Comments on: My Search Sucks! http://www.baynote.com/blog/2009/10/13/my-search-sucks/ Intelligence Collected Fri, 12 Feb 2010 19:28:57 -0800 http://wordpress.org/?v=2.8.4 hourly 1 By: Sam Mefford http://www.baynote.com/blog/2009/10/13/my-search-sucks/comment-page-1/#comment-678 Sam Mefford Tue, 13 Oct 2009 18:46:17 +0000 http://www.baynote.com/blog/?p=437#comment-678 Scott, I couldn't agree more with your points; precisely why we encourage our clients to use Baynote rather than allocate dollars toward attempts to custom-build solutions to address relevancy. My search clients tell me their end users are furious their search experiences aren't bringing the right documents to the top. Just an example of your #1, one client told me they re-branded one of their products, but their field agents continued to search for the product by the old name which was nowhere in the newly published documents. This took months to discover as a problem, and their end users experience countless similar examples that are never discovered, even if someone were assigned to manually find and fix such search problems. Funny thing is that same client also experienced #2 when they rolled their own (expensive custom-developed) document ratings system, then pulled it after a year because users were gaming the ratings. And as far as #3, I've found organizations struggle enough to deal with search on its own --- I think it's overwhelming for them to imagine the work it would require to have site search results influenced by other factors such as browsing and organic search terms. I'll add a #4 to your list: "Users' Needs Vary Over Time". As much as I educate my clients about how they can fine-tune their search engine with lots of custom relevancy rules, realistically no manual process can economically keep up with the ebb and flow of fickle user interests. I love the AT&T example where one day a search for "insight" might mean understanding or clarification, suddenly LG's hot "Incite" phone has users excited but unaware they're misspelling the product name, luckily Baynote's Social Search figured out that when users search for "insight" they're currently looking for LG's hot new phone. Scott, I couldn’t agree more with your points; precisely why we encourage our clients to use Baynote rather than allocate dollars toward attempts to custom-build solutions to address relevancy. My search clients tell me their end users are furious their search experiences aren’t bringing the right documents to the top. Just an example of your #1, one client told me they re-branded one of their products, but their field agents continued to search for the product by the old name which was nowhere in the newly published documents. This took months to discover as a problem, and their end users experience countless similar examples that are never discovered, even if someone were assigned to manually find and fix such search problems. Funny thing is that same client also experienced #2 when they rolled their own (expensive custom-developed) document ratings system, then pulled it after a year because users were gaming the ratings. And as far as #3, I’ve found organizations struggle enough to deal with search on its own — I think it’s overwhelming for them to imagine the work it would require to have site search results influenced by other factors such as browsing and organic search terms.

I’ll add a #4 to your list: “Users’ Needs Vary Over Time”. As much as I educate my clients about how they can fine-tune their search engine with lots of custom relevancy rules, realistically no manual process can economically keep up with the ebb and flow of fickle user interests. I love the AT&T example where one day a search for “insight” might mean understanding or clarification, suddenly LG’s hot “Incite” phone has users excited but unaware they’re misspelling the product name, luckily Baynote’s Social Search figured out that when users search for “insight” they’re currently looking for LG’s hot new phone.

]]>