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

# Personalize Outbound with a Deepline Play

> Build a Deepline play that researches each prospect, derives a relevant angle, and writes reviewable outreach copy from structured account context.

[Human guide: work with your agent →](https://deepline.com/docs/human/plays/play-personalize-outreach.md)

> [!NOTE]
>
> Prepare a reviewed outreach file whose proposed messages and next actions
> follow supported evidence.

Use this play to turn structured prospect and account research into a specific outreach angle and a reviewable draft for each row.

Use it when the output needs to become a campaign. This page owns the
prospecting-to-draft workflow after you have a qualified account or signal. Use
[signal playbooks](https://deepline.com/docs/plays/signal-playbooks.md) to qualify a trigger first, and
[job changes and CRM cleanup](https://deepline.com/docs/plays/play-job-change-alerts.md) to detect and
review changes in existing contact data.

## When to use it

* Build a prospect list from an ICP definition
* Find decision makers at target accounts
* Map multiple roles at the same account
* Add public or private signals before writing outreach
* Create personalized openers or sequences from verified context
* Run local-business prospecting from a category and location

## Run it from an agent

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
/deepline-gtm Build an outreach list from accounts.csv. For each account, find the right decision makers, enrich verified work emails, add recent signals, and write a short personalized opener with the evidence used.
```

With Codex:

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
codex "Build an outreach-ready CSV from accounts.csv with contacts, verified emails, signals, evidence URLs, and personalized openers."
```

## Workflow

**1. Define the audience**

Start with either target accounts, an ICP description, or a local-business
query such as "dentists in Austin." Write the constraints before asking for
contacts.

**2. Find companies or accounts**

Use search, CRM exports, or a company database to create the account list.
Keep the first run small so you can inspect whether the audience is right.

**3. Find the right people**

Search by role and seniority. For account mapping, ask for a few specific
roles per department instead of every possible title.

**4. Enrich and verify contact data**

Resolve LinkedIn URLs, work emails, and other contact fields. Use waterfall
tasks for email and contact discovery.

**5. Stack signals**

Add the context that explains why now: job changes, hiring, funding,
competitor engagement, recent news, CRM status, or product usage context.

**6. Write outreach context**

Generate the opener or sequence from the selected signals. Include the
source so a rep can verify the claim before sending.

## Define the review output

The following fields are a suggested custom workflow contract, not a fixed
prebuilt API response. Author and validate the policy before a pilot.

| Column | Description |
| - | - |
| `company` | Target account |
| `domain` | Company domain |
| `contact_name` | Person to contact |
| `title` | Current title |
| `linkedin_url` | LinkedIn profile when available |
| `email` | Work email when found |
| `email_status` | Validation or confidence status |
| `signals` | Signals used for prioritization or personalization |
| `source_url` | Evidence URL for the strongest signal |
| `personalized_opener` | Short opening line or email angle |

## Pilot and review before activation

Run a representative five-row pilot before using a full file. For each row,
inspect the evidence, contact match, and generated copy. Generated openers and
sequences are drafts: check factual claims, campaign fit, suppression and
consent rules, and account ownership before any send or writeback.

Ask an agent to return a clear review file:

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
/deepline-gtm Use outreach.csv to build an outreach-ready pilot for five
representative rows. Enrich the account and person, preserve the strongest
evidence URL, generate a concise opener and three-step sequence draft, and
write outreach-ready-pilot.csv.

For each row, return personalization_signal, opener, sequence_step_1,
sequence_step_2, sequence_step_3, evidence_url, and a review reason. Do not
send messages or update the CRM.
```

## Useful variants

### Prospect list from an ICP

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
/deepline-gtm Find 25 B2B SaaS companies in the US with 100-500 employees. At each company, find the VP Sales and Head of RevOps, get verified work emails, and write prospects.csv.
```

### Account mapping

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
/deepline-gtm Map buying-committee contacts at each company in target-accounts.csv. Focus on RevOps, Sales, Marketing, and Finance leadership. Include LinkedIn URLs and verified emails.
```

### Local-business prospecting

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
/deepline-gtm Find med spas in Miami with websites, phone numbers, ratings, and owner or manager contacts where available.
```

For local-business source details, see [SMB data sets](https://deepline.com/docs/features/smb-data-sets.md).

### Multi-signal outreach

```text theme={"languages":{"custom":["/languages/deepline-prompt.json"]}}
/deepline-gtm For each account in target-accounts.csv, combine job changes, hiring, funding, and competitor engagement. Score the account, find the best contact, and write a 3-step sequence using only verified signals.
```

## Quality rules

* Ask for a representative five-row pilot first.
* Keep the ICP constraints visible in the prompt.
* Prefer role lists over vague phrases like "decision makers."
* Do not generate personalization from unsourced claims.
* Put evidence in a column, not only in the email copy.
* Separate list building from copywriting if the first pass needs review.
* Treat generated copy as a draft until a person checks campaign fit and
  account-contact safeguards.

## Related pages

* [Find decision makers](https://deepline.com/docs/plays/task-find-decision-makers.md)
* [Find work email](https://deepline.com/docs/plays/task-find-work-email.md)
* [Waterfall enrichment](https://deepline.com/docs/features/waterfall-enrichment.md)
* [Signal playbooks](https://deepline.com/docs/plays/signal-playbooks.md)
* [Lead scoring and signal discovery](https://deepline.com/docs/plays/play-qualify-lead-fit.md)

## Verify this task

Check recipient identity, evidence for each claim, suppression, ownership, and the review-only output. Keep sending as a separate action.

Ask your agent to explain a failed check before repeating work or expanding the scope.


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