Facebook spam, the growth strategy that fizzled off too soon.
Every startup wants to nail a viral loop for its product. There’s tons of literature out there about how to go viral and watch your hockey stick develop, but there’s very little about how not to go viral.
The problem with viral design is that it has become less of a design issue and more a fad in recent times. Viral designs are forced onto products because they’re supposed to work. However, as with all fads, the execution is often counter-productive.
New products that try implementing viral design often do it in ways that ultimately harm them.
Spamming was never cool on the internet, still isn’t!
Designing for virality is essentially the internet offering you a megaphone. And megaphones can be irritating if used inappropriately.
Here’s a quick story told in 2 pictures:
BranchOut (Feb – Mar)
BranchOut (Apr – Jun)
What went wrong?
BranchOut achieved rapid success piggybacking on Facebook’s social graph. But in designing for virality on the social platform, it became a nuisance the way a lot of social apps on Facebook already have. Its ribbon game layer (the more invites a user sends out, the flashier the ribbon) was effective in prompting users to invite friends. Not only does the user inadvertently send out invites to 50 friends (Facebook imposed maximum), the service is designed in a way that repeats this process many times over. But users are far less tolerant of spam on Facebook than they are on email and there was considerable backlash against Branchout.
At some point, effectively every user on Facebook had received the invite more than once, and the whole thing had become a major nuisance. Users who had signed up without clearly experiencing the value proposition stopped using it; new sign ups stopped growing and users frustrated with the spamming started protesting, cumulatively leading to a mass drop off.
Viral invites work best when they are targeted
At the end of it all, you need every user to bring in, at least, one other user to go viral. This can either be done by spraying and praying (spamming friends list and hoping a percentage sign up) or by targeting invites to a chosen few for whom it’s relevant. Turn-based games do this rather effectively. Users send out invites which are contextual to their friends who might be interested. As a result, turn-based games have been fairly successful in new user acquisition.
No invite without proposition
The challenge of a social product is that users often see value in the community only after engaging with it. Since it makes sense to have the current user invite other users who would make the community more valuable to the current user, the invite loop should be triggered only after the user engages with the community. In contrast, community products often have an invite loop right at the time of signing up. If the proposition is clear initially (as in the case of Skype or Facebook), that would be a great point to get more users on the service. If not, it would just lead to spam.
The internet is characterized by overabundance and spamming is not going to help differentiate in the middle of a million other competing services.
Lecture at the MIT Media Labs discussing platform economics and growth.