Build a LinkedIn ad audience from companies with a buying signal this month

<context>
I'm building a LinkedIn company audience for: [campaign name and goal].
My ICP: [industry, company size, country].
Signals that mean "worth an ad now" for us: [for example funding news, a new executive, a hiring surge], in the last [30] days.
Exclude: [current customers, open opportunities and competitors, as domains or a CSV].
</context>

<task>
1. Credits. Run account_usage, which is free, and estimate the cost per page of results before searching.
2. Search. Find companies in Lusha that match my ICP and have at least one of those signals in the window. Pull [number] companies.
3. Clean. Drop companies whose industry doesn't match my ICP, anything on my exclude list, and duplicates. List every company you dropped and why.
4. Rank. Put companies with two or more signal types first, then companies with one.
5. Build two outputs:
   - An upload CSV for LinkedIn Campaign Manager with the columns companyname, companywebsite and linkedincompanypageurl, nothing else.
   - A working table with the same companies plus the signals each one showed and the signal window, for reporting.
</task>

<output>
- A summary: companies found, companies dropped by reason, companies in the final audience
- The upload CSV
- The working table
- The date to rebuild the audience, 30 days from today
- The credits the run used
</output>

<constraints>
- Company level only. Do not search for or reveal any contacts.
- Do not upload anything. I'll upload the CSV myself.
</constraints>

The situation

We ran the prompt on October 25, 2026 for an example seller of sales software. ICP: US software companies with 201 to 1,000 employees. Signals: financial news, executive news or a hiring surge between September 25 and October 25, 2026.

Summary: 25 pulled, 7 dropped, 18 in the audience

The industry filter let through six companies outside software, so the clean step is not optional. One company on the list was a direct competitor of the example seller, which is why the exclude list belongs in the prompt and not in a later review.

Signal mix: why the order counts

Every company in the pull had a hiring surge. The ones with executive or financial news on top of it go to the front of the audience, so a limited budget reaches them first.

The upload CSV (first five rows)

The working table carries the same companies with the signals each one showed and the window, for reporting after the campaign runs. Download LinkedIn’s current company list template in Campaign Manager before the first upload and match the column headers to it.

Credits the run used

52 credits for 25 companies, about 2 per company, including the signal filter. The credit check at the start is free. At that rate, a 300-company audience costs roughly 600 credits.