SOQUIJ
SOQUIJ wanted to understand what law firms really need when they look for legal information, and where its own tools got in the way. Between February and September 2021, I facilitated framing workshops with SOQUIJ's product team, a discovery workshop with a large law firm's research team, seven user interviews, an alignment workshop, a quantitative survey, and interviews with buyers at large firms.
The content was still SOQUIJ's strength, especially in labour law, and its reputation with senior lawyers was excellent. The friction was everywhere else.
What the research showed
Searching is stressful
Finding the right terms, refining results, and the fear of missing the one decision that matters. A search could take 15 minutes or 3 hours.
The pricing fought the tool
Users paid for documents that turned out irrelevant, and knew their usage set next year's price. So they started their searches in free tools.
Too many places to look
Each search engine had its own commands and its own login. Users wanted one place and one way to search.
The conclusion
Clean the data, then AI
In 2021, before large language models reached the market, the recommendation was already clear. To use AI one day, SOQUIJ first had to decide which kind of AI could help, then structure its databases for it.
The research showed why. Court records came back as images of text. Naming differed from one database to the next. Terminology, synonyms and metadata were missing, so decades of content stayed invisible to search and unusable by machines. Clean the data, rebuild the information architecture, and the content becomes ready for AI.
Credits
- Facilitation and UX research
- Jean-Sébastien Daigle
- Client
- SOQUIJ