Skip to main content

​
Build your first prospect list

In this tutorial you will run a natural-language search, read what came back, download the results as a CSV, and open them. By the end you will have a working prospect list saved to your machine and a clear mental model of how the CLI search loop works. What you will need: landbase-cli installed and authenticated. If you have not done that yet, complete Install Landbase CLI first.
Open a terminal and run: 
landbase-cli "find B2B SaaS companies in New York with 50 to 500 employees"
The CLI sends your query to Landbase, builds an audience, and streams the response back. You will see a JSON object printed to stdout. It will look something like this (truncated): 
{
  "run_id": "run_abc123",
  "session_id": "ses_xyz789",
  "dataset_id": "ds_def456",
  "status": "complete",
  "content": "Found 142 B2B SaaS companies in New York with 50–500 employees...",
  "created_at": "2026-06-05T10:00:00Z"
}
Notice the content field — that is a plain-English summary of what Landbase found. The dataset_id is the ID of the audience dataset it created in your workspace.

​
2. Inspect what came back

Before downloading, ask for a peek at the raw records: 
landbase-cli datasets peek <dataset-id>
Replace <dataset-id> with the dataset_id value from the search response. You will see a few sample rows so you can confirm the shape of the data before downloading the whole thing.

​
3. Download the results as CSV

Now download the full audience: 
landbase-cli "find B2B SaaS companies in New York with 50 to 500 employees" --download=prospects.csv
The --download flag controls both where the file is saved and the output format. Using .csv triggers a publish workflow — Landbase formats the data as a spreadsheet. This takes about 30 seconds. When it finishes, you will have a prospects.csv file in your working directory.

​
4. Open the file

open prospects.csv
Or on Linux: 
xdg-open prospects.csv
You will see your prospect list in a spreadsheet. Each row is a company with fields like name, website, industry, employee count, headquarters city, and LinkedIn URL.

​
5. Save a different format next time

You picked CSV for this tutorial because it opens in Excel or Sheets. For programmatic use, .jsonl is more practical: 
landbase-cli "find B2B SaaS companies in New York with 50 to 500 employees" --download=prospects.jsonl
JSONL keeps all available fields and is easy to process with jq or load into Python. See Export data in different formats for a breakdown of all options.

​
What you learned

  • The CLI takes a plain-English query and returns a structured JSON response
  • The dataset_id in the response identifies the audience in your Landbase workspace
  • --download lets you pull the results to disk in one command
  • The file extension controls the output format
Next step: Try Enrich a contact list from scratch to add emails and phone numbers to the people in your list.