Re: real marketing or just catchy slogans?

[Dan Winship <danw novell com>]

Sri Ramkrishna wrote:
> I met the guy who did firefox's community (and release manager I
> believe)stuff (and I think marketing) at OSCON.  He said he would be
> happy to talk with us about what he did to help Firefox.

Gnome is not like Firefox. End users can see an ad for Firefox, decide 
that it's cool, download it, install it, and go. But end users can't 
download and install "Gnome". The closest they can come is to download 
and install a Linux distribution that is *based on* Gnome, which (even 
ignoring the huge difference in scale between a web browser and a 
distro) is a totally different thing. How would we tell users to install 
GNOME if we had a New York Times ad? Would we pick a preferred distro? 
Or let anyone who wanted to contribute money to the ad be able to put in 
a plug for their distro (even if that distro was really hard to install 
and was likely to end up driving users away)?

We can't sell ourselves directly to end users. We need to sell ourselves 
to Linux distros, and get them to sell *themselves* to end users. We're 
not like Firefox, we're like Intel! [Cue "Intel Inside" chimes] The vast 
majority of our "customers" don't "buy" our product directly, they're 
getting it as an integral part of someone else's product. Even if they 
do understand that this other product contains our product, they aren't 
going to be able to explain exactly what our part does for the combined 
product, where our part of the product ends and the other vendor's part 
begins, or how the possible alternatives to our product would make 
things different for them. At best, they'll be able to say "well, this 
one has 2.8 and that other one has 2.6, so I'll get this one because it 
has a bigger number!"

Of course, this doesn't necessarily mean we want to market ourselves the 
same way Intel does. Intel definitely markets itself to end users, but 
that's just part of its strategy to sell chips to PC manufacturers, who 
are its real customers. By convincing end users that PCs with Intel 
chips are better/faster/more-likely-to-get-them-laid than PCs with AMD 
chips, they keep the demand for Intel-based PCs high, which keeps the 
manufacturers buying lots of chips, which keeps Intel in business.

We could apply the same technique: convince end users that GNOME is 
better for them, so that they will preferentially install distros that 
use GNOME, so that distros (our real customers) will use GNOME as their 
preferred desktop. But there's a problem. (Sri, you might want to stop 
reading here :-). Intel only markets itself to end users because its 
products *aren't* any better than its competitors'. If their chips were 
unambiguously better than AMDs, then the PC manufacturers wouldn't need 
to be convinced to stay with Intel, it would just be the obvious choice.

The same principle should hold for GNOME. If we are actually better than 
our competitors, than all we have to do is make sure that the distros 
realize this (by marketing ourselves *to the distros*), and we win. And 
if we *aren't* better than our competitors, then we're working against 
users' interests if we try to convince them otherwise.

(And what are we going to convince end users of anyway? "Use GNOME! It 
has Epiphany! [Unless you're using Red Hat, SUSE, or Ubuntu. Or anything 
else.] It doesn't have an office suite!" GNOME isn't a whole story unto 
itself. "Desktop Linux" is the story, but that's not a story we can tell 
on our own.)

-- Dan