When people search for their job, they have complex information needs. They could express what they want in a paragraph. A financial analyst might want to know:

I want to understand what social media is saying about the Netflix-Warner Brothers merger to gauge how engaged the public is on the topic

Traditional search forces this user to translate their need into a few short keywords social media netflix warner-brothers Oh that didn’t work! Let me try just netflix warner brothers merger. What was I doing again?

We force users to review results, repeat, iterate. When we could just work on solving their initial research need.

Replace this with an agentic approach. An agent could be told the full narrative above wants, use search tools to iterate, evaluating results against the users original research question.

With modern agentic search, there’s little reason to force users into complex “hunt and peck” research loops. Let them work with an agent - do it before your competition does!

-Doug

(Luckily there’s a whole training course on this - Cheat at Search with Agents!)

Services: Training (use code search-tips) · Consulting

This is part of Doug’s Daily Search tips - subscribe here


Upcoming course: Build your own vector database

Build your own vector database

Want to understand what makes embedding retrieval fast, relevant, and useful in real AI systems? Join Build your own vector database and build the core pieces yourself, from embeddings and indexing to search and retrieval.

Doug Turnbull

More from Doug
Twitter | LinkedIn | Newsletter | Bsky