Way back on November 9, 2009 I attended Louis Rosenfeld’s workshop in Washington DC on site search analytics. As part of a book release Search Analytics: Conversations With Your Customers, co-authored by Marko Hurst (who recently presented at HUGE) this workshop proposed a practical strategy and  useful tips and tricks for getting your site to perform better, be smarter and be more contextual to your users needs.

I’m looking forward to the release of the book for additional insight, but here’s what I took away in a nutshell from the day:

Data Gets You Hot – Remember the game “hot-cold” from the playground? You’re getting closer; no you’re not – you’re ice cold! In the context of web, data from your analytics software or search logs is like a cheatsheet to this game. Search analytics gives you the confidence and justification to drive your content program in a focused, backed-up way. It gives you the foundation to make content for what your users are looking for – or tells you when you are “hot” or “cold.”

A Site Search Analytics Strategy is like an Archaeological Adventure – Wading through the remains of the past in search for valuable golden trails of information is the art that resides between hard data reporting and user experiencing guestimating (of what users want). Within this data, you have a window into the lives of your community – you can live in their shoes through exposed seasonal keyword trends and see what time they search for certain things. You can reconstruct (through design, architecture, linking and metadata) the paths through which they would typically flow (or want to flow); serving those who are yet to come.

Data Makes the Subjective Substantial – Interpreting search records to inform content decisions gives you footing to make smarter content decisions. Have a creative editorial vision? In addition to (or maybe in lieu of) spending time and money on fancy focus groups, take a look at your search data to see if you can identify clusters of terms and/or untapped topics. But this can’t be done by a computer. Believe it or not, humans are still smarter than computers – and the absence of a human interpretation means a loss of true meaning. Allocating human resources to your data analysis offers you interpretation of context and identification of subtle relationships between items in your site.

The Middle Torso is the Wild West – These are Louis’s own words and what he means is paying attention to the middle torso of the long-tail curve. More specifically, according to him “the middle torso is where the new wave of terms appear and where short head terms go to die.” When you graph your data, you’ll get a group of keywords that are used most frequently and most often.  This short head (at the front of the curve) is the most stable aspect of what your users are looking for. The rest is a fall towards a long-tail of unique phrases that are generally too specific for content targeting.

Grasping Search Stats Helps Your Metadata Plan – In times of poor tagging and scarce resources site search analytics can offer you a means of employing a metadata strategy that prioritizes the most important content (most valuable in terms of search). In this scenario, only the premier content gets an overhaul and energy dedicated to optimizing it. Some nerdy librarian-type might argue this as a triage approach to metadata application — one that ignores moving towards standardized metadata. Why put your valuable time and money against terms that are not your highest performers? In times of resourcing, budgeting and time despair, I think this strategy is spot on.

Check out the shared presentation: