# Enrichment

> Learn how enrichment fills in the missing details on the contacts, companies, and properties in your CRM.

Source: https://www.lev.com/docs/learn/originate/enrichment

Last updated: 2026-08-06

---
**Enrichment** fills in the details missing from the contacts, companies, and properties in your CRM, from a single phone number to a full picture of the transactions a firm has done. Every value it adds carries a confidence score and cites the sources behind it, so you can tell what's confirmed from what still needs a look.

## Enrichment on a record {#enrichment-on-a-record}

There are three main ways enrichment can help fill data gaps, depending on the missing value.

- **Field enrichment** handles the missing cells. An empty phone, email, or LinkedIn field isn't a dead end, and filling it happens in place on the record.
- **Research enrichment** covers the answers that are paragraphs rather than cells, for example the kinds of transactions a person buys, or whether a company has any legal or bankruptcy history.
- **Relational enrichment** fills the connections between records, for example a contact to their company, a company to its people, and both to the properties and transactions behind them, so one record opens the wider graph.

Each empty field on a record shows an *Enrich* affordance, and clicking it fills the value in place. For the answers that don't fit in a field, ask the **Lev Agent** in the chat on the record, and the research it returns is saved to the record as a note for the next person who opens it. Because that chat is scoped to the record, you can ask anything about the contact or company from any of its tabs without re-establishing context. Type `@` to pull a related company, contact, or property into the conversation when you need the wider picture.

## Confidence, sources, and feedback {#confidence-sources-and-feedback}

Enrichment is only as strong as the quality of the data behind it, so every value it fills carries its own quality signals. Each enriched value gets a **verification state**, either **Verified** or **Unconfirmed**, along with a per-field **confidence** score built from source quality, freshness, and how many sources corroborate it. Records also carry an **Enrichment freshness** score that shows how recently the record was updated and how many sources contributed to it. Together, these let you tell a corroborated phone number from a single-source guess before you act on it.

Hover any confidence score to see the breakdown behind it, including how many sources support the value. Every enriched email and phone shows its verification state and where it came from, whether that's a public filing, a data provider, or the agent's own research, so you can accept what's right and report what's wrong. Both signals feed back into future enrichment, which means correcting a record today sharpens the results you get tomorrow. You can also ask the agent for the underlying records, like a company's most recent transactions, and trace each one back to its source.

## Enrichment across a list {#enrichment-across-a-list}

Enrichment also works in aggregate, across every record in a list rather than one at a time. Set a rule on a column and it applies to the entire list, enriching that field for every record in the pipeline and for anyone added to it later.

Records are also scored for fit, ranked against your criteria across industry, keywords, size, and location, and hovering a score shows the breakdown behind the match. If you're importing your own list via a CSV or Excel file, the import maps your columns to the pipeline's fields, then enriches and scores every row the moment it lands.