This is a very interesting post on the future of eBay, by Brian Dear. He keys off a question I asked Louis Monier on stage Weds…
One of eBay’s main selling points for years has been: trust and safety. You’re gonna be fine if you buy or sell on eBay, even if the other person in the transaction is a total stranger halfway across the world. And that is true. Most of the time, things are fine. Fraud happens occasionally, but the vast majority of the time, even big transactions like computer and car sales go smoothy.
But now think 2005. Why might we need eBay less and less?
Consider craigslist with RSS, or, better yet, a notification service tied to RSS or email. “Notify me when a sofa with the following attributes and in the following price range and in the following general geographical area goes on sale.”
And maybe hours or days or weeks or months later, you get that notification, and your dream sofa is for sale, by someone you don’t know, but who lives nearby.
Why do I think it might be nearby? Consider for a moment how much PC/Internet household penetration there is now. And how much high-bandwidth penetration there is now. There’s a much better chance in 2005 that a whole lot of people who live in your own neighborhood or general vicinity will have stuff you want, and you certainly will have stuff they want, and both of you will have ways to find out about each others’ haves and wants. Does eBay’s trust and safety cushion still offer as much value in such a world?
The issue is spam/trust in most open systems. You cannot have the proposed system be viable (on a large scale anyway) without an intermediary that manages the spam/trust issues through reputation, for example. If you built such an open system, then someone would pounce on the opportunity to act as a filter to solve trust issues, and then that intermediary would have incentives and opportunity to extract the same rents that eBay does now.
How trust relationships can evolve, and move beyond the closed systems is also well analysed in an article on Ftrain.com, which brings the Semantic Web into the picture.
It’s easy for Feedster to support these two commenters. We’re a new sort of intermediary that uses [lightweight] semantics to aggregate very precise and timely offers for jobs, couches, etc. jobs.feedster.com is the first of many examples in which we’ll do just that.
On the precision side, if you want a software-related job in the eBay ecology and don’t really care where, try this. We’re only getting few so far when the search is this precise, but it’s growing quick.
On the other hand, if you take a look at a broader slice, like all accounting jobs in Seattle, the posting volume is already healthy.
Right, but isn’t it relatively easy to spam those listings? People may not do that now, and you may have a healthy set of listings, but what it is in the end is a kind of (gasp!) semantic web. That’s useful and leads into a discussion of why not have a semantic web. Search engines could create one by virtue of their market power, but perhaps they don’t because of the spam problem.
Brian misses a fundamental benefit of eBay — ability to trade across geographies and across the world. What happens when you cannot inspect the item? eBay’s right — trust, safety and the feedback system are still required in a global marketplace. And I think eBay delivers such availability notices to customers today…
…and what about impulse buying, where I don’t know that I want something until I happened to be somewhere where stuff is sold and see a good deal. Because eBay is a destination, I know that when I go there, there’ll be stuff that I might or might not be interested in. This as opposed to simple “request stuff I need” model, which is questionnable how many people purely ascribe to. Most do both type of buying.