Company Leads Research Package

Main problem

Building a fresh, duplicate-free list of good-fit companies with real decision makers takes many manual hours.

Impact

• Reps spend time researching instead of calling.
• The same websites are checked again and again.
• Directories and listing sites clutter the results.
• Decision makers and direct phone numbers are often missing.

Current workaround

Manual Google searches, copy and paste into spreadsheets, and buying generic lead lists.

Weaknesses

• No memory of which searches and websites were already used.
• Spreadsheets lose call status and notes across a team.
• Generic lists often do not match the exact company type the team sells to.

Primary outcome

A growing, deduplicated call list of companies of one lead type in one country, each with address, main phone, a short call summary and the decision makers named on its website.

Key benefits

• No repeats: searches and websites already used are remembered per lead type and country.
• No directories: listing and ranking sites are filtered out and remembered.
• Real data only: the Agent saves only what it finds on the company’s website, and leaves fields empty otherwise.
• Optional enrichment: direct contact details for decision makers from the Orbitype lead pool.
• One place to work: call status, notes, filters and CSV export in the App.

Time to result

After setup, the first companies appear as soon as the first small batch finishes. The App’s own guidance says a run of 20 companies takes about 20 to 45 minutes.

Inputs

Country (CH, DE, AT, UK, US), lead type (default agency), number of new companies, enrichment on or off, and an optional description of a good fit for other lead types.

Steps

1. In the App, a rep starts a run. The App sends it to the Agent in small batches of up to 3 companies.
2. The Agent loads its memory, plans local-language searches and runs them one at a time. Known websites and directories are removed.
3. For each new website it reads the home page and up to 4 key subpages (legal notice, contact, team, about), then decides whether it fits.
4. Good fits are saved with company details and decision makers. When enrichment is on, those people are enriched from the Orbitype lead pool. Non-fits are remembered so they are never checked again.
5. Reps work the list in the App: filter, search, set call status, write notes, export CSV.

What is included

• Company Leads database schema (install SQL for 4 tables)
• Agency Research Agent (search, read website, mark domain, save company, save contact, enrich contacts, load memory, notify)
• Company Leads App (run launcher, lead list, status, notes, CSV export)

Prerequisites

A SQL connector with a connector API key, plus credits.

Limits

• Up to 50 companies per run in the App, up to 30 batches per run.
• The App page must stay open while a run works.
• A run stops after 3 batches in a row with no new companies or 3 interrupted batches.

1. Agency outbound in Switzerland

A team selling to agencies needs new Swiss web and marketing agencies → starts a run for agency in CH with enrichment on → gets companies with address, main phone and named owners or managing directors, ready to call.

2. A new segment in Germany

A team wants to test IT service providers → enters lead type it_service with a fit description → the Agent searches German terms and saves only companies that match that description.

3. Weekly list top-up

The call list for Austria is running low → the team starts another run for AT → searches and websites from earlier runs are skipped, so only new companies are added.

4. Call tracking for a small team

Several reps call from the same list → each sets the call status and writes notes in the App → everyone sees who was called, who is interested and who needs a follow up.

5. Handing off to another tool

A manager needs the list in a spreadsheet → filters by country, lead type and status → exports CSV with one row per decision maker.

Alternatives

• Manual research in spreadsheets.
• Bought or generic lead lists.
• One-off scraping scripts.

How this differs

• Memory across runs: searches and checked websites are stored, so each run looks for new companies only.
• Website-first data: company details and decision makers come from the company’s own website, and missing fields are left empty instead of guessed.
• Any lead type: describe a new segment in plain words, no code changes.
• You own everything: the data lives in your workspace database, and the App, Agent and SQL are editable.

Limitations

• Results depend on what companies publish on their websites and what web search returns.
• Five countries are supported without changes.
• Runs happen while the App page is open, not on a schedule.

The Template

The Template itself is free.

Usage that may cost credits

• Research runs: the Agent’s AI reasoning, web searches and website reading use Orbitype credits.
• Contact enrichment (optional): looking up direct contact details in the Orbitype lead pool costs credits for each value returned. Turn enrichment off in the run form to skip it.
• Notifications: the end-of-run inbox notification to the sponsor uses credits.

Costs depend on usage: run size, the number of searches, and whether enrichment is on. A per-run cap is set on the Agent’s goal settings, and you can change it.

Primary action

Install the Template and start a small test run.

Button text: Install Company Leads

Objections and answers

• “Will it invent contacts?” No. The Agent is told to save only data found on the website and to leave fields empty otherwise.
• “Will I pay for the same companies twice?” Searches and websites are remembered per lead type and country, and companies are deduplicated by website.
• “Do I need outside accounts or API keys?” No outside accounts. You only need a connector API key for your own workspace database.
• “Can I use it for something other than agencies?” Yes. Type another lead type and describe a good fit.
• “What does it cost?” The Template is free. Runs use credits based on usage, and enrichment can be turned off.