I'm going to be carrying out research into how to rebuild a search engine from scratch for a large University data base. Existing search is slow and cumbersome.

I'll be researching user needs that will try and figure out:

What users are looking for; how they formulate queries; what they do with the results etc.

My question: has anyone got any researching guidelines for specifically search engine user needs?

EDIT There is no quantitative data capture set up. I wish there was.


Getting to the specific data you're looking for is likely to require:

  1. Carrying out a large scale survey to determine user types asking questions that will help you identify things like frequency of use, motivation for use, and level of user ability.
  2. Interviewing a couple of archetypal users for each identified user type to find more detailed information about those user's usage patterns.
  3. Running some task-based user testing sessions on the existing product with more users based on the information from the interviews to differentiate between self-reported and actual usages.

From there you should be able to get a strong idea of who is searching for what and how - This is essentially the same for any research looking into product usage

| improve this answer | |
  • And I'd love quantitative data for search terms so I can prioritise, but they have nothing to track this data. I don't want to run any tasks on the existing system as that's already been done by previous UX and we're totally rebuilding the engine. But usage patterns [and how users interact with other systems] is a good point. – colmcq May 2 '19 at 9:52
  • Wouldn't you be able to get quantitative data for search terms from the logs? – Andrew Martin May 2 '19 at 9:55
  • Edited question. – colmcq May 2 '19 at 10:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.