Blogpost: A staggering amount of data forced us to rethink the way we did filters and search. So far searching was queries and predefined filters.
This is what i learned from that project:
We were working on a people database for social media and our users should be able to both finding one person, they previously had engaged with, or define target groups based on attributes. The databases varied in size from thousands to millions of entries.
Searching, unlike the challenge,
should be simple – yet powerful.
The way we define problems and ask questions, shape how others understand, answer and act. If it’s to narrow it leads to tunnel vision. If it’s to broad we lose focus. We tried to balance these two options and provide a broad, yet focused way, of finding profiles and create target groups.
After all, search is about findability.
Do you find what you are looking for?
Typically searching is quick look ups. You do it often without ever thinking about it. Querying like that is one way of searching.
Asking, browsing and filtering are other very used ways, and we rarely pays any attention to it. But we do notice when one of the options are gone or if our search fails. Imagine that you ask somebody where you can find the local library, if they tell you the wrong directions you’ll probably blame them, and their lack of knowledge and not you own, that you didn’t find what you were looking for. That’s why searching has to be smart.
Try to remember last time you searched on Google. You typed keywords, got results, filtered to only show images, and started browsing before finding the photo you looked for. That’s a very common search flow, and it contains not just one way of searching, but multiple search disciplines. That’s why searching is complex.
When we search, we learn.
We go through different mental stages when we search. It starts out with optimism that you will actually succeed, finding what you are looking for. Hopefully it turns into curiosity and actionable results when you get your first answers. Actionable results are good surprises and you could discovering new and unexpected answers that will help you gain understanding and make you feel a lot smarter than when you started. That’s the joy of searching.
Search is a wicked problem
without one definitive solution.
Searching and how we do it is about context. One way of search might be useless in one situation but useful in another. The solution for our problem was combining query searching and browsing. We created a search bar that users would recognize and understand.
For each search parameter that a user inputs, we update the summary of their search. But we also write the search conditions in a easily understandable language, helping the user to reconstruct their query on the fly. The summary is a clickable area, that can be used for filtering down on particular areas of interest, making searching more fun on exploratory.
We didn’t have the resources to supports questions. But we believe in it and i wouldn’t be surprised if we in the future even were able to search our database by typing “Males from Copenhagen talking about Startups within the last 7 days”.