Communities

One of the things we’ve noticed over the past year since Gittip launched is that Gittip grows community by community. It’s not enough for an isolated individual to join the site, either as a giver or a receiver. They’ll have a disappointing experience. Gittip is a network and it depends on the network effect to be valuable: the more of your friends and colleagues that are on there, the more fun and useful it is.

Through historical accident, Gittip’s growth in its first year has been in the open source software community in general, and in the Python programming language community in particular. There are other communities represented on the site, but because the homepage to date has shown only the top ten givers and receivers on the site as a whole, and because Python was the first community to start using Gittip, it’s been hard for other communities to see themselves on Gittip.

That changes today with the launch of a new “Communities” feature on Gittip. Now, self-organizing communities of 150 members or more each get their own homepage on Gittip. Why 150? Because it’s the Dunbar number, the rough upper limit of the people in a social group that we can each keep track of in our own head. The communities feature is designed to split the Gittip homepage up into multiple homepages, one for each large community or ecosystem. In the coming months we’ll be introducing additional features on Gittip to better model projects and other smaller groups.

We actually soft-launched this communities feature two weeks ago, and today the Korea community became the first to cross the 150 mark, thanks to a call to action from Hong Minhee.

한국에 계시는 오픈 소스에 기여하는 개발자 분들, Gittip 한국 커뮤니티에 참여해주세요.gittip.com/for/korea/

— Hong Minhee (@hongminhee)

May 13, 2013

Unfortunately, international credit cards are a rarity in Korea, so the Korea community’s actual giving and receiving won’t be very robust until we’re able to improve Gittip’s non-U.S. payments story. That said, it was a wonderful surprise today to discover Korea appear out of nowhere on the list of communities on Gittip, and quickly reach the 150 mark. We’ll get payments sorted out sooner or later, and when we do they’ll still have been the first.

What will be the next community to tip into viability on Gittip? That’s up to you! What communities are you a part of? Go find or add them on Gittip!

Update: Python and JavaScript have become the second and third communities on Gittip!

New Design, New Team

Gittip has a new look, thanks to a collaboration between Damon Chin, Nick Sergeant, and myself (Chad Whitacre). The site launched seven months ago with a spartan user experience, and early on we recognized the need to do better. Within a month, Damon had contributed design files for a new visual design for the site. Damon is an employee of Gittip’s close partner, Balanced Payments, and I’m grateful to both for the contribution. The design languished, however, for lack of someone to implement it in markup and integrate it with our web framework.

Fast forward to 2013, and we’re starting to build the Gittip team in earnest. I announced a month ago that Gittip is hiring, and since then we’ve added over a dozen active contributors working on many different parts of Gittip. We’ve deployed important work from Joonas Bergius in refactoring the codebase under the hood, as well as smaller contributions from others, but we haven’t had a major user-facing contribution until now. I’m grateful to Nick Sergeant for stepping forward to convert Damon’s design into markup. I picked it up from there to integrate the markup with the site itself.

Refreshing the visual design of Gittip and recruiting contributors were two of five priorities I identified in October. Therefore, with this update, I’m satisfied that we’ve met those two goals. A third, Twitter integration, was finished in November. The two remaining goals from those five are non-U.S. payouts, and integrating with RubyGems.org. I look forward to announcements on those in the near future.

---

Chad Whitacre is the founder of Gittip, a platform for inspiring generosity through weekly cash gifts.

Design Gittip

Design community, I need your help. There’s a stigma associated with Gittip that it’s only for open source programmers. To be sure, some of that comes from the name (which we tried hard to change, but which we just need to live with for now). My sense is that the bigger problem is that the visual design that Gittip launched with seven months ago is holding us back from being taken seriously by the design community and by design-conscious communities such as musicians, writers, and artists. This is a bummer because that’s exactly who we want Gittip to work for. I want to support my favorite musicians on Gittip.

Within weeks of launching we started a ticket to “change the visual design of Gittip.” It proved to be a fairly popular ticket, and a month later we received a design contribution from Damon Chin (an employee of Gittip’s payments partner, Balanced). This languished for want of an implementor until last week, when Nick Sergeant stepped forward to convert Damon’s design to markup. That’s definitely going to move the needle for us, and I’m grateful to Damon and Nick for diving in. This post is about how to go even further.

Three weeks ago I announced that Gittip is hiring. This caused some controversy due to the fact that, as an open company, Gittip doesn’t compensate its employees, but the fact is that we now have a dozen people actively contributing to Gittip. This is fantastic! Our bottleneck, though, is front-end expertise.

Nick’s sweet spot for us is going to be converting design files to markup. Damon’s contribution came through his employer, and, while I love Balanced, we need to not lean too heavily on them. Gittip needs to build its own team. Therefore, I’m looking for someone with a passion for web product design to join the Gittip team.

What is it we’re building?

Core Product

Here’s a high-level overview of Gittip as a product:

• Profiles. We enable individuals to tell a story about how they’re making the world better.
• Giving. We enable individuals to participate in each other’s stories through long-term, no-strings-attached, financial gifts.
• Communities. We enable communities to tell and sustain a collective story.

Extension Points

Here are the ways Gittip interfaces with other sites:

• Connecting Accounts. We support connecting as many other services as possible to a Gittip profile. A few of these other services can also be used to authenticate with Gittip and/or move money into or out of Gittip.
• Partnerships. We seek out partnerships with third parties to add Gittip links in their products. We practice engineering as marketing.
• Browser Extension. We provide extensions for all major browsers, adding Gittip links on websites with whom we’re not yet partners.
• Widgets. We provide ways for participants to embed Gittip functionality on other websites.
• API. We provide APIs that enable third parties to build tools for participants to work with their data stored in Gittip.

Information Architecture

Below are the pages we have and need for Gittip.

• / - Homepage focuses on high-level aggregate data.
• /about/ - More charts and other info.
• /%participant_id/ - Allows individuals and groups to tell their story, connect accounts elsewhere, and manage their giving.
• /%participant_id/funds/ - Allows users to manage “funds.” Think mutual funds; see this ticket for a description.
• /communities/%community/ - Aggregates information about givers, receivers, and funds within a certain community. Community pages are described in this ticket.
• /on/twitter/, etc. - Represent accounts on other services that haven’t been connected to a Gittip account (example), or groups of accounts on other services (example).

That’s what we are building, and I believe that Gittip has nothing but open road in front of it. From our charts you can see that we experienced strong early growth, and then hit a rough patch while I transitioned out of my full time job in order to focus on Gittip. Now, here in 2013, we are building a team and are starting to grow again. Our goal is to grow two orders of magnitude over the next 18 months, and to do that we need a strong product designer on board.

Do you want to join us? You can get in touch with me, Chad Whitacre, as whit537 on Dribbble and Twitter, by email at chad@zetaweb.com, or by phone at +1-412-925-4220. IRC is another great way to connect; we hang out in #gittip on Freenode.

Thank you, and I look forward to hearing from you! :^)

---

Chad Whitacre is the founder of Gittip.

25¢ Growth Special

Gittip launched just over six months ago, on June 1, 2012. In the first 8 weeks it grew to 151 funders giving $955/wk. Gittip has only added 128 funders and $470/wk in the 23 weeks since, and, in the past 13 weeks, Gittip has only added 6 funders and has lost $117/wk. Interestingly, total users has continued to grow roughly linearly, with bursts in weeks 8 and 16, while funders has flatlined. Why?

(source)

I’m aware of three possible explanations:

1. Gittip has saturated the market. Most of the people who are remotely interested in giving money through Gittip are now doing so.

I don’t think this is true, because the number of givers on Gittip is really, really small (279) compared to the number of people who, e.g., are on the Internet (2,450,000,000 [source]). Yes, there will be additional growth when Gittip becomes easier for companies to use, but surely there are more than 279 individuals in the world who “get” Gittip.

2. Gittip has been neglected. According to this explanation, the initial burst of growth in weeks 1 through 8 was due to heavy product development and marketing effort, and the slow growth since then is because this effort fell off.

I think this is part of what’s going on. I left my job at YouGov in week 8 in order to pursue Gittip. I spent the next 22 weeks transitioning to a situation where I can sustainably fund my work on Gittip until I’m fully funded on Gittip itself. I worked on Gittip for 29 hrs/wk before I left my job, and 19 hrs/wk after I left my job through the end of 2012. While I was able to finish Twitter integration during this period, we also got hit with a significant fraud incident, which took up a good deal of time without adding direct value to Gittip users. The most glaring result of this neglect is Gittip’s continued inability to pay out to people without a U.S. bank account.

The reason I spent so little time on Gittip is that I was doing other things: l taught a programming class and I took a philosophy class (these were set in motion before I left my job), I made up for lost time with YouGov, and I searched for new consulting work. Happily, I’ve now made good with YouGov, and we’ve arranged for me to continue as a contractor on an hourly basis. I’m grateful to YouGov, and here in 2013 I’m able to focus on Gittip again.

Moreover, the past three months has made especially clear to me the importance of building a team to build Gittip. I can’t do this by myself. Therefore, Gittip is hiring.

3. The $1 threshold is too high. In week 8, we upped the minimum tip from 25¢ to $1. The thinking at the time was that this would increase the amount of money flowing through the system. It’s possible, though, that it had the opposite effect.

This seems to me a plausible explanation for why the number of users has grown consistently since week 8, while growth in the number of funders and the volume of giving has trailed off. New people are showing up on Gittip, signing up, going to leave a tip, and then balking at the price tag. In the long run I expect to implement a sliding scale, inspired by Humble Bundle. In the short term, I’m going to run an experiment for the month of January: I’ve added back the 25¢ tip amount. If you’ve avoided tipping on Gittip because the price tag is too high, this is your invitation to revisit Gittip and tip someone who inspires you. Thanks! :-)

A Growth Target

My goal is to see Gittip grow an order of magnitude in 2013. That is, I want to see us with 1,000 or more funders giving $10,000 per week or more by the end of this year. The average giving per person right now is about $5, so if that holds we would need to see 2,000 funders to reach $10,000 per week. Can we do it? Who inspires you?

---

Chad Whitacre is the founder of Gittip.

Gittip is Hiring

Building a team is as vital for Gittip as it is for any other company, but Gittip is an open company, which means (in part) that Gittip doesn’t compensate its employees. The idea is that people working on Gittip are volunteers who give their work away as a gift. Maybe they receive reciprocal gifts, maybe this happens on Gittip itself, but there’s no guarantee, no salary. This goes for myself, the founder of Gittip, as well. I’m not making a profit on Gittip, only receiving gifts like everyone else.

Without traditional compensation, the incentives Gittip can offer to employees are rather different. Here are some reasons I can think of to join the Gittip team:

Find a nice living. While I can’t offer you a salary, I will note that I personally receive about $150 per week in gifts on Gittip. That’s not yet enough to live on in Pittsburgh, USA, but it would go further in Hyderabad, India. At this stage in Gittip’s growth, funding your life via Gittip is a real possibility for someone in the right part of the world. My personal goal is to fully fund my life via Gittip within 18 months.

Learn. Education is changing, and apprenticeship is in the air. Working on a real web app with experienced programmers would be a great way to learn how to program and start building your reputation.

Support your open source habit. I had a Skype call with Mike Bayer the other day. He’s the author of the SQLAlchemy library, which Gittip uses. I realized during our call that I was basically trying to recruit him to be in charge of Gittip’s data layer. If Gittip were a traditional company I would have been offering him a salary and benefits and stock options, but Gittip’s not a traditional company. Instead, what I can offer Mike and others like him is the chance to build a platform that can eventually sustain his work on the project he loves, SQLAlchemy. I’ve got my own favorite open-source project, a web framework called Aspen. The idea of getting to work freely on Aspen was one of my inspirations for starting Gittip in the first place.

Have fun. I love Gittip. It’s deeply innovative and it’s got tons of potential and I’m really eager and excited to push it hard. I want you to share my joy. :-)

Work without a boss. Since you’ll essentially be a volunteer, I’ll never be able to coerce you to do something you don’t want to do. We’ll be partners freely collaborating to build Gittip together.

Change the world. Startups commonly talk about disrupting industries. Gittip sets the plow even deeper: with the open company model, we’ll be disrupting industry itself. That sounds like baloney; I actually believe it. There are some innovative workplaces practicing various forms of industrial democracy—Semco, W.L. Gore & Associates, Mondragon, Lincoln Loop, etc. Gittip dials this to eleven. Gittip is the most innovative company on the planet right now. If we pull this off, Gittip is going to have a far-reaching and unpredictable impact on our world.

Positions Open

Gittip’s immediate need is for:

• Developers. The stack is Python-based; see the README.
• Visual and user interface designers.
• Community advocates. This encompasses both marketing and customer support.

What “Hired” Means

“Hired” means we list you on the “About” section of the Gittip website as being part of the Gittip team.

You will have no formal obligations, as discussed above. The informal expectation is that you’ll be spending a good chunk of your life and energy building Gittip. Right now I’m spending about half my time, for example. If and when you end up lessening your commitment, we’ll simply unlist you from the Gittip website after a discussion with the whole Gittip community. No hard feelings. :-)

Gittip will continue to have a lively community of people occasionally participating in discussions and contributing code. What I’m looking to do here is build a core team of people working heavily on Gittip.

How to Apply

Create a new issue on GitHub introducing yourself and suggesting that we hire you. We’ll review these on a rolling basis, and decisions about who to hire and when will be open to the whole Gittip community to participate in. If you have questions first, hit me up at @whit537 on Twitter.

I look forward to your application! :-)

---

Chad Whitacre is the founder of Gittip.

Gittip Anyone on Twitter

Gittip is a platform for sustainable crowd-funding. It allows you to set up a small, anonymous, weekly cash gift to a person who is doing great work. Gittip launched six months ago, and currently has about 550 active users, exchanging about $1,400 per week in gifts. At launch, Gittip required an account on GitHub (a social network for programmers) in order to sign in. Now, it’s also possible to sign in using a Twitter account, and to pledge support to any Twitter account.

This change is important because it significantly expands the number of people that you can pledge support to using Gittip. GitHub is approaching 3 million total users, and Twitter announced today that they have 200 million monthly active users.

Giving to someone on Gittip is a way of saying thank you, and also a way to support future work with no strings attached. Now you can use Gittip to pledge support to musicians, artists, writers, activists, organizations and anyone else with a Twitter account whose work you feel is important and inspiring. Here are just a few examples to get you thinking about who inspires you:

• KvonBengtson is pursuing open source space flight.
• lisakristine is photo-documenting global slavery.
• rstevens is the author of Diesel Sweeties.

Once signed in with Twitter or GitHub, you can connect the other kind of account, too. If you accidentally create two Gittip accounts, you can merge them together by connecting both accounts to the same Gittip profile. All tips are moved to the new account.

Requiring an account on one of these networks in order to join Gittip helps to prevent fraud. It also helps with security, because it keeps Gittip out of the business of storing passwords (for now).

Per our roadmap from a couple months ago, the next priorities for Gittip are:

• Implementing payouts to non-US bank accounts
• Landing a visual design refresh
• Partnering with RubyGems.org
• Increasing code contributions

Now go find someone awesome in your Twitter feed, and Gittip them!

---

Chad Whitacre is the founder of Gittip.

The Delpan Incident

Gittip is a platform for sustainable crowd-funding. It allows you to set up small weekly gifts between $1 and $24 to people you believe in. Gittip is about five months old, and currently has about 550 active users, with about $1,500 changing hands per week.

Last week, we determined that stolen credit cards were being used on Gittip. We started investigating further to understand the nature and extent of the fraud, and we started taking steps to undo it and prevent future fraud. This is part two of that post. I named this incident the Delpan Incident after the account we first suspected of fraud, and you can find a detailed incident report here.

I conclude that $567.89 of stolen money was injected into Gittip over a period of seven weeks, representing 6% of credit card charges by dollar volume, and 5% by number of transactions. The impact of the fraud is apparent in the following chart, where the weeks in question are shaded red (the Gittip gift exchange takes place every Thursday).

I apologize for this fraud, especially to the original victims of the credit card theft, and to the ten innocent bystanders on Gittip who were affected. I’m sorry.

I have investigated the network of relationships stemming from the five accounts identified as fraudulent last week (a sixth account turned out to be legitimate). I have also reviewed all accounts that moved money into or out of Gittip in the past, and specifically those that had credit card failures in the past. With help from the fraud and risk officer at Balanced (our payments provider), I looked at account activity on GitHub accounts that were linked to Gittip accounts that also have a bank account attached. My thanks to the employees of Balanced and GitHub who helped out, as well as those anti-fraud professionals who reached out in confidence via email or publicly on Hacker News and GitHub to offer their expertise and support.

We now have an is_suspicious field in the Gittip database, with options “yes”, “no,” and “maybe” (technically, true, false, and null). Accounts start in the “maybe” category. Only accounts where is_suspicious is “no” are allowed to move money from a credit card into Gittip, or from Gittip out to a bank account. Accounts in the “maybe” category may exchange gifts within Gittip, but can’t move money between Gittip and the outside world. Accounts in the “yes” category are not included in Gittip’s weekly gift exchange at all, nor are they permitted to login. Whenever an account first links a credit card or bank account, it goes into a queue and is reviewed before being included in the weekly gift exchange.

As a result of this investigation, a total of 22 accounts have been marked suspicious, out of 6,308 (0.3%). None of these introduced money into the system last week, and, as shown on the above chart, the dollar volume returned to an amount in keeping with Gittip’s normal growth during the past three months. Therefore, I believe that we’ve identified all accounts that have fraudulently participated in the Gittip economy to date, and I have whitelisted all other accounts that have successfully moved money into or out of Gittip in the past. There are currently 431 whitelisted accounts on Gittip (7%).

Here’s a summary of the new categories:

Is suspicious?

Yes - 22 (0.3%) - Can’t move money at all; can’t do anything

No - 431 (6.8%) - Can move money; unrestricted

Maybe - 5,855 (92.8%) - Can move money, but only inside Gittip

Total - 6,308

Making Good

I have refunded the $567.89 of stolen money that was injected into Gittip. I have notified Balanced of the bank accounts linked to suspicious accounts that were used to withdraw $379.80 (67%) of the stolen money, and I am waiting to hear whether they are able to recover any of that money. $104.00 (18%) of the stolen money was given to ten innocent bystanders on Gittip, and will be recovered from those individuals’ existing balances and future gifts. $54.00 (10%) is still escrowed within Gittip, and another $30.09 (5%) went to fees for Balanced and Gittip, whence it will be recovered.

Then, there will be chargebacks. Victims of credit card theft have 120 days to file a “chargeback” for each fraudulent charge, which then takes a month or two to hit the affected merchant, Gittip in this case. Ideally, we’ve identified all of the fraud on Gittip to date, and all stolen money has already been refunded. However, a $15.00 fee applies for each fraudulent transaction that we didn’t refund in time, before the chargeback process began. (Chargebacks remove the moral burden of being complicit in fraud that I expressed concern about in my prior post.)

I count 29 fraudulent transactions between 2 and 9 weeks ago, so in a worst case scenario Gittip is looking at an additional $435.00 in fees. Assuming in this scenario that the money already withdrawn to bank accounts is unrecoverable, Gittip is looking at a burden of $814.80 for this incident, or about 400% of the approximately $200.00 that Gittip has earned since launching. Ultimately, whatever this burden turns out to be is the responsibility of Zeta Design & Development, LLC, the legal entity behind Gittip, and its owners, namely, me.

Lessons Learned

It turns out that Gittip is particularly suited to a certain step in the black market for stolen credit card numbers, where low-level agents purchase long lists of numbers and then verify which numbers are actually good by performing small transactions with them. This is often done in the form of small donations to charities that have simple, unsecured, online donation forms. However, the money is lost from the point of view of the fraudster. With Gittip, it is possible to set up a bank account on the other end to recover some of that waste. The upshot is that Gittip is potentially useful for a certain kind of fraud, even though Gittip doesn’t lend itself to quickly unloading large amounts of money from any given card.

Fraud was bound to happen on Gittip sooner or later. Now we know one form it will take. While the amount of bad money injected into Gittip was small this time around—only $567.89—I much prefer to gain experience containing fraud while the stakes are comparatively low than to have been overwhelmed by an even greater degree of fraud now, or to learn an even harder lesson further down the road. Gittip is better prepared for next time, with systems in place that we didn’t have two weeks ago:

• A whitelisting policy for transactions with the outside world
• The ability to blacklist fraudulent accounts entirely
• A fraud dashboard and incident reporting infrastructure
• Experience in data mining for fraud detection
• A start on communicating additional fraud detection signals to Balanced
• Relationships with anti-fraud professionals at partner companies and in the broader tech industry

Morever, this incident has given us a chance to test the principle of “maximizing transparency” that is at the heart of what it means for Gittip to be an open company. Rule #0 of anti-fraud is that you never tip your hand in the arms race with fraudsters. Obfuscation and so-called “information asymmetry” are the name of the game. What is the information that is obfuscated? Fraud signals, machine learning algorithms that mine as much data as you can get your hands on to find slight perturbations in the corporate cognitive field, slight disturbances that warn of fraud. It’s considered a truism that publicizing these algorithms gives fraudsters a much easier time inventing new work-arounds.

I accept the starting-point that fraud will always exist. The task is not to extirpate fraud—that’s impossible. Rather, it’s to keep it to a low enough level for normal life to proceed apace. It’s like insanity. Each of us has a tinge of insanity, but as long as we’re 99.5% normal, nobody notices. I also grant that sociopaths walk among us. Heath Ledger’s Joker has been my mental image during this episode. Why commit fraud? Wrong question. And yet, at this point, my own question is, can I make my fraud algorithms public, and still keep fraud below that 0.5% threshold? Because if I can, I want to. That said, I respect the right of Balanced and Gittip’s other partners to manage fraud on their own traditional, closed terms. There are fairly clear layers here between Gittip and our partners, and I believe Gittip can experiment with openness without jeopardizing the integrity of our partners’ anti-fraud efforts.

I don’t know exactly what this looks like yet, or whether I’ll end up giving up and pursuing an information-asymmetric arms race per the status quo. The incident report I published is a first step in exploring how transparent Gittip can be with regard to fraud. I hope to push that envelope further as Gittip grows. Stay tuned …

---

Chad Whitacre is the founder of Gittip.

Stolen Money on Gittip, Part 1

Gittip is a platform for sustainable crowd-funding. It allows you to set up small weekly gifts between $1 and $24 to people you believe in. Gittip is about five months old, and currently has about 550 active users, with $1,400 changing hands per week.

On October 10 (27 days ago), I noticed a new user named delpan in the top ten receivers list on the Gittip homepage. This user was suspicious, because the associated GitHub account was recently created, and empty: no repos; no followers, starred, or following; no name or location or avatar. Gittip has grown by word of mouth, so to have a new user unconnected to the rest of the Gittip social graph is unusual. But I didn’t want to jump to conclusions. There’s a Counter-Strike player that goes by delpan; maybe Gittip was breaking out into a new community? We adopted a wait and see approach.

Since then, it has become clear that delpan and other accounts on Gittip are in fact being used to steal money. The basic pattern is to create two Gittip accounts, one linked to a stolen credit card, the other to a bank account, with a tip set up from the one to the other. On payday, Gittip pulls the money in from the credit card and deposits it in the bank account. It does this via Balanced, Gittip’s payment processing partner.

I have identified six Gittip accounts that I strongly believe are linked to stolen credit cards. I determined this by manually reviewing the top 100 or so givers on Gittip, and looking for patterns. My heuristic boiled down to the following:

• The Gittip account is associated with an empty or suspended Twitter or GitHub account.
• Tips include significant tips to other Gittip accounts with empty or suspended Twitter or GitHub accounts.

Secondary factors included:

• The Gittip account is anonymous on the giver’s leaderboard.
• The “I am making the world better by …” statement is vacuous.
• The username is CamelCased.

I have reported these six accounts to Balanced. Together, these accounts have been used to steal $488.15 since September 27. The total charge volume during this six week period was $8,414.92, so the money stolen through these six accounts comprises 6% of Gittip’s volume during that time. However, we had an unusual number of credit card failures during the October 18 payday, and unless new stolen cards were associated with the same Gittip account, which I only noticed in one of the six cases, these would not have been reflected in the top 100 at the time I conducted the review. Consequently, I expect that even more stolen money has been funneled through Gittip. I need to do more research to determine how much.

Where is the stolen money now? Again, I need to do more research to fully answer the question. Anecdotally, most has gone to suspicious Gittip accounts, though a significant portion has also gone to legitimate Gittip users. Some remains escrowed within Gittip, some has been regifted, some withdrawn to a bank account—again, both by suspicious and legitimate Gittip users. Some has been paid to Gittip and Balanced as fees. The ideal is to get the money back to the people it was stolen from. Is this feasible? Do Visa and Mastercard make this possible?

The uncomfortable truth is that Gittip, Balanced, and our legitimate users are financially incentivized to turn a blind eye to fraud, because we have benefitted and are benefitting from it. I have stolen money in my bank account. Heck, I pretty much have stolen money in my pocket right now. The difficulty of unravelling the flow of money once it’s in the system makes this even less comfortable. We’re accidentally complicit in the crime, with no easy way to make good.

What do we do?

The most important thing is to prevent stolen money from entering the system in the first place. Therefore, I am instituting a whitelist: every giver will be reviewed and approved before Gittip charges their credit card. Gittip has about 350 credit cards on file and is only adding a few each week right now, so for the immediate future I will just manually review all paying accounts before payday each Thursday. I’ve already started adding admin UI to facilitate this and updating the payday script to enforce it, and no money was charged last week to the six accounts I identified.

As Gittip scales, we’ll need to rely more on algorithms and less on human intervention, though it’d be nice to avoid the nightmare scenarios that people run into with PayPal. The fact that Gittip accounts must be linked to a Twitter or GitHub account comes in handy here, and an automated or semi-automated approach based on the heuristic above seems like a good next step.

We still have the problem of recovering stolen money once it has entered the system. I need to do more research to understand what that looks like and what our options are with Balanced before I know how to proceed there, and that’s why this is Part 1. Stay tuned …

UPDATE: Here’s part two: The Delpan Incident.

---

Chad Whitacre is the founder of Gittip.

Let’s Apply for a Shuttleworth Fellowship

Update: Denied.

Gittip is funded on Gittip. That means the interests of Gittip as a company are perfectly aligned with the interests of Gittip’s users. It also means I can’t take any traditional venture capital to fund the early development of Gittip.

Mark Shuttleworth made a fortune during the dot-com boom, and has since established Ubuntu as the preeminent open-source operating system. He also has an eponymous foundation that gives generous grants to “dynamic leaders who are at the forefront of social change.” A few points in the Shuttleworth Foundation’s philosophy and model jump out at me. First is the focus on people more than projects:

We support individuals through fellowships (rather than project funding) because we believe that great people are the true change agents.

We want to give the person with the fresh idea a chance to try it and see if it brings about the positive change they believe it will.

We hope that by freeing up 100% of a persons time to follow their dream, it will become a reality.

The more we expose the thinking, working and practices of our organisation and our projects, the better.

Working together is faster and smarter than working against each other.

The Foundation does not […] fund initiatives with the development or advocacy of free and open source software as the primary objective. We rather apply the free and open source philosophy as the underlying principle to our work.

Lastly, the foundation’s “skin in the game” expectation is something I can certainly get behind, since I left a comfortable corporate job to pursue Gittip:

We make sure the Fellow invests, both personal resources as well as money into their projects.

My long-term goal is to find my living on Gittip itself—and if Gittip works for me, it will be because it works for many of us. I think we can get there in a year or two. How do we bootstrap Gittip to that point? It seems that a Shuttleworth Foundation grant would be a big help.

So, in the spirit of openness and community, I would like to invite your input and support as I apply for a Shuttleworth Foundation grant, starting with the application itself. Besides a resume and a five-minute video, the application is just four questions and two pages long:

1. Describe the world as it is. a description of the status quo and context in which you will be working
2. What change do you want to make? a description of what you want to change about the status quo, in the world, your personal vision for this area
3. What do you want to explore? a description of the innovations or questions you would like to explore during the fellowship year
4. What are you going to do to get there? a description of what you actually plan to do during the year

Applications for the spring cohort are due a week from today. I’m going to draft mine here, and I welcome your input. Let’s show the Shuttleworth Foundation that Gittip is a community worth joining.

---

Chad Whitacre is the founder of Gittip, a platform for sustainable crowd-funding.

Status Report, and Priorities for Fall 2012

Gittip, a platform for sustainable crowd-funding, is now just over four months old, having launched on June 1. The core of the service is small, anonymous, weekly gifts to people who do great work.

Here’s a snapshot of Gittip’s current status:

• 5,780 people have registered with Gittip.
• 1,773 people (31%) have filled in the blank, “I am making the world better by …”
• 341 people (6%) have a working credit card on file with Gittip.
• 275 people (5%) gave money on Gittip today.
• 40 people gave at least $10 today.
• Four people gave at least $50 today.
• We’re currently moving over $1,600 per week, and growing 5-10% per week.
• 30 people (1%) are set up to withdraw money to a bank account.
• 225 people (4%) received money on Gittip today.
• 36 people received at least $10 today.
• Three people received at least $100 today.
• There are 188 open issues, and 144 closed issues.
• 15 people have contributed code.
• Many people have participated in the Gittip community, via GitHub, Twitter, and Hacker News, and IRC, in particular.

That’s not nothing. I am humbled and excited by the way you all have gotten behind Gittip and used it and broken it and fixed it and constructively criticized it and generally pushed it this far in four months. Thank you!

A couple nights ago over drinks I was telling a friend about Gittip, that multiple patrons on the site are giving away thousands of dollars a year each. He was floored when, in response to a comment of his, I told him it wasn’t even tax-deductible. “That’s amazing that there’s a community that would share that much with each other without even getting a tax break for it!” No kidding!

Priorities

Let’s keep flooring people! :-)

We’ve got a lot of pressing issues with Gittip, and I admit that I can tend to get distracted by conversations that maybe aren’t the highest priority. Sorry about that. Stepping back, here’s what I’m seeing as Gittip’s top priorities for the next few months:

Twitter. Update: Finished. I started adding Twitter support, but it’s half-baked. There’s no way to link both a Twitter account and a GitHub account to the same Gittip account, and if you pledge to an unclaimed account it ends up garbled on your profile. We need to finish baking the Twitter integration.

Non-US Payouts. We need to find a solution for payouts outside the US. There are 22 people on Gittip who have accumulated $100 or more, and eight of them (36%) are outside the US. There’s a fair bit of pressure to add bitcoin support, which could solve this in theory, but in fact those affected are generally not supporters of bitcoin and wouldn’t take advantage of it. The clearest way forward at this point is probably adding PayPal, or at least implementing a manual process. Maybe we mail monthly checks?

Visual Design. Update: Finished. Damon Chin from Balanced has contributed a new visual design, which has yet to be implemented. I think this would make Gittip much more accessible to those further back on the adoption curve, along with giving us a chance to make the site just plain easier to use.

Open Registration. We’ve got an offer from RubyGems to add Gittip integration once we have email-based registration. This would be a great partnership to have in place, and email-based registration would effectively throw Gittip wide open in the process, further expanding our reach.

More Cooks. Update: We’re off to a good start. Now, as you may know if you’ve been following along, Gittip is funded on Gittip. It’s the only way I could think to do it that is true to the ethos of the site, but that means we can’t just take venture capital and hire a bunch of people to build this out. I left my corporate job two months ago, and have been able to spend about 20 hours per week on Gittip since then. We are seeing some code contributions from others—thank you! Can we do more of that? Do you want to run a Gittip install party at your user group or as part of a Python or other tech conference? I’d love to at least Skype in or otherwise help you out. The more people hacking on Gittip, the better.

There’s plenty of other issues, but those are the ones I am seeing as priorities. Do you agree? What do you want to see Gittip focus on over the next few months? How can I help you build Gittip?

---

Chad Whitacre is the founder of Gittip.