> ## 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.

# Account-based marketing: select, understand, act

> Pick twenty companies in an industry where you already win, understand what they showed interest in, and take a defensible next step.

Account-based marketing starts small: one audience, one problem, one outcome. This recipe takes you from your own customers to a shortlist you can explain to Sales.

**1. Learn who you win**

Ask the API to describe your customers:

```bash cURL theme={null}
curl --request GET \
  --url "https://data.leadinfo.com/v1/profiles/icp/suggestions?tracker_code=LI-5A14070B93509" \
  --header "Authorization: Bearer <token>"
```

The draft names the sectors (SIC major groups), sizes and countries that describe the companies you tagged customer or hold in your CRM, with the share and the lift of each value against all your visitors. If you have no customers on file yet, `basis=cohort` describes who similar Leadinfo customers sell to. Save the reviewed draft as your default ICP profile so the fit pillar reflects it.

**2. Build the shortlist**

Search in the sectors you win, among prospects only, with some interest and a decent fit:

```json theme={null}
{
  "start_date": "2026-09-01",
  "end_date": "2026-09-27",
  "sort": "score",
  "limit": 20,
  "relationship": "prospects",
  "filters": {
    "conditions": [
      { "field": "branch_international", "operator": "starts_with", "value": ["73", "50", "87"] },
      { "field": "icp_score", "operator": "equals", "value": 100 },
      { "field": "sessions", "operator": "greater_than_or_equal", "value": 2 },
      { "field": "url", "operator": "not_contains", "value": "log-in" }
    ]
  }
}
```

`relationship=prospects` removes CRM-linked companies, customers, competitors, AI assistants and shared-network traffic in one go. The `url not_contains` rule drops existing users when your product lives on the tracked domain.

You can also start from your own target account list: `{"field": "domain", "operator": "in", "value": ["sendcloud.com", "..."]}`. Companies from the list that never visited are simply absent.

**3. Understand each company**

For each shortlisted company, `GET .../companies/{company_id}/score` gives the pillars, matched and missed rules, the pages viewed and what is unknown, and `GET .../companies/{company_id}/pages` the full page trail. Separate observed behaviour (pricing page viewed three times over two weeks) from interpretation (they may be comparing tools) and from missing information (nobody identified; no CRM record).

**4. Find the role that owns the problem**

`GET .../companies/{company_id}/contacts?persona_id=...` lists people persona-first. Fall back to `dmu` or `department` for small companies. Contacts from the database did not visit; identified visitors (`/visitors?company_id=...`) did.

**5. Act and learn**

Write the message around a plausible need and a useful insight, propose a small next step, and let Sales review it. Re-run the same search over the next window to see which accounts moved: `sessions`, `intent_score` and `score` per company are the progress signals.

> [!TIP]
>
> A prospect that reads pricing pages but has never been identified is a research task, not a sales email. The intent pillar tells you the topic; the people pillar tells you whether you know who.


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