Fixing Stale Results
Current Search Engines are Good for Real-Time ("Time-Bound") Results
Modern search engines pride themselves on their almost real-time results. That is, they are quite good at returning results which include activity (usually in the media or social networks) that has occurred quite recently ("time-bound"). This works moderately well for current events but isn't much help for "time unbound" queries. For example: searching for a historical event, a technical tutorial, a scientific article, etc.
Search Now: Bad for "Time-Unbound" Results
In unbound queries the chance that a better and/or newer resource moves up the rankings without significant expenditure of marketing funds is unlikely. It doesn't matter if you just published the most extensive, most beautiful, best organized site on chemistry in the world - it will languish unless you have power in the from of money, time, and/or social connections.
Fixing Time-Unbound Results
This is extremely unfortunate as it encourages stagnation in the results and discourages efforts by would-be contributors who give up before (maybe) seeing some traffic results from all their hard work. So how do we fix this?
Individuals and organizations will be able to submit sites to the engine for indexing and ranking. A small number of sites may be submitted for free, additional sites will require a nominal fee (to discourage spammers from mass submissions). End users will be able to view recently added sites and vote on the value of such submissions and will be incentivized to do so via:
- Monetary Incentive - Users who discover new sites of high value will receive a portion of Phoebe's revenue associated with that site.
Quora's partner program is somewhat similar to what we are envisioning here, although we believe Quora's particular implementation is flawed and causes an increase in spammy, duplicate, and useless questions on the service.
- Authority Incentive - The discovery of valuable sites is one way to increase one's "real life" authority on a topic. If people regularly see you providing high quality results on a topic they will look to you as an authority on the topic.
This is similar to how authors develop authority over time. At some point one can safely say, "I recommend books by _____." Without needing to qualify that some particular work is of poor quality.
- Discovery Incentive - Everyone loves discovering and sharing something valuable with others, part of human culture - and internet culture in particular - is the creation and sharing of valuable content.
We understand others have tried this with search and failed. We believe we can learn from those failures and that there are great examples of successful discovery oriented businesses doing it right, e.g. Reddit and StackOverflow.