David Hornik pointed me to a write-up by Christine Herron of the talk he made at FOO-camp. My blog is not ready yet, but the points he made are too close to what I’d doing now to let the opportunity go.
This is pretty much a point-by-point comment, so my recomendation is you read Christine’s write up and come back here, or best, you open them in two windows next on to another. In bold is the orignal idea, as I understood it, then a paragraph of comments.
Speads depend on social conventions, like hand-shakes:
Using social conventions is indeed a great deal; however, besides the handshakes, most habits are local, and this point might mean: hire local experts to translate your product; well, “hire”. . . Enroll them in the viral marketing scheme, which means: make a product that suits international travelers, diplomats and flight attendants;
AIDS spread faster in certain groups:
Regarding AIDS diffusion in the Gay community, Barabàsi makes a point in Linked (ch. 9): “Patient Zero” was one the first person diagnosed with it, and not only gay, but extremely handsome and a flight attendant with little fidelity; there certainly are different kind of holes, but there mostly are different kind of people. If O’Reilly says on his blog: “Service X is the future on mankind”, he’ll have more of an impact that I would on my non-existing-yet blog.
More importantly regarding the example you give, the AIDS rate is very different among man and women in Africa, where most relations are heterosexual: how many partners whom a man and a woman have (their “degree”) follows very different distributions, even if the average number of heterosexual partners has to be the same; this is actually mostly due to the fact the epidemics don’t transmit with the same probability from a man to a woman than the reverse. There is more in structure than groups, and Two-Sided Markets litterature tackled that quite well.
Diseases in essential aspects spread faster:
Be essential: e-mail & cell spread faster than kites, that much is for sure; however, good food practices are difficult to impose too — because it is essential; you have more competition in the core, and you need to stand much higher to make a difference worth changing the life habits of your users.
Diseases spread faster if they can pass most barriers:
Super contagion: what makes some trend even more contagious? The fact that it makes sense to everyone you don’t know: you would tell you neighbor about Skype, not that new financial management tool; this pervasive-ity might actually be the answer to core-service resilience.
Pleasure is the most efficient vector:
I shouldn’t probably comment on that, but 100 millions MySpace users can’t be there just because of the fun typographic tools.
Killing your user (fast) is bad:
“Lethal” might be too strong a word when talking about a tech service, but think of all the service that make you love coding: aren’t geeks somehow dead to the world? Mentioning a service without being able to make references to cypher-like notions certainly kills you as an ambassador to the masses. That is close to KISS, but not quite.
Dormant virus are more efficient on the long term:
Does asymptotic and unknown applies to services? Spreading demands the bearer to mention it, after all. . . However, understanding or perceiving the cognitive, social, conventional changes it demands might be a downer; human mind is the most plastic thing ever, giving room to any new idea, and such changing of its shape is usually unnoticed: consciousness that something went different tends to appear when coming back to one’s peers—how to avoid that goes beyond the syphilitic comparison, but is a key aspect. Cell phone didn’t seemed to change our land-line habit, so we applied them eargerly at first—and they did turned everything up-side down.
Somes places are cramed, and better at diffusion:
There are better people, and better places: can the good first one help you travel from one good second to another good second? Said otherwise: this is very true, next challenge is “How can I go from one room to another?”
Be a part of your carrier:
Would you call YouTube and MySpace a symbiosis? Then tell R. Murdorch, ‘cause he is about to cut his baby in two.
In a nutshell, I’d recomand a new service to consider many user profile, and imagine not only wether they would love the service, but wether they would have interest in recommanding it around them, provide some support to the new users that they know.
Of course, I am working on formal models to prove all that; progress is good so far, in spite of reading too many blogs and , and I’ll link them as soon as they are spell-checked.