Our enterprise search team has just completed several significant Google Search Appliance implementations. (My hat is off to the development team - some of the time lines were insane!) Sharing success is a key part of the non-linear culture. We recently held a de-brief on the Google Search Appliance deployments and identified the key factors that drove success.
Download the NLC whitepaper on Proven Approaches for Success with the Google Search Appliance (registration required)
The initial list of lessons learned had 22 entries, but we quickly realized that six items are most significant. Interestingly, the most important tips are the least technical - they have much more to do with understanding search users (and protecting your job) than with hard core coding.
We’ve found that paying attention to these six tips, will ensure that your Google Search Appliance implementation:
- Meets the needs of your key users groups
- Meets these needs better over time
- Leverages other elements in the technical ecosystem
- Let’s you keep your job by keeping information that should be hidden, hidden
Between now and the release of the white paper, I’m going to write a few blog entries that give the sense of the most important lessons we’ve learned.
Tip Number One: Know Your Users
Anyone typing anything into a search engine is seeking something. That sounds obvious, and it is. But it’s also a powerful idea.
Understanding what motivates a person to search lets you anticipate the kinds of searches they’ll perform, the types of content they’ll need to access, and the barriers they might face.
As our whitepaper goes into considerable detail about the importance of scenario-driven design when contemplating a Google Search Appliance implementation but the central idea if very simple:
put yourself in the shoes of your most common or most value audiences and think through what they need search to do for them to declare success.
It requires you to exercise your mental muscles, but scenarios need not be driven entirely by your imagination. Web analytics, search analytics and customer support centre calls can help you identify what your most important visitors seek when they arrive at your web site or extranet or log in to the intranet.
Later this week I’ll discuss Tip Number Two - Defining Success and then Measuring Towards It.