Hello.

For this purpose, I would recommend to use Index type: Document crawler. Documents crawler type indexes directly parse the HTML output generated by documents, which means that all text located on or associated with a document is searchable. This allows document content to be searched more accurately than using a Documents type index. However, building and updating a Documents crawler index may require more time and resources, particularly in the case of large indexes and complex documents.

For more info please refer to http://devnet.kentico.com/docs/devguide/index.html?smart_search_definign_index_content.htm

Best Regards,
Radek Macalik