Reaching Sustainability

Winning Gittip’s Second Year

Gittip has had a great first year. It’s achieved enough traction to demonstrate the concept of small, weekly gifts and actually start moving real money ($44,724.13 as of this writing) into the bank accounts of people doing awesome work. All the people who have worked to create Gittip should be very proud of the truly difficult thing they have accomplished.

So, the obvious question is: What’s next? Chad Whitacre answers this question with the goal of seeing “people reaching the bottom rung of sustainability on Gittip within the next year.” It’s not clear exactly what this means, though he could be referring to a goal of some people making enough per week to live 100% on Gittip income. As of this writing, Chad himself is the only Gittip user whose weekly receipts ($320/wk) exceed that of a full-time US worker making the federal minimum wage ($290/wk). However, once you subtract the amount that Chad gives on Gittip, his net receipts are about 3/4 of minimum wage, similar to the second highest receiver, Jesse Noller. While Gittip incomes are still low (and the federal minimum wage is hardly a measure of “sustainable” for people living in the US), it is clear that a reasonable partial income for a handful of Gittip users is within reach.

But how large would Gittip need to grow in order to reach a “sustainable” level? That is hard to project because as the number of people giving grows, so does the number of people receiving. The Gittip payday API does not actually report the number of people who receive tips each week, but it does report the number of transfers to users’ bank accounts. Counting bank transfers may be a better measure, in fact, because it represents the number of users who are both interested in being paid by Gittip (some users regift all their tips) and making enough in tips to warrant a withdrawal. As of Gittip week #57, this number was 37 people. Although that may seem small, compared to the number of users who tipped at least one person in the same week (790), it isn’t too surprising.

Modeling Growth

It’s dangerous to try to extrapolate the growth of a young and volatile system (especially an economic one), but I think the exercise illuminates the various factors that go into the Gittip economy.

Increases in Tips

If we are going to extrapolate the performance of Gittip, we need a model, so let’s pick a simple, if extremely bad, model. I’ll based it on trends I saw in my writeup on plotting Gittip data over on Wakari. Let’s assume the number of Gittip users continues to follow the roughly linear growth of 360 users/week shown after week 50:

Note that a “Gittip participant” has only claimed a Gittip account, but does not necessarily tip anyone.

In addition, let’s assume that the fraction of participants who choose to tip at least one user remains stable at 5%:

We also should include the unexplained long-term trend of falling contributions per user, currently sitting at an average decline of 1.8 cents per tipping user per week:

Following these trends out to the end of year 2 (week #103), we would project the amount of money transferred that week with Gittip to be approximately $6400/wk. That is a growth of 85% over the $3449 that was transferred in last week of year 1 (week #51).

Increases in Payouts

The other end of the equation is the increase in the number of Gittip users receiving bank payouts:

Extrapolating this trend linearly from week 50 onward would project 87 people receiving bank payouts at the end of year 2, which is 135% growth!

More Money Chasing Even More People

While both of these growth values are impressive, they unfortunately do not predict an increase in Gittip incomes to the level required to sustain an individual (in the United States, anyway). In fact, the trend goes the other way, suggesting a potential drop in the average amount of money funded users take home!

Lies, Damn Lies, and Linear Extrapolation

Let me be clear: this simplistic model is not evidence that Gittip is doomed. Accurately predicting the doom of any system requires more information and much more effort than I’ve invested here. It is smart to be wary of over-interpreting analyses, especially when there is no assessment of uncertainties.

Moreover, the future is not yet written! People are highly non-linear creatures, and (quite literally) anything can happen in 46 weeks. What I want to highlight with the above discussion is that the current status quo will not get Gittip to its next milestone. I find that exciting, because it means we cannot sit back and idly watch the Gittip chart page week after week. There are interesting and difficult problems yet to solve!

Aiming High

In fact, the structure of the above model is designed to illustrate the components of the Gittip economy that we must focus on if we want Gittip’s Second Year to be successful.

We need to increase the product of all of these things:

1. The number of people who have Gittip accounts. (Growing!)
2. The fraction of Gittip users that tip other users. (Stagnant)
3. The per-user average amount of weekly tips. (Falling)

Gittip’s First Year has been profoundly successful at item #1, which makes sense. Gittip is an unusual idea in the world of crowdfunding, and communicating that message in an engaging way has taken a lot of effort. The combination of presentations by Chad, website improvements, news articles, and word-of-mouth has worked very well.

However, items #2 and #3 are equally significant in the product. Raising the fraction of tipping Gittip users from 5% to 9% (assuming everything else stayed the same) would have as much effect as all the projected growth in the total number of Gittip accounts over the next year. Moreover, there is also the unexplained long-term drop in average tips which is going to eat 20% of Gittip’s growth over the next year unless we can figure out the reason and try to stop, if not reverse, it.

At the same time, efforts to increase items #1-3 will generally also increase the rate of growth of the number of Gittip users receiving funds. That’s a great thing for the diversity and robustness of the Gittip economy, but paradoxically, it might also move the goal posts just as fast as we run toward them. We should be prepared for that, and recognize that sustainable incomes might not be the only goal at this early stage.

Listen to the Users

Before concluding, I want to step back and reflect on the people behind the data. We need to remember that the current and future Gittip community is not some mechanical system to coldly manipulate to make some number bigger. Instead, we should look at the data as one way the ever-growing Gittip user base can communicate their preferences to us:

• Do they believe in the Gittip vision, but can’t find anyone they would like to tip?
• Does their initial enthusiasm for Gittip cause them to overcommit, causing them to taper off donations over time?
• Is engagement dropping off because we don’t foster a connection between givers and receivers?
• Maybe reaching sustainable incomes for receivers first requires a much larger Gittip community?
• Am I even asking the right question?

Year Two of Gittip is going to require us to find answers to these kinds of questions. I’m looking forward to it!

Head over to the Gittip Issue Tracker to start digging in.