« Web 2.0: Floorwax or Dessert Topping? |
| I got Soul, but I'm not a Soldier »
Stewart Butterfield says you can check in anytime you like... and starting soon you can never leave you CAN leave (maybe)... but only if your next Roach Motel lets you check out too.
Interesting take... i like it. So does Thomas Hawk.
Posted by Dave on Saturday, June 17, 2006 at 10:09 AM in Geeks, Tech, Startups | Permalink
| Save to del.icio.us
Yeah, that's a good point Dave -- I actually hadn't thought about it in comparison to the blog auth credentials. Of course *we'd* never do anything bad ... just the other guys ;)
Stewart Butterfield |
Sunday, June 18, 2006 at 11:26 PM
hey stewart, thanks for commenting... particularly since your comment was a lot more rational than my tongue-in-cheek post ;) apologies if i was being a little too trite with the headline, and i appreciate the clarification.
however, i think your question is an interesting one, as well as your answer -- if you do end up supporting industry standards & enabling APIs, i think your competitors will have some level of access anyway. but i take your point about enabling competitive access to user data.
curiously tho -- when Flickr asks for your blog credentials (username/pwd for TypePad, WordPress, etc), that's also an incredible amount of power to use on behalf of your users for a different platform. imagine if Yahoo suddenly launches/acquires a major blogging platform... that would put Flickr in the interesting situation of competing with external solutions where they have user credentials.
a more amazing example of this was when i worked at PayPal (before we were acquired by eBay), users were asked to provide their eBay username/pwd credentials (so that they could embed PayPal info in auction listings, and so they could handle close-of-sale invoicing & billing). it used to blow me away that so many people would willingly provide that info... which if abused, could lead to significant financial liability.
but overall, i think your point about customers is the most challenging item for discussion -- if someone wants to leave they likely will... so how much time do you spend on featureset to make them leave happily (or at least just neutral), as opposed to pissed off?
obviously you want to be customer-focused, but now you're in the situation of deciding whether to build features for your staying customers (fans) vs leaving customers. that's a tough call... most times you'd say screw it and focus on those that stay.
however, there *is* probably some value in at least making sure leaving customers aren't pissed off (not to mention some moral high ground too). this is sort of like how you treat people in the exit interview when they leave your company, or how you handle people you interview but don't decide to hire... it probably pays not to be an asshole, since pissed off people tend to hold a grudge they may even up on later.
regardless, i'm impressed Flickr is having the discussion. such introspection is usually lost within most companies, even moreso among those recently acquired by large gorillas.
keep up the navel-gazing. i'm sure some good will come of it :)
- dave 'flick fanboy' mcclure
Sunday, June 18, 2006 at 03:49 AM
To be clear, since we opened our API shortly after launching, you've always had the ability to check out: the API is open for non-commercial use, there are a number of export tools and that's all fine with us and always has been.
The question is: do you want a direct competitor to have access to the API, including handling authentication tokens for users? There's a risk there beyond just "users might leave" ('cause if someone wants to go, they'll go anyway).
(In any case, implementing support for vendor-specific APIs is not the best path to real interoperability. A better way is supporting standards: then import/export, though slightly less convenient, will generally work.)
Stewart Butterfield |
Saturday, June 17, 2006 at 09:25 PM
The comments to this entry are closed.