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

# Understanding match confidence

> Learn what the four match confidence tiers mean, how Landbase determines them, and when you can act on each tier's results.

# Understanding match confidence

When you run `landbase-cli match person` or `landbase-cli match company`, the response includes a `tier` field that describes how confident Landbase is in the match. Understanding what each tier means helps you decide when to act on a result and when to treat it as uncertain.

***

## The four tiers

### `strong`

Landbase has found a unique, high-confidence match. Multiple independent signals align: name, company, domain, LinkedIn URL, title, or location all point to the same record. You can treat a `strong` match as reliable for CRM updates, enrichment, and automated workflows.

### `likely`

The match is probable but relies on fewer signals. For example, a name + company domain match where the name is not globally unique. A `likely` match is good enough for most practical purposes, but worth a spot-check if the downstream use is high-stakes (like sending email to the matched address).

### `plausible`

Landbase found a candidate that fits the input but could not fully disambiguate. For example, "John Smith at Acme Corp" where there are multiple John Smiths at similar companies. Use `plausible` matches with caution — treat them as leads to verify rather than confirmed records.

### `unlikely`

The candidate Landbase found is a weak match. The name or company partially aligns but multiple signals point in different directions. `unlikely` matches should generally not be used directly; they are included so you can see what the closest match was and decide whether to discard or manually review.

***

## How tiers are determined

Landbase evaluates a combination of signals when matching:

* **Name similarity** — exact match vs. partial match vs. common-name ambiguity
* **Company alignment** — does the domain, legal name, or LinkedIn match the supplied input?
* **Geographic consistency** — does the person's known location match what was supplied?
* **LinkedIn URL** — when present, this is a strong disambiguating signal
* **Title/department** — secondary signal, used to break ties between similar candidates

More signals → higher confidence tier. The tier reflects the quality of evidence, not a probability score.

***

## What to do at each tier

| Tier        | Typical action                                                                     |
| ----------- | ---------------------------------------------------------------------------------- |
| `strong`    | Use directly. Safe for CRM writes, enrichment, automated outreach.                 |
| `likely`    | Use for most tasks. Spot-check high-stakes cases.                                  |
| `plausible` | Review before using. Good for generating a candidate list for manual verification. |
| `unlikely`  | Discard or manually review. Do not use in automated pipelines.                     |

***

## Tier in the response

The tier appears in the `result` object:

```json theme={null}
{
  "result": {
    "tier": "strong",
    "match_reason": "Matched on LinkedIn URL and company domain",
    "candidate": {
      "member_name_first": "Daniel",
      "member_name_last": "Saks",
      "member_websites_linkedin": "https://www.linkedin.com/in/danielsaks"
    }
  }
}
```

The `match_reason` field gives a plain-English explanation of which signals drove the match — useful for debugging unexpected tiers.

***

## Related

* [match reference](/docs/reference/match) — full match command reference
* [How enrichment works](/docs/explanation/how-enrichment-works) — what to do after a successful match