> ## Documentation Index
> Fetch the complete documentation index at: https://docs.leadinfo.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Companies

> The account list: every company identified on a tracker with its activity, history, firmographics, tags, CRM status and scores, plus the drill-downs behind one company.

`GET /v1/trackers/{tracker_code}/companies` is the centre of the API. Each row is one company identified on the tracker in the date window, with everything Leadinfo knows about it. `POST` on the same path accepts a nested filter tree; see [Filters](https://docs.leadinfo.com/concepts/filters.md).

## Listing companies

```bash cURL theme={null}
curl --request GET \
  --url "https://data.leadinfo.com/v1/trackers/LI-5A14070B93509/companies?start_date=2026-09-01&end_date=2026-09-27&sort=score&limit=20&relationship=prospects" \
  --header "Authorization: Bearer <token>"
```

**sort** `string` (default: last\_seen)

`last_seen`, `first_seen`, `sessions`, `pages`, `duration`, `unique_visitors`, `visitors`, `persona_visitors`, `form_submissions`, `sessions_total`, `lead_score`, `engagement_score`, `icp_score`, `intent_score`, `score`, `name`.

**relationship** `string` (default: all)

`prospects`: not in a connected CRM, not tagged customer, competitor or not-interesting, not an AI assistant and not shared-network traffic. `existing`: in a CRM or tagged customer. `all`: no relationship filter.

**segment\_id** `integer`

Apply a saved Portal segment. Rules the API cannot translate are listed in `meta.unsupported_rules`.

Quick filters on `GET`: `tag_ids` (comma-separated, carries any of them), `exclude_tag_ids` (comma-separated, carries none of them), `country` (comma-separated ISO-2 codes), `name` (contains) and `url` (viewed a page whose URL contains the value). For anything richer, use `POST` with `filters`.

`exclude_tag_ids` is how you leave out your own relationship tags, such as Partner, Login or Trial user, which `relationship=prospects` does not know about: it only excludes the built-in customer, competitor and not-interesting tags.

## A company row

**Identity and firmographics**: `company_id`, `name`, `domain`, `website_url`, `country`, `city`, `province`, `postal_code`, `employee_range`, `employees`, `branch_code`, `branch_local` (the national industry code, for example SBI in the Netherlands), `branch_international` (SIC 1987, the code every company carries whatever its country), `legal_form`, `sales_volume`, `founding_date`, `linkedin`, `logo`.

**Activity in the window**: `sessions`, `pages`, `duration`, `unique_visitors`, `active_days`, `first_seen`, `last_seen`, `last_source`, `last_medium`, `visitors` (identified people), `persona_visitors`, `form_submissions`, `leadbot_submissions`.

**All-time history**: `sessions_total`, `pages_total`, `duration_total`, `is_new` (no sessions before the window), `is_returning`.

**Relationship flags**: `tags`, `in_crm` and `crm_integrations` (the connected CRMs holding this company), `is_ai_agent`, `is_shared_network`.

**Scores**: `score` and `score_breakdown`, `icp_score` (100 when the company matches an active ICP, else 0), `intent_score`, `people_score`, `profile_scores` (stored scores: the value per active profile and persona), `engagement_score` and `engagement_tier`, `lead_score`, plus the 90-day counts behind engagement (`engagement_sessions`, `engagement_pages`, `engagement_unique_visitors`). See [Scoring](https://docs.leadinfo.com/concepts/scoring.md).

### Two flags worth knowing

`is_ai_agent` marks the fixed company ids Leadinfo assigns to AI assistants fetching pages for their users: Claude, ChatGPT, Perplexity, Meta AI and Gemini. They are useful for understanding how assistants read your site (generative engine optimisation) and never a sales target; they stay in the companies list but are excluded from opportunities.

`is_shared_network` marks the signature of an ISP or shared office network: many distinct visitors with about one page per session over the last 90 days. Treat the activity as noise, not interest. These companies are excluded from opportunities and from `relationship=prospects`.

## Drill-downs for one company

| Endpoint | What it returns |
| - | - |
| `GET .../companies/{company_id}` | Full firmographic record. |
| `GET .../companies/{company_id}/sessions` | Every session in the window with source, medium, campaign, click provider, pages and duration. |
| `GET .../companies/{company_id}/pages` | The pages the company viewed with views, first and last seen and average duration; sort by `views` or `last_seen`. |
| `GET .../companies/{company_id}/contacts` | People who work there from Leadinfo's contact database, persona matches first; filter by `persona_id`, `department`, `dmu` or `position_keyword`. |
| `GET .../companies/{company_id}/score` | The score explained: pillars with matched and missed rules, pages, known facts and what the data cannot tell. |

> [!NOTE]
>
> The website identified the company, not a person. Contacts from the contact database did not necessarily visit. Only identified visitors (see [People](https://docs.leadinfo.com/concepts/people.md)) are known to have been on the site.

## Opportunities

`GET /v1/trackers/{tracker_code}/opportunities` ranks companies by engagement over the 90 days to `end_date` and adds the best person to contact: a persona-matched identified visitor, then a persona-matched contact from the database, then any identified person with a job title, then any database contact. It defaults to `relationship=prospects` and always excludes AI assistants and shared-network traffic; `exclude_tag_ids` leaves out companies carrying your own relationship tags. `icp_only=true` keeps companies that match an active ICP Builder profile; `persona_id` restricts the recommended person to one persona.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.