While the information architecture is fine, the site is really ugly. There has to be something functional and beautiful, and not just functional and bare like craigslist.
The rating system is idiotic -- an early form of karma whoring where the marginal cost of high ratings is too low. Perhaps it would make more sense to gain a point a per day or hour, and a 10 point rating system, where you need to allot scarce points when giving reviews.
The payment problem is solved with PayPal and others, so that is a non-issue. People actually trust it.
Better templates that allow for easy and beautiful sales pitches would be nice. Some of the 1996 era junk makes me want to run for the nearest far-more-pretty myspace page.
Community would make a lot of sense for particular items, like bobble-head dolls and replica light-sabers. Buyers of one rarely have the other, which beckons for sub-culture. Though I did know a guy who had mock light-sabers and a collection of StreetFighter 2 bobble-heads.
Instead of trying to break into the news business, for example, reddit build a new kind of thing, and people started using that instead.
So if you want to beat something huge and established, the way to do it is to try to answer the question: twenty years from now, when they write this history of this period, what will people say replaced it? Such questions are very hard to answer, but they do have answers, and with some effort you can often see the beginning of one.
One example: Everquest. Its network-effect was dwarfed by World of Warcraft, and the gameplay didn't keep pace which led to a gradual weakening.
You're right about that. Having a better model doesn't work unless you can convince a critical mass of people to use your system.
It reminds me of the story of the Arizona Stock Exchange (http://www.businessweek.com/1997/09/b3516129.htm), which was (theoretically, anyway) better for trading than the NYSE, but failed because they couldn't convince enough brokerages to leave the NYSE and use AZX instead.
Blockbuster. For a long time, BlockBuster was thought to be invincible in the video rental space, because they bought up all the prime real estate close to video renters. Because you could always find a BlockBuster near you, people became regular BlockBuster renters, which gave them a steady revenue stream and reliable borrowers so they could afford to buy up even more prime real estate, which enhanced their brand, which got them even more people hooked. They are in the process of being killed by NetFlix, because NetFlix realized that with the Internet, you didn't need a physical store presence at all, rendering BlockBuster's real estate advantage moot.
IBM. IBM came to own the computer market by offering CIOs a single model (the System/360) that would serve all their needs - it could run all the software people needed, and it was expandable with compatible, more powerful models for when you outgrew its computing capabilities. As a result, "nobody ever got fired for buying IBM". As more Fortune-500 CIOs bought IBM, it became more unthinkable to buy anything else, because the whole industry was used to IBM machines. They were killed by Microsoft and the PC revolution because this dropped the prices of computers & business software to the point where CIOs were no longer making the purchasing decisions. The department heads and small business owners who now made the buying decisions were used to PCs at home, so they bought PCs.
The Rothschilds. The Rothschilds provided liquidity in European financial markets for most of the 19th and early 20th century, frequently serving as the "buyer of last resort" during financial panics. In doing so, they developed very strong connections to the ruling governments in Europe. These personal relations let them charge a premium when loaning money, which helped them build up capital, which increased their power to ask as a source of liquidity, which improved their power connections. However, they've been hurt badly by the rise of America as a major financial center (they have few or no connections to the American government) and by the massive growth of capital markets. They no longer have the funds available to compete with Wall Street giants like Goldman Sachs, Bear Sterns, etc.
If you look at these, there are basically two ways to break a network-effect monopoly:
Identify the factor of production that lets them operate so efficiently. Search for a technological breakthrough that makes that advantage irrelevant. Use every advantage the technological breakthrough gives you to deliver the service better and cheaper than the incumbent. (Eg: Netflix).
Identify a group of people who form a potentially large market, yet are marginalized by the customers of the monopoly you want to break into. Offer a product that makes this marginalized group more powerful than the customers of the monopoly. Profit as they become the dominant customer group. (Eg. IBM, Rothschild)
However, I don't think those examples have the same power of network effects that ebay leverages. By powers, I mean that in the mathematical sense. For a network of N nodes, the examples above have some network effect of N^1.3 or N+log(N), but I think ebay is closer to N^2 or greater...
The only way around ebay will be to somehow go around the edges (aggregation?) or fight a long hard battle against them with a superior product. I do think someone can take them out, but it will not be easy.
I bet there are more people like me. I currently make do with local services like marktplaats.nl which works better for me, but the ability to buy or sell niche materials isn't good. (the network effect applies to idiots/scammers/spammers so there aren't nearly as many).
In other words: There's this huge market that basically isn't being served.
But I wonder if maybe market fragmentation and not a single competitor will beat them. The network value math above might be off. As ebay has grown the fraud and moron problems have become serious. The future may be in the direction of finer grained, more vetted communities. Sorta like usenet got ruined by morons and the discussions largely went to more private web forums.
A superior auction/marketplace meta-search application could also help kill Ebay. If people preferred to use it, buyers would see auctions at competitors' sites, not just Ebay. Creating a meta-search site might be a good first step for an Ebay-killer.
I recently had some issues with eBay where I had to relist an item 3 times because of scammers. Why does your item disappear before the deal is DONE? Why not have ME decide when it is sold, and not the buyer?