From: Mike Masnick (email@example.com)
Date: Thu May 04 2000 - 22:48:19 PDT
Okay, so you're point is if you can convince people you have enough data
you might find some suckers who are willing to pay money to you? I'm not
sure that's a sustainable business model (as some of the portals are
discovering now). I haven't seen anyone *prove* that all this data they've
collected is actually valuable in terms of convincing people to purchase
stuff (I could definitely be wrong on this, so if anyone has data, I'd love
to see it). Anyway, in 4 years we very well may be hearing stories about
the "excess" of companies like GM plunking down $70 million to advertise to
a bunch of freeloaders who certainly didn't care to have cars advertised to
Anyway, in terms of the Netzero deal, that's quite different than anything
Snapfish could put together. The info that NetZero has is significantly
more than something Snapfish will have, in that Netzero knows all the sites
you surf. That's a bit closer to "valuable data" and it's specific to the
user, and not general "psychographic" info.
At 10:29 PM 5/4/00 -0700, Nicolas Popp wrote:
>All I know is that if you have a large network of users and if you know
>enough about them, some people are willing to pay you a lot of money in
>order to be able to market to them.
>The best example I know is NetZero, the free ISP. Look at the type of deals
>they have been able to strike with GM
>"Providing NetZero fulfills certain performance criteria, the base agreement
>could represent $68 million in revenue to NetZero over the next four years,
>with up to $8 million occurring in the first year of the contract."
>Just 18 months ago, nobody was giving a chance to the free ISP concept...
>From: Mike Masnick
>To: Nicolas Popp; 'FoRK@XeNT.CoM'
>Cc: 'firstname.lastname@example.org '
>Sent: 5/4/00 7:16 PM
>Subject: RE: Someone please tell me why Snapfish.com is worth $40 million in
>I'm familiar with PRIZM and other psychographic and geographic profiling
>systems, and I don't consider that valuable data for "highly targetted"
>promotions... In general they give you some idea of the people your
>shooting at, but it's still a pretty weak crap shoot. Most of those
>systems, on their own, are close to useless (though, this does depend on
>what you're trying to sell or promote). When combined with other
>of info they can help somewhat, but you're still taking a lot of
>and you're going to be wrong many more times than you're right.
>Valuable data has to be much much much more specific to the individual
>you're targeting. The single data point (where you live) doesn't give
>nearly enough info to do really targetted promotions. I've spoken with
>people who do profiling and they'll admit how weak a lot of data is if
>not specific to that person (i.e. exact buying patterns of *that*
>Besides, if knowing our address was such a valuable piece of data,
>Amazon be profitable by now? Hell, they know my address *and* what I
>and yet their "highly targeted" promotions don't hit home very often.
This archive was generated by hypermail 2b29 : Thu May 04 2000 - 22:50:01 PDT