GiveCrypto Monthly Update — February 2021
Automated ambassador program launched, growth hacking strategies and the unexpected consequences of a bug
Automated Ambassador Program Launched
As discussed in January’s update, we recently pivoted away from the donation marketplace to an automated ambassador program. We decided to pivot because the donation marketplace wasn’t working as intended — simply put, we weren’t able to convert a meaningful percentage of visitors into donors. Instead, we decided to focus our efforts on automating and scaling the ambassador program, which has already proven its potential. To see what we mean, check out the results of the randomized controlled trial that we conducted in mid 2020.
We made the decision to pivot in mid-January and were able to launch (i.e. begin inviting ambassadors) on February 24th, about a month after we started development. We were able to move so quickly because we had already built much of the required infrastructure while working on previous programs.
So where does the ambassador program stand today? We’ve onboarded 30 ambassadors to be part of the program. Those ambassadors have invited 26 recipients who have requested $1,815 of help. Thus far, we have deliberately focused on selecting ambassadors from countries that score the lowest on the economic freedom index.
Part of the recipient’s onboarding process asks them to describe how they will use the funds. Here is a breakdown of the self reported ‘use of funds’ by recipients.
These numbers are small and based on only a few days of usage, but it is exciting to see real people flowing through the program.
Ambassador Recruiting Strategies
One of the primary challenges with this program is finding ambassadors who will act in good faith. Since we delegate the targeting decisions (i.e. who gets our help) to the ambassadors, it’s essential that we can trust them to select recipients who are actually in need. What makes this challenge even trickier is that we don’t have many personal contacts in our target countries.
With that said, we see this challenge as an opportunity to bring creative solutions to the table. As such, we’re considering different methods to prospect would-be ambassadors. One idea is that we can utilize publicly-available data to seek out potential candidates. Services that might be particularly useful for this include freelancing platforms, social media networks, donation sites, and peer-to-peer crypto exchanges. Aside from helping us locate potential ambassadors, this publicly-available data can help us scrutinize a would-be ambassador’s potential trustworthiness.
We have already started ambassador recruiting experiments on Upwork, Twitter, and Reddit.
Upwork
Upwork is one of the largest global platforms for connecting freelancers with those seeking their services. By using Upwork, you can find a freelancer who can help you with just about anything from software engineering to legal support to online research. What makes the platform function so smoothly is its robust reputation system. By combining a review system with other metrics like “Total [Money] Earned” and “Jobs Worked,” Upwork provides its users a level of transparency that’s hard to find elsewhere. And because Upwork is so successful at connecting freelancers and enterprises, both parties are incentivized to maintain a high feedback score on the platform so they can continue to work with highly-rated users.
With its high implied trust, we decided to begin recruiting ambassadors on Upwork. Our first strategy targeted freelancers of any type who lived in one of our target countries. We initiated contact by sending them a request for a proposal. In the proposal, we offered to pay the freelancer $30 to recruit three people in their community who were in need. One of the stipulations of this agreement was that they’d be expected to help teach the recipient how to use a crypto wallet and upload images and videos of what they did with the funds.
We realize that Upwork’s acquisition cost probably makes it unfeasible as a scalable source of ambassadors. But, its high implied trust should provide a baseline for ‘good’ ambassador behavior.
With Twitter, we’ve decided on a two-pronged approach. The first consists of finding Twitter users who live in our target countries. From there, we use various criteria to rule out less established profiles for reasons of trust. For example, if the user has been active for less than three years or has fewer than 500 followers, we rule them out. We are currently experimenting with different ways to engage with these Twitter users and are using a combination of tweets, DMs, and follows.
The second part of our Twitter strategy consists of running Twitter ads that directly target users in our desired countries. We direct them to a basic Google form, where we ask them to provide some information to determine whether they might be a fit for us.
We used a similar tactic on Reddit, but since Reddit doesn’t publicly disclose user locations, we tweaked our strategy by searching country-specific subreddits. We looked for active users in these subreddits and assumed that they either lived in the country or had contacts there. Like the Twitter approach, we are experimenting with different ways to engage with the Reddit users; chats, posts and comments/responses. We limited our scope to people that had a “karma” (Reddit’s feedback score) of at least 100.
As the table above makes clear, there are a number of other approaches we could take. In this table, we’ve scored each strategy by its scalability, cost, and implied trust. Eventually, we’ll want to determine which acquisition channels are generating higher-quality ambassadors. There are a number of interesting ways to determine the “quality” of an ambassador:
- Funnel completion rate of their recipients (i.e. how many of their recipients have uploaded evidence of what they did with the funds)
- Quality of the content that an ambassador’s recipients upload
- Blockchain analysis of the recipients’ downstream transactions
As is the case with standard growth hacking, we’ll perform these activities manually at first. This will be labor-intensive at the onset, but once particular strategies begin to show more promise than the others, we will begin automating the acquisition processes.
A Happy Accident: The Unintended Consequences of a Bug
As mentioned above, we pivoted away from the donation marketplace at the beginning of the year. Before we did so, something unexpected occurred. Due to an especially pesky bug, some of our donation recipients had their payments delayed for a few weeks. This delay coincided with a significant ETH appreciation, resulting in the ultimate value that was sent being about 3x what the recipient was initially expecting. Most of the payments ended up being around $1,000 each.
This is quite a bit more than we have given recipients in the past, and as such, it allows us to evaluate what recipients might do with a larger donation. To get an idea of what kind of impact those funds have had, we contacted the recipients and asked them how they have used the funds so far. Below are some of the video responses we got.
Staying Connected
As always, we love to hear from supporters. Please follow along on Twitter and our blog for information on our programs.
