Four weeks ago, we launched a “coming soon” page for a new product called SpringSled. The page had a referral program built in to it and we expected to acquired somewhere between 500 and 1,000 sign-ups, anticipating about 150 from our efforts and the remainder from the viral marketing effect.
In the first four weeks, 150,000 people signed up for SpringSled.
Yeah, no kidding.
If we had used Google Adwords to acquire the same amount of traffic, it would have taken 37 months and cost $2,230,792.
Here’s exactly what we did:
Setting up the website and referral program
1. We built up a simple, clean “coming soon” page
You’ll notice that the only piece of color on the page is the green “Get Early Access” button, providing an extremely clear call to action.
2. After someone signs up, our referral program prompts them to invite 5 friends in exchange for (substantial) free stuff
*We didn’t start showing what # in line people were until we had about 140,000 signed up.
3. We email them with virtually the same content prompting to refer friends
4. To nurture the viral marketing loop, we follow up through email each time someone gets a sign-up through their link
The follow-up emails are really the magic behind this system. We had tested the exact same referral program without follow-up emails a few weeks prior to launching SpringSled, and that launch (for a different product) brought about 4,000 sign-ups total, compared with 150,000 with email follow-up.
Getting the first people to sign up
To get the ball rolling, we submitted SpringSled to BetaList through their expedited review process ($59), which brought in our initial sign-ups, who then began sharing. One user shared it on ProductHunt, which was another great source of early traffic.
The rest of this post goes over interesting occurrences and findings, but is not directly actionable information.
The first user to sign up via referral was number 70, and she signed up almost exactly 3 hours after the first person. The table below shows the first 500 people to sign up. People who came in via referral are colored in orange, giving an indication of the sharing acceleration.
After posting to Betalist, we didn’t do anything to push the product, but as you can see, referrals became a very significant source of traffic. Of the 150,000 sign-ups, 99.79% came in through referrals.
Our home page has a conversion rate of 42.5% from visit to email submission. We don’t know how many people are sharing their referral link, unfortunately, but we do know that 15% of signed-up users are referring at least 1 person.
So for every 10,000 people that visit our page, here’s what happens:
For each user that refers, the average number of people they bring in is 5.8 (with the impossibly high outliers removed). So here’s the rest of the story:
And then the story repeats again, and again, and again. When we had about 140,000 sign ups, we calculated that for each initial user, we had acquired an additional 480 through the referral program.
Here’s the gist of it: Facebook is the only traffic source. Twitter brought nominal traffic, but I wonder if we’d have had more traffic overall if we hadn’t offered a “Share on Twitter” button and instead had prompted everyone to share only on Facebook… here’s the data:
It just felt right.
Luckily, it also turned out to be an effective number. Here’s a histogram showing how many referrals (x-axis) were generated by how many people (y-axis):
What else do you want to know?