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

# CRM Cleaning

> What CRM Cleaning changes on its own, what it asks you about, and what it reports

CRM Cleaning keeps your connected CRM tidy. It writes values in one standard format,
fills blank fields, links contacts to their companies, merges duplicates, and reports
the problems it cannot fix by itself.

It needs the CRM Cleaning add-on and a connected CRM (HubSpot, Salesforce or Attio).
Some fixes need HubSpot: see [capabilities by provider](https://docs.clustr-ai.com/crm-cleaning-capabilities.md).

## How it works

```
Scan your CRM  →  run every check  →  Do it: written to your CRM
                                   →  Ask me, or not sure: waits in Review
                                   →  cannot be fixed by Clustr: listed in Findings
```

1. A **scan** reads your CRM and runs every [check](https://docs.clustr-ai.com/crm-cleaning/checks.md) that is not
   Off. Start one by hand, or turn on the [schedule](https://docs.clustr-ai.com/crm-cleaning/history.md#scheduled-scans).
2. Each check has one setting: **Off**, **Ask me** or **Do it**. See
   [Modes and settings](https://docs.clustr-ai.com/crm-cleaning/settings.md).
3. What a check is certain about, under Do it, is written to your CRM straight away.
4. Everything else that can be fixed waits in **Review** for you to apply or dismiss.
5. Problems Clustr cannot fix go to **Findings**, with a link to the record in your
   CRM.
6. Every change is kept in **History**, and most can be reverted for 30 days, one at a
   time or a whole rule at once.

While the schedule is on, a contact or company created in your CRM is also checked for
duplicates within seconds, not at the next scan. It uses the same checks and the same
merge settings as a scan. See [new records](https://docs.clustr-ai.com/crm-cleaning/checks.md#new-records).

## What is automatic and what asks

Until you change anything:

| Done on its own | Asks you first | Only reported |
| - | - | - |
| Formatting: phones, emails, websites, domains, LinkedIn URLs, names, countries (HubSpot) | Linking a contact to its company | Records without an owner |
| Company LinkedIn URLs in one form | Setting a company's domain from its website | Deals without contacts |
| Merging duplicates whose match is certain (HubSpot) | Merging duplicates whose match is not certain, one pair at a time | Lifecycle anomalies |
| Filling blank fields, once you turn on enrichment | Replacing a value someone entered | Required fields that are empty |
| | Fixing a name an import damaged, like `lastSilva` | Possible subsidiaries, shared junk identifiers |
| | Fixing a typo in a mailbox domain, like `gmial.com` | Unusable values: junk names, unreadable phones, throwaway emails |
| | Removing emoji or symbols from a name | Duplicate deals |

Paid enrichment and the schedule are off until you turn them on. Every scan writes the
changes of the checks on Do it, including a scan you or your AI assistant start by hand.

Clustr never writes over an edit someone made in your CRM after the scan. A
formatting fix never changes what a value means: phone numbers never gain or lose a
digit.

## Review and Findings

The two lists never mix.

* **Review** holds changes waiting for a decision, grouped by kind, for example "Safe ·
  Format & casing · contact · phone". Apply or dismiss a whole group, or one change at
  a time. A waiting change that no scan finds again for 45 days closes on its own.
* **Findings** holds what you fix in your CRM: a record with no owner, a deal with no
  contacts, a stage that does not add up, an empty required field, a duplicate. Duplicates
  can be merged from here on HubSpot.

More in [Review, Findings and merges](https://docs.clustr-ai.com/crm-cleaning/queue-and-findings.md).

## Where things are

**CRM Cleaning** has these views in the sidebar:

| View | What it shows |
| - | - |
| **Review** | Changes waiting on a decision. This is where CRM Cleaning opens. |
| **Findings** | Problems to fix in your CRM |
| **History** | Every change, applied or not, revertible for 30 days |
| **Overview** | Completeness of your CRM, what cleaning did in the last 7 or 30 days, enrichment credits, and any changes that could not be saved |
| **Properties** | HubSpot only. Every property with how many records fill it, and the custom ones nobody fills. See [property insights](https://docs.clustr-ai.com/crm-cleaning/property-insights.md) |
| **Standardise values** | The spellings of a field's values, grouped, and the mappings that rewrite them to one value. See [standardise values](https://docs.clustr-ai.com/crm-cleaning/standardise-values.md) |
| **Runs** | Every scan, with **Scan now** and the schedule's On / Off switch |
| **Settings** | Modes, schedule, alerts, enrichment, duplicate rules, summary email recipients, recipes |

## Keep up to date

While the schedule is on:

* One [weekly email](https://docs.clustr-ai.com/crm-cleaning/health-and-summary.md#the-weekly-summary-email) every
  Monday sums up what was fixed and what is waiting, with a short
  [Worth a look](https://docs.clustr-ai.com/crm-cleaning/health-and-summary.md#weekly-summary-highlights) note when
  many new duplicates appeared or a field got emptier.
* An immediate email goes out only when changes fail to write to your CRM.

## Go further

* [Every check](https://docs.clustr-ai.com/crm-cleaning/checks.md), with examples and the format standard
* [Enrichment](https://docs.clustr-ai.com/crm-cleaning/enrichment.md) and its monthly budget
* [Scans, history and revert](https://docs.clustr-ai.com/crm-cleaning/history.md)
* [The API](https://docs.clustr-ai.com/crm-cleaning/api.md) and [the MCP tools](https://docs.clustr-ai.com/crm-cleaning/mcp.md) to drive CRM
  Cleaning from code or an AI assistant


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