Fixing Stale Results

Search Now: 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. 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 these unbound queries the chance that a better and/or newer resource moving 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 money power 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 because:

  • Monetary Incentive - Users who discover new sites of high value will receive a portion of NeSE'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 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 warn that some particular work will be of extremely low 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 right now to look to - Reddit and StackOverflow are great examples.