It's not like you can pay to have your customers refer more friends.
Anecdotally, I've heard Amazon shoots for $23. Thus they'll put basically unlimited money into any system bringing in customers for < $23.
You can run into a problem of volume though, i.e., you overfish the pond. Then you'll need to find new sources of customers (e.g., Facebook ads if you've only done SEM).
Unhappily for me, I'm capable of saturating all of Google's advertising inventory conveniently available at $15 per customer. (Its sort of like oil: the more you are willing to spend, the more exists in the world. I guess if I were an environmentalist I would describe my current state of affairs as Peak Bingo, but with some sustained effort into AdWords I could probably locate more sources of customers between $15 and $30, or alternatively in different mini-verticals.)
BTW, still chuckling at Peak Bingo. Well put.
This is a long way of saying that, while I'm happy with my volume, I'd be more happy with, oh, enough volume to hit the limits on my credit cards. In fact, I'd avail myself of pretty much every source of credit less odious than juice loans from yakuza or opening an account with CapitalOne.
I have paid for advertising elsewhere. The main issue is that a) Google has a de-facto monopoly on it in no small part due to the fact that b) their competitors s<%= "u" * (10 10) %>ck. Microsoft, for example, takes every bit as much work to get right as AdWords does, but you have to do it within their application, and they have less than one tenth the inventory Google does. To add insult to injury, it is at higher prices. I got a note in the email from them that my credit card had expired and could not bring myself to logging in to give them the updated credit card number, that is how useless they are to me.
I'm surprised prices @ Bing are higher, I'd assumed lower volume would mean lower competition and thus lower prices. But I definitely understand how the volume they can deliver isn't worth the time spent optimizing their system.
Thanks for elaborating.
Even more interesting, you can see which actions are predictive of referring friends (e.g., leaving a comment on the suggested features list). You can use the same strategy to predict the conversion rate from free to paid.
http://20bits.com/articles/almost-viral-a-hybrid-acquisition...