campaign and hook the user in. Old projects or old fans leads the area to be used as a reminder of all the value they provide for the fans on the fence.
In my mind, Gittip is far more about tipping to projects and people whose stories you are already familiar with, and far less about using the site to find new people.
Ultimately, everyone has their own opinion, but Chad is in a position to ensure everything we do is coherent and aligned.
The site needs to stay up when a popular twitter personality, or blog mentions us. (John Resig/Khan Academy being the most recent to take the site down).
Areas are being identified to make the architecture more robust, but the tradeoff is less time spent elsewhere, and so we should be mindful of over engineering.
We absolutely need a cache. Requests seems to hit the database too heavily, and adding a caching layer is relatively straightforward and would make the most difference. Redis or memcache are the most likely options.
Equally an asynchronous task queue such as celery will likely be useful as we have more background processing.
Efforts to reduce friction in development and deployment by embracing devops are also good.
There is renewed vigor for moving toward json rest api and js heavy frontend client. I’m a huge fan of this architecture, but as thing stand it is a lot of effort for little gain given the issues that seem most pressing in the short term.
Longer term, gittip should be able to provide something for everyone and all communities. Allowing non-US payouts is a short term priority, but general multiple currency support is a big job that will help push gittip global with less friction. One off tips are an important long term feature, as although they fit less with the sustainable vision, it is desirable for gittip to cover the various use cases someone may have for giving.
We want to provide more accessibility to corporate sponsors, potentially allowing KhanAcademy style tipping through a UI on the site itself.
Equally we want to enable charities to use the site, by ensuring we are compliant with any applicable laws, so no money is lost.
Even longer term, I could talk about the possible effects a system like gittip could have on the economy as a whole, and the way we view money and labor, but it’s far too early for that, and this piece is long enough as it is.
Increase conversion through social discoverability and better explanations of use cases. Grow, grow, grow by more aggressively encouraging open source projects and technology companies, then move out into other industries. Keep things polished, keep the site running smoothly, and ensure there is a tight coherent vision.
Gittip is small, but large in the minds of those whose attention it captures. It will be a long road ahead, but there are great signs that a real community is forming for getting work done and driving the project forward. I am very excited to be involved in what I believe will be a part of history.