I recently wrote about how Watsi became the first Y Combinator non-profit and grew to the $4,500,000 donation level. We explored transparency as a core value and surprisingly effective fundraising tactic. There is more to that story.
Watsi co-founder Grace Garey also shared how Watsi grappled with the technology of transparency. Without viable technology, transparency is just another pillow-soft idea on a mission statement. Garey also reacted to my comments on personalization, as I was a bit critical of how First Round characterized Watsi's approached to personalization.
Transparency as a technical discipline
When it comes to re-inventing non-profits, technology and transparency are linked. Garey:
In the non-profit world, we’re at an inflection point; it’s very clear that donors and funders want very granular information about where their money is going. There’s actually an opportunity to tell a more real and honest story that way, and also to attract more money and resources. Having a technical component to your organization is starting to emerge as a key to this - whether or not you’re a tech startup like Watsi is.
From a technical standpoint, transparency wasn't so easy to achieve. The foundational tech? A Google spreadsheet, manually updated:
We had this huge transparency document - literally just a Google doc. We manually published twenty fields of information about every single patient posted on Watsi, from the name of the doctor who diagnosed them, to the cost of their care, to their expected treatment date, to a link to the screen shot of the fund transfer from our bank account to the hospital’s bank account.
The Google doc turned out to be a simple/effective solution - except for the manual updating:
For almost two years, we populated all that information manually. We were literally transparent until it hurt. Now, we’re starting to automate some of this stuff. That’s making it easier to maintain the level of transparency we launched with. It’s also opening new doors to what we can be transparent about.
Watsi's transparency architecture now includes a number of third party tools. The Google doc is now hooked up to Watsi's back end, using data clips powered by SQL queries. Garey explains:
Heroku makes this possible. You can save a SQL query, pulling various information straight from the database of your app. You can send the results from that query out into other tools. We now have a perpetually running data clip that sends all of these numbers, pulled straight from our back end, into a public Google spread sheet. This is version two, if you will, of the spreadsheet we used to manually maintain. The next version hopefully will be easier for people to digest in its design.
Let partners visualize your data
Watsi uses visualization tools to open their data up to third parties. A partnership with tech company Segment took data sharing to another level. Watsi used those visualizations to essentially open source their data, allowing third parties to explore data about everything from revenue growth to patient impact.
This collaboration led to a public event, Analytics for Good, blogged about by Segment's Diana Smith. At the event, analysts from Looker, Chartio, and Mode presented insights they found in Watsi’s data. They also advised on activation, retention, and conversion. One example - a glitch was discovered in renewal behavior:
[Chartio] also found that, on average, it takes Watsi donors more than a month to make a second donation. This insight was a bit surprising because after a month people usually completely forget about products they’ve tried.
That sparked a fruitful discussion, with Garey hypothesizing that donors responded down the line, once they received a notification that their patients were fully funded. Methods to loop donors back in sooner were hashed out.
These early successes have inspired Watsi to look at open sourcing more of their data to their tech-savvy user base:
We've been discussing building something ourselves, so it lets people query topics like the average treatment cost of a procedure online, or treatment cost variations by country. On the docket for us is to publish is the actual pricing breakdown for each treatment, down to things like the number of surgical gloves used, or the dosage of medication that a pediatric heart patient takes in Haiti after surgery, or the cost of transportation to the hospital.
We have all that information now. There are incentives for both the public and our internal operations to build a system that can structure and automate reporting on all of that data.
The implementation of transparency takes time. Garey's advice? Don't wait for the perfect tools to get started:
The biggest lesson we’ve learned in this process is that if you have a value that you believe is important, you need to make sure that you can go the extra mile to live it out - even if you need to patchwork it together in the beginning, like we did.
How transparency and personalization intersect
Transparency intersects nicely with personalization goals. With that in mind, I wanted to know what Garey thought about my hyper-personalization critique. I wrote:
I see hyper-personalization as a truly personal message to one recipient. Watsi’s text-versus-HTML campaign, at least the way it was described, was not a truly one-to-one message, and therefore not what I’d call “ultra personalization.” That doesn’t mean it wasn’t effective. The text-based email campaign clearly worked – thanks to its directness and transparency, and because it was sent to members of the same tribe if you will (Watsi monthly donors).
It’s a myth that every message we receive should be ultra-personalized. We usually don’t mind receiving messages that are sent to an entire tribe we self-identify with.
Turns out Watsi isn't using much of what I'd call hyper-personalization, or deep customization of each person's messages. Hyper-personalization is still difficult to scale. In many cases, it's not necessary. If a company has achieved a tribal belonging amongst its constituents, broader communications still feel personal (Example: U2 fans don't need a personal concert from U2 to feel a personal connection; their tribal identity suffices). The other key is to give customers - or donors in this case - easy access to self-serve their own queries.
Watsi has taken advantage of both: segmented messaging to smaller groups and easy self-service on queries:
Luckily, the nature of our product is such that there's a lot of personalization baked in. On Watsi, the crux of what most users’ actions hinge on is this very personal activity where they are connecting with an individual person. They’re funding that person’s health care. They’re hearing back about how that person is doing after treatment. People welcome these updates because they want this personal connection.
These interests translate well to segmented email messaging:
If we are looking for a marketing e-mail to send to our donors, and we can send them a follow-up e-mail that shares more information about a doctor at the particular hospital where the patient they supported received care, that's what we do. That information is easy for us to come by, because of the level of transparency we maintain, I can’t see why we would send them a general and less relevant story.
Watsi also sends emails triggered by actions within their app. These emails might go to individuals, but most often, they go to segemented groups:
Customer.io enables you to send e-mails based on what people do or don’t do in your app. We have events associated with the patient you supported, and what country they’re from. We can put together e-mail messages based on the information we’re tracking about donors, so we can deliver a more personal message to them.
Watsi's communications are fueled by the goal of making people feel their contributions, no matter how modest, had impact. Garey pointed me to her own profile page as an example. Under her bio, you can see the patients she has successfully funded. Each donor profile auto-populates with personalized data, such as this from Garey's profile:
Grace's most recent donation traveled 1,500 miles to support Andres, a 9-month-old baby boy from Guatemala, for a hernia repair.
The wrap - for now
Reinventing fundraising is one of those audacious goals you are guaranteed to be humbled by. Garey told me having a clean slate to create an organization using new tools and new ideas turned out to be a big advantage. Where Watsi ends up remains to be seen. But the lessons from their transparent adventure are worth noting.
Image credits: All photos provided by Watsi, from their team trip to donor locales in Guatemala. Except: screen shot from Watsi site, and Looker graphic with Watsi data via Segment.
Disclosure: diginomica has no financial ties to Watsi. I plan to become a Universal Fund donor. Salesforce, owner of the Heroku platform, is a diginomica premier partner.