Search recipe: short queries, many of them, every language

LinkedIn post search is an AND over every word in the query. A natural pain phrase like "tired of prospecting by hand" needs every one of those words in the post and finds almost nothing. Short topic queries find the same post and hundreds more.

Measured on one real day, same intent, same account

Query style	Queries written	Posts found in one day	Target post found
Pain phrases, 4 to 6 words ("tired of manual prospecting", "still building lists by hand")	16	18	no
Topic keywords, 2 to 3 words ("Sales Nav", "prospect list", "cold outreach")	20	638	yes, ranked first by engagement

The target post was a founder's rant about paying for a sales tool and still building lists by hand. The short form of the tool's name found it; the full name did not. Never rely on one query.

Rules for writing the queries

• 2 to 3 words per query. Tools, brands, chores. The short form of a name and the long form are two different queries.
• 15 to 20 queries. Fewer misses posts; more returns the same posts again.
• Every language the buyer speaks. A French buyer complains in French. Write the chore words in each language ("prospects froids", "liste de prospects").
• Sort by date, filter to the past 24 hours (or since the last pass). Relevance sort surfaces old high-engagement posts already seen.
• Recall is the job of this step. Precision comes from the engagement filter and the post score. A longer query is not a filter, it is a hole.

Building the query list from the intent

Given an intent sentence and anchors, write three groups:

1. Anchors as-is. Each tool, brand or chore name, short form and long form.
2. The chore. What the buyer does by hand: "build lists", "find emails", "write sequences", "clean the CRM". Two words each.
3. The cost. What they pay or lose: "100 a month", "hours a week", "no replies".

Twenty queries across those three groups, translated into each language, is the daily list. Run all of them; the engagement filter throws away the noise for free.