SOLR is a standalone enterprise search server with a web-services like API (per definition). Many reputable site has used SOLR as their search backend. See a sample on AOL or the complete list here.

So, will it work? What do we sell?

• Sell grid power. This is no-brainer. But shouldn;t be the first in the priority list. For most site, search is a non-dominant action. So, they may already have the computing power needed to run their site (includes search activity)
• Tuneable search experience. User will no longer depending on what their CMS provide, instead they can create their own search experience using their defined field and search weigthing.
• Leverage content discoverability. Some site may have a big bulk of data inside their server. Indexing this data (on SOLR) may give them added value. Eg: faceted data can be displayed to attract visitors. It’s like “simple” data mining.
• Targetted ad . We can push ad onto the search page

Then, who are our potential customers?

• e-commerce site. There are many e-commerce site in Indonesia. many of then is simply built from scratch of using off-the-shelf FOSS CMS such as osCommerce which is lacking good search feature. yes, MySQL Fultext search just won’t cut it.
• document intensive site. Eg: library (OMG, there are lots of library out there), .go.id site
• any rich content site

Ok, so how are we going to execute it, technically?

• provide REST API, or
• simply expose the SOLR endpoint for integration and searching purposes

Former option would enable us to push our ad. And the later seems fit premium service, where user can process raw SOLR data.

So, anyone interested? Or desperately wanting one for your own site? Or, you suddenly want to start your own web crawler? That’s an interesting idea … Going vertical, anyone?

Photo source: swisscan