Many of you asked, so I decided to write this blog post, dissecting the launch of iubenda and the viral system we invented.
The base graph
To date we have 2700 waiting list subscribers and 1700 registered users. The product is still private, and only a couple hundred beta testers can use the product today. The launch day was on June 10, the waiting list subscribers are very in target, we went from nothing to this result thanks to what I’ll explain below.
The viral system
Instead of using launchrock or building a custom launchrock clone, we made the effort to move forward. We built our custom viral system based on the best examples out there, adding some further fresh ideas we had during the process. I’ll go through the explanation like a story more than a broth of data, this way you can understand the whys behind our choices.
The home page
Most viral pages don’t explain the product, and in my opinion this is a horrible mistake. Do you want to get emails, or people interested into your product? Because getting emails is worthless if you don’t reach the right people, and you’ll never reach the right people without explaining the product. So, the first thing we did is to explain the product, and show the product itself.
Another thing we did was to launch this page when the product was already on a beta stage, you’ll understand why.
The call to action is clearly the subscribe button, linked to the next page.
The sharing page
I named the sections to explain them one at a time.
When the user reaches the page, he realizes that the private beta is closed, and that there are no more seats. The beta is closed since we can’t afford too many users yet, so there’s an actual reason behind the choice. Here the user has two ways to react: do nothing (or maybe confirm the email address), try to get early access. Instead of building something cryptic, we made everything very clear. There are points, there is a position in line, and on a chosen day the first X users in line get early access. The reward is very explicit and the way to achieve it is measurable for the user, connected to some key actions.
- “A” Section, aka ‘when the hell will I gain access?’ → The point here is give to the user a measurable reward for the sharing actions. Launchrock asks to invite 3 people, while we went for something different. We have invitation batches every couple of weeks, and the user has to gain points to reach the first positions in line and gain the right (on the field) to be a beta tester.
- “B” Section, naked sharing → Most viral systems only consider the referred people, while we tried to encourage people to share without caring of the reward. If you click on these buttons you gain a few points just for the sharing action (we track the callback), doesn’t matter if you have no followers on twitter and you created the twitter account just for that, you always get rewarded.
Here and on the other steps the number of positions gained for every action is explicit. This gives perception of what’s happening and of how the action will impact on the final goal (reaching the first positions in line).
- “C” Section, like and follow → Since we wanted to encourage people to follow us on social networks, we came up with this idea of placing a reward connected to the follow/like. We went from zero to 700 followers within 3 months (and counting): it definitely payed. This step is quite easy to get, we track the callback to assign the points. This is also good for people not wanting to spam their followers, but still willing to gain some points and stay up to date.
- “D” Section, traditional referring system → Here the user realizes that the sharing link can lead to gain some extra points if other users use that referring link to subscribe. Most people end up sharing twice, so we differentiated the precompiled sharing text to make it less repetitive.
- “E” Section, email confirmation → Even this step has a reward. You can check your position in line only if you confirm the email address, and this leads to an extremely high confirmation rate (nearly everyone). Moreover, the referring system only counts confirmed emails, so cheating is not so easy (like it’s instead with launchrock).
Turning emails into users
After clicking the confirmation link (from the email we sand at subscription), the user lands receives the request to choose a password:
After choosing the password, the user can finally read the position in line:
Again, there’s a reward behind (displaying the position in line), and we end up having not just emails, but valuable registered users.
TagliaBlog is a personal blog about web marketing, and we wrote an article on privacy policies in general, providing iubenda as a simple solution, but explaining the legal requirements and how to meet every obligation.
Ninjamarketing is a web magazine on web marketing and general web topics (covering viral stuffs, mobile apps, and so on), with a wide team behind publishing several articles ad day. On their website we wrote an interview talking about the product, with some questions provided by a lawyer that covers the legal topic on the magazine.
Fanpage is a sort of Italian Huffington Post, getting very big lately, and there I released an interview on the experience of this startup.
On the same day we also announced a seed round of funding from some great Italian investors.
Launch by the numbers
To date we have:
- 2700 emails
- 2400 verified emails (89% of emails)
- 1900 users (70% of emails, 80% of verified emails)
- 700 follows on twitter and likes on facebook.
- Thousands of shares on Twitter and Facebook. Here you can check: http://twitter.com/#!/search/realtime/iubenda%20include%3Aretweets
- 1100 users referred by other people (46% of verified emails, the only to be counted as valid)
- 1000 users who referred other people (42% of verified emails)
- K-Factor of 1.1 (users_referred/sharers, but there are other ways to calculate and intend it)
Explaining the spikes
There’s a reason for every spike.
- A → It’s the launch. As you can see, it performs very well, and the strategy to combine multiple sources payed back.
- B → Here we merge a previous list of subscribers, so this spike is mostly drugs.
- C → This spake is the result of two different events. As first, on July 1st, we invite the first Italian beta testers (the reason behind the spike on the referred users, green line). Then, a few days after, we send a general update to all our subscribers.
- D → Seedcamp rules here. We pitched at Seedcamp London, the experience was really awesome, really, and iubenda got featured on The Guardian and on The WSJ.
- E → Here two major events happen. We release a freebie, wp-bootstrap, WordPress theme based on the Twitter Bootstrap toolkit, downloaded a few hundred times (powering this blog too). Moreover, we publish a deep blog post called A New Generation of Privacy Policies, result of months of thoughts and experiments on improving privacy policies. The post trended on Hacker News for several hours.
- F → We get selected for Mind The Bridge, the most important Italian startup competition.
- G → We send this email announcing a new awesome redesign (still WIP, sneak peek here), a new beautiful email template and the first international invitation batch, scheduled on October 25. This was our best action in terms of effectiveness.