How to make your application viral - A guide to getting users
danieltenner.com
danieltenner.com
1. Utility - it must be high utility, even if nobody else is in the system.
2. Network Effect - As people are drawn into the system, it needs to scale in value more than linearly
3. Residual value - Value taken out of the system, sometimes as revenue. Typically ads.
I think that it's not good enough to do these alone, however. You must construct something that isn't just viral but that makes it such that the fact that some person is not on the system is PAINFULLY ANNOYING to the actual users of the system. Consider the growth of email, in general: "Please just go sign up for an email address somewhere so I can email you!"
If you construct it in this manner, you do not have to spam people. Delicious, which has a lot of users, never emailed anything other than password resets, for example.
Firefox extensions and RSS helped a huge amount. And it was close to one of the first user data sites that had an API (before Flickr, I think...)
I think in 5 years time delicious & flickr will probably be the two webapps i still use.
Doesn't this simply exclude some domains. email, facebook, dating sites. Are you suggesting that these domains need to be approached by inserting some basic (slightly unrelated) utility that is useful without other people?
I was just trying to demonstrate pieces that live in the "social software" space which would be difficult to build while conforming to the rule. I suppose it would be possible, but it would be a squeeze of some sort.
I could imagine Skype starting as a phone book or email starting as a way of organising your faxes for the day, but it seems hard.
Say you've made skype. It's really good at being a phone, but it needs more people signed up to become Skype. Now you decide that in order to make more people sign up, you should make it into a great phone book. A lot of people might want to use a desktop phonebook already. If you made a really good one people would use it & once some of their friends used it, they could skype call each other too.
The problem I see with this idea is that you need your Skype to be a good phone & a good phonebook. If we're talking 0 cost, it probably needs to be the best at both. Being the best at anything is hard, two things is really hard.
BTW, I'm not really against the idea of utility. It's a way of cracking the chicken-egg problem. But would you discount the other ways?
Skype did not evolve in a vacuum. It piggybacked off the phone system. "Make calls for cheap" - which is a direct call to utility.
Remember that any system that gets its users laid or makes them money can have the users jump through hoops to use it. Anything less than that has to have increasingly more finesse.
Remember that any system that gets its users laid or makes them money can have the users jump through hoops to use it. Anything less than that has to have increasingly more finesse.
Could I distil this to: "Skype got away with not having this because.."?
BTW, I'm not saying that creating something that is useful without a network is bad. It's good. I'm just saying that in many cases it's very hard. Say skype had gone down this path. They may have failed to find some way (eg phone book) of making it useful to just one person.
Forcing the approach seems like unnecessarily inserting a another binary hurdle.
On the other hand, the slightly less elegant approach of getting your small number of users that want to save on phone bills to nag their friends seems more suitable. The existing users have in-built incentive to do this. It's pretty much guaranteed to produce some results.
Taking this screenshot as evidence:
http://en.wikipedia.org/wiki/File:Skype_1.4.0.58_alpha_for_L...
Skype 1.4 already offered this. I'm pretty sure SkypeOut was available even before that, too.
We don't have to stick with Skype either. Facebook solved their (enormous) chicken-egg problem by targeting specific Universities.
Of course, to gain social validation of your product, it needs to be useful and something people will enjoy. There are some counterexamples, like hi5, that are fairly popular and got to that point because of their random email invitations (I don't know who you are Aaron, Britney, and Alexa), and to that, I'm not quite sure what to say. I guess they found a market. But as far as virality goes, in the genuine sense, I think Facebook/Myspace/Twitter beat hi5.
You don't want to make users want to spread your app, you want them to want to spread your app.
The "Invite people" forms are a really bad way of doing things IMHO. It's like asking your wife/hubby to fill in a form detailing how much she loves you.
I'd also say a good strategy is to give users something for free, that they think should cost money - they'll be sure to tell their friends off their own back.
Still good ideas in the original article though.
If I like something, I go out of my way to recommend it. If not - emails aren't going to convince me you have a good product - if people like it, it ought to spread itself.
Lets look at an example that I had to deal with about a week ago: The company I work for is having a pretty large event coming up, something that our customers would most likely be very interested in. My boss's reaction was that we should grab the email addresses that we have harvested from our website (signups require an email address), and send those people an email telling them about the event.
I'm sure that the majority of you can predict what my reaction to this was...something along the lines of "No, that is spam, that is bad, part of my job is preventing exactly the type of email you want to send out from making it onto our network. By harvesting email addresses from our website you are violating our customers trust; spam spam spam, BAD!" Naturally, the PHBs did not understand this and, despite my best efforts to stop them, they ended up sending the spam (although I forced them to use a different mail server than we use for sending mails).
After much anger, and much thinking, I have come to a conclusion about this. I have changed my mind. Now...I, as a geek/admin, DESPISE the sorts of mass mailings that they sent out, but why? I hate them because I hear "mass mail" and immediately think "0mG G3t V14GRA N0W!", something that I hate. I know the sorts of slimey tactics that spammers use to evade spam filters, and actively fight against them.
The "problem" is that the majority of my dislike for "mass mailings" are technical. I hate them because they are not what the email system was intended for, and because they are a pain in my ass. You know what, though, they're things that most of our customers want to hear. Looking through my email box, I see that there are some things (mailing lists, google's newsgroup summary emails that come every day telling me what happened in comp.sys.ibm.as400.misc that day, emails from a website called steapandcheap [like woot, but for outdoor gear], backup confirmation emails etc.) that are automated, and might seem evil at first (the commercial stuff like steapandcheap especially) but are actually things that I WANT to see.
Where am I going with all of this?
There is a good chance that your hatred for the sorts of tactics he describes here come from your nature as a geek. The unfortunate reality is that the majority of users just simply don't care. Look at things like myspace, look at success stories like that 15 year old girl that is making 7 figures with her website that does myspace layouts (err..I think she is like 19 now...she was talked about in the 30 woman web entrepreneurs making big $$ article the other day), look at facebook applications like honesty box, these things are HORRIFYINGLY bad from the perspective of a geek. But to a normal person, to the type of person that is keeping the "Click here for a FREE iPoD!" advertisers in business, and to the type of person you're probably trying to attract, these things that we might find "obnoxious", they find useful.
It is a sad reality, but it is a reality nonetheless.
Net promoter score is an easy tool you can use to determine trends in your users' experience. It can be used as a leading indicator of the 'N' part of your viral coefficient. eries goes in to detail in "Net Promoter Score: an operational tool to measure customer satisfaction" (http://startuplessonslearned.blogspot.com/2008/11/net-promot...).
Additional reading from HN members:
jfarmer's "Three myths of Viral Growth" (http://20bits.com/articles/three-myths-of-viral-growth/)
Social / Viral plays these days have a tough time because of how quickly these channels become saturated. IMO you can disregard a lot of the rules in the article as long as your app has network effects and a unique viral channel.
I will never use proflowers again; They don't respect my time. So far I think Amazon does it best.
Let's take for example, a mafia game in Iphone, when you play, you play offline (you are the only player) and your enemy is the computer. Will the idea grow?
Yes, if it's good, but after 3 or 4 years, may be the device change, or another startup takes your part. So solution? Connected gaming, here's the same game but a little bit different
* You download the game and run it
* The game shows how to play and tell you the vitality of having other players with your to help you
* you tell 2 or 3 of your friends and play together
* game over, your friend now create their band and call another friends, and friends call friends.
you'll finish with millions user in few days, it's like a chain and it spread quickly.
Look at most of the successful games of those kind, they are very simple but succeed
I recall that FB also started out by mass-mailing the Harvard mailing lists.
I guess virality often is just spamming or tricking the user into spamming.
The downside of forcing virality is reflected in additional metrics like retention rates and engagement.
There are uncounted examples of political statements packed into some music videos of marginal and pseudo-underground bands and the like.