note: This is the first in a series of posts that go under the hood of the experiments (successes and failures) from the Monkey Inferno in 2013.
WaterForward
WaterForward, was an experiment.
Could we take everything we knew about social networks & viral growth and apply it to the non-profit sector.
In other words, could we make charity go viral?
The journey outlined below is my best recollection of the trial & error process we went through on the road to reaching over 5 million people and raising ~$850,000 for charity in 2013.
As you’ll see — this is not meant to be a ‘how-to’ (in fact, you’ll probably see more things you should not do). Instead, I woke-up this morning to write this post because I believe we need to hear more stories. Real stories. Ones that show the ups and downs, the warts, the moments of confusion and doubt — rather than stories ofovernight successes or brilliant visionaries.
Without further ado — let’s get started:
a problem worth solving:
Charities are in a tough spot.
Every year, charities need more revenue to complete their mission(like a for-profit business)
Money is easiest to get from your existing donor base.
So, to get more money, you need to get a bigger donor base.
But…to grow your donor base, you need to raise awareness and likely spend money on marketing.
…BUT every dollar that goes to marketing (for the future), is a dollar that does not go directly to the field, to save lives today.
This is what I call the ‘philanthropist’s dilemma’.
our ‘genius’ solution:
What if we use what we know from social networking & viral marketing to raise a little money, from a lot of people. In short, we wanted to “make charity go viral”.
Our team had built both Birthdayalarm.com and Bebo.com to over 40M members each through organic viral growth, so this should be easy (right…right guys?…where is everybody….?)
Attempt #1 — ‘what if we created a giant pay-it-forward campaign?’
What it was:
‘WaterForward’, a website designed to be a simple “pay-it-forward” campaign.
Here’s the analogy — I buy you a coffee, but instead of you paying me back, you buy a coffee for someone else. And the chain of giving goes on-and-on, until we’re all hopped up on venti Frappuccinos.
genius! and look — we’ve already hit TechCrunch!
Oops. Not the best headline. Somehow our charitable intentions were interpreted as some kind Nigerian Prince Ponzi scheme.
The #PRfail was the least of our concerns however, since we had a fundamental problem:
what went right:
- beautiful design
- the ‘celeb factor’ — we got Richard Branson, Jack Dorsey etc.. to kick off the campaign with big donations.
- the viral coefficient actually wasn’t far off (~0.7)
What went terribly, horribly wrong:
- took nearly one year of development time to go live (the next time someone says ‘let’s use Google App Engine!’, I will light their chest-hair on fire)
- WAY too complicated — sorry, my body revolts whenever I try to type out the explanation of how this site worked. You’ll have to trust me, it was complicated.
- we didn’t ‘tell a story’ about WHY it was important to donate, so even though the viral engine was revving, the emotional fuel was missing.
The moment of truth — Push Forward, Pivot, or Give Up.
It would have been easy to keep going, after all, the viral coefficient was ~0.7. But despite the data, our instinct was that we were on the wrong path. The site was too complicated, and users weren’t getting joy out of donating (there was no emotion, just transactions).
Giving up also would have been easy. After all, we had no upside in the venture. This was purely a charitable donation on our part, and we were getting itchy to start building companies inside the Inferno.
Instead, we decided to pivot. We had another idea. A very simple, crazy idea, that we wondered if it could work. Although there was no financial upside for us to keep working on it, we had an itch that we needed to scratch.
“It seems that curiosity got the best of us” — Michael Birch
Attempt #2 — “Remember that one thing we did at Bebo…?”
Not one to be demoralized, Michael Birch and Alex Tew hatched a slightly kooky idea that stemmed from Michael’s Bebo days.
Once upon a time — he created a little easter egg on Bebo.com. For 1 out of every 1,000 visitors, a small red button would appear in the corner of the screen.
If the user clicked the button, they would see a simple, nonsensical message (eg. “don’t click this bowl of spaghetti”. Of course…they clicked the spaghetti) which led to another page, with another message, and most importnatly — another big red button. This continued for, oh you know….100 pages.
The weird thing is that when they looked at the data, they realized that once somebody clicked 2-3 pages in….they were very likely to go all the way to 100! (lesson learned, never underestimate the power of a big red button and sunk costs).
Here’s how we adapted this to Water Forward.
Everybody told us we were stupid. That 20+ pages is WAY too long in this ADHD world we live in…But, like any good entrepreneur, we employed selective hearing.
We rationalized that we were going to learn our lesson from version 1.0 and make this less about the transaction, and more about the story. After all, Scott’s telling of the story is what started the charity:water movement that has led to over $100M in donations over the past 5 years.
We gave ourselves an irrational deadline (30 days) and got started.
*Cue the Trial & Error Montage*
Remember how Attempt #1 took a year to go live? This time we did things differently, and shipped over 45 versions in 30 days.
Some were minor tests (eg. A/B testing the landing page of the flow)
Others were conceptual — testing a happy story vs. a sad story. (lesson learned: people want inspiration, not depression — when surfing the net).
We changed buttons, copy, content, and calls to action until we finally had a flow that was working.
We used every tool we had:
- we tested the concepts/mockups with usertesting.com
- measured dropoff rates through each step on mixpanel
- ran experiments using a homegrown a/b testing tool
- and in the end ran 45 major experiments in 30 days.
And after all that…it still wasn’t growing virally.
We sat in a meeting room, fairly depressed. We did all the right things. We were data driven, we were lean, we were agile, we had every buzzword in the book…but that motherf*cker still wasn’t growing!
I have a theory that I throw around the office, which I call the Golden Triangle.
It was time to get lucky.
Luck Strikes
Luck came in two strokes for us.
First, a strategic change. Instead of using Facebook & Twitter as the growth platform, we switched to email.
In hindsight, this seems obvious. Charitable giving is less of a public broadcast and more of an intimate personal request. Also, we had built Birthdayalarm.com and Bebo up past 50M members each off of email sharing.
Second, we added “take 3 minutes to help us end the water crisis” to the footer of our daily emails for Birthdayalarm.com. On 1M emails a day, we were only getting ~200-300 clicks for the charity link…but importantly, it got us a few users from India — where it turns out our story really resonated! (as an Indian, I’ll assume that I subconsciously targetted the site this way)
Here’s what it looks like when your luck changes...
BOOM!
The elusive hockey stick graph.
We watched as people invited all their friends: 2,000 invites…8,000…200,000…finally 500,000 before we stopped sending the emails.
Why did we stop? Because we needed to optimize the whole site before we burned all our traffic.
OK this post is getting irrationally long (if you’re still reading, I’ll buy you a beer) — let’s fast-forward to the end…
Happily Ever After?
The ending of the WaterForward story is a bit anti-climatic.
No, we didn’t raise $100M like I had hoped we would (came up about $99M short)
Yes, we ended up going viral — but not in countries that mattered the most (US, UK, Australia, Japan, Korea) for donations. The countries where it was growing the fastest(eg. India) had the lowest donation rates (0.01%) —AND trying to processing payments in various foreign currencies turned out to be a nightmare.
The growth was unprecedented (5M people in 2 months), the journey was insightful, and the financial impact was meaningful, albeit not mind-blowing (~$850,000 donated).
Alas — 7 lessons learned from WaterForward
- If you want someone to take their credit card out of their pocket, emotion is the most powerful lever you have
- The best products are the simplest ones. Big images, big fonts, with one big button.
- Copywriting is a precious skill (eg. We doubled our open rate with a subject line “finally. charity done right!”)
- “donate” and “share” are loaded concepts that work totally differently in different cultures
- Knowing when to quit, pivot, or persevere is the hardest thing to do as an entrepreneur
- Find out what you are actually providing at a core level to the user. In our case, it was inspiration, that there are good people, doing good work somewhere in the world — and that you can be a part of it with a click of a button.
- I’m damn lucky to work at a place where we can take on a project like this, with no financial return on investment. There’s great joy and learnings that come from doing a project like this.
ps. if you’ve read this far, might as well go to charitywater.org to donate to the cause ☺
