I am a huge fan of Flickr.
I love the Lemonade-out-of-Lemons domain name - you just know that they sat there with Godaddy, punching in every combination until a misspelling finally came back as available: Nowadays people don't decide upon a business name and then try to find a correlating domain. Instead they check out random or loosely correlated domain availability and compromise. yafla had just such an origin (though it has really grown on me since).
I love the simple interface design of the Flickr website. I love the features (intuitive "Web 2.0" features like in-place edits of titles, captions, and so on, coupled with massive capacity and bandwidth with a remarkably liberal usage policy. Even on the desktop Flickr is a winner, with a simple yet powerful resize & upload utility that makes incrementally populating one's online photo archive a breeze).
I love occasionally browsing through some of the beautiful photos uploaded by other users for a bit of entertainment and inspiration. I love the ability to combine photos into sets. I love the distributed keyword method of loosely categorizing photos.
I especially love the way that you can set relationships to other people, allowing me to limit certain photos (such as those of my children) to family, other photos to friends, and so on. While I don't use it to build new cyber-relationship networks, I do find it worthwhile as real-world friends and family join the service (mostly at my incessant urging. I don't spam out emails full of pictures of my kids anymore, but instead upload to Flickr and provide access for the appropriate people. If they're interested, they can look. If they're not then at least I haven't filled their inbox quota).
If Flickr keeps going with the current philosophies and designs, it will continue to be a winner. Its competitors will have a really tough time doing something better, unless they start sending out cheques to users for using their services. If there was one possible weakness in Flickr's armour, it is that its competitors could use feature-rich desktop photo indexing software to kickstart their web venture.
I am not Flickr's optimal customer, however.
A large selection of llama inspired gifts, jewelry, art, collectibles and stuffed animals from Nose-N-Toes. Com.
While I use it to store and share my photos, I never click on ads (not only are they something I naturally tune out anyways, the keyword correlations makes for some really ridiculous impressions. I recently uploaded a picture of a llama at a country fair I visited. Now I'm getting ads selling llama goods and services), and the stickiness of the site is limited for me. Instead of uploading my photos and then participating in late night discussions with other amateur photographers, oohing and ahhing about each of their photos and hoping for the same in return, I'm very utilitarian in the way I use the site. My relationship network is not built in Flickr, or on the online world, but instead I crystallize my real-world network in Flickr.
This interests me because the business model of a large number of sites, particularly "Web 2.0" sites, rely upon a considerable amount of stickiness - Not only will you visit, but you'll hang around. Just like television, these sites hope to draw you in for a period of time under the premise that not only are you more likely to see some revenue generation that interests you (e.g. ads), but you're also more tightly bound to that community.
Time is finite, however. I've discarded countless web ideas because while the services might be utilitarian and useful for some needs of some people, they were too marginal to realistically charge a fee, yet there was no way I could rationally establish enough stickiness (you can only create discussion groups about so many things).