> For the complete documentation index, see [llms.txt](https://support.parvenu.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://support.parvenu.io/features/enrich/icp-scoring.md).

# ICP Scoring

Score every contact in a list against your ideal customer profile

{% hint style="info" %}
ICP Scoring costs **0.1 Credits per contact scored**, including contacts outside your ICP's area. Contacts already scored with the same ICP are skipped for free unless you choose to score everybody again.
{% endhint %}

## What You Need

An ICP saved in Parvenu. Build one first on the ICP page, then point this enrichment at it.

## How to Score a List Against Your ICP

* Go to your [Lists](https://app.parvenu.io/app/lists) page and open your list.
* Click "Add Enrichment".
* Choose your ICP from the list.
* Under **Who to score**, keep **People not scored yet** or pick **Everybody**.
* Wait a few minutes, then sort the list by ICP Score.

Want to check contacts against several ICPs at once? Use Multi-ICP Score.

## Who to Score

* **People not scored yet** (the default) - Contacts that already have a score for this ICP keep it and are not charged again. If you edit the ICP, everyone is scored again, because it is asking a different question.
* **Everybody** - Score every contact on screen again, including ones already scored.

The same choice is in the API, the connector and the ICP Scoring Flow step.

## What It Fills

| Column          | What it holds                                       |
| --------------- | --------------------------------------------------- |
| ICP Name        | Which ICP the contact was scored against            |
| ICP Match       | Yes or no, based on the threshold                   |
| ICP Score       | 0 to 100                                            |
| Match Threshold | The score a contact has to beat to count as a match |
| Explanation     | Why it scored the way it did                        |

Click an Explanation cell to see one card per check: what it checked, whether it passed, and how many points it added to the score.

## How the Score Works

Your ICP has must-haves, like the type of company and the job title, and nice-to-haves.

* **Must-haves decide the match.** If a contact clearly misses a must-have, or there is not enough information to check it, it is not a match, however well the rest scores. The Explanation says which must-have stopped it.
* **Job titles need a clear fit.** Owner, General Manager or VP of Sales pass. A vague title like Manager or Vice President passes only when the company clearly fits your ICP.
* **Software and IT titles are not buyers.** A title built around a tool or system, like ServiceTitan Manager, CRM Admin or IT Director, does not count as a buyer.
* **No location? The website ending helps.** If a contact has no city, state or country, the company website ending is used instead, so .ca means Canada and .co.uk means the UK. Endings like .com, .co or .io don't name a country, so they are left alone.
* **Unknown company size is left out.** If the row has no company size, size does not help or hurt the score.
* **Empty rows are skipped for free.** A contact with no job title and no company is not scored and not charged.
* **Thin company info?** Fill in a company description first with Company Profile, Company From Domain or Scrape Website with AI, then score. More to go on means fewer contacts stopped for missing information.

## Use Cases

* **Work the best rows first** - Sort by score and give your best hours to the top of the list.
* **Trim before you spend** - Score first, drop the low matches, then run the expensive enrichments on what is left.
* **Sanity-check your targeting** - The Explanation column shows you what the model is keying on, which usually tells you whether your ICP is written well.