Outbound & prospecting

Reach out when the timing is right

Choose the signals that matter, check the context, and decide when a conversation would be useful.

Copies a prompt for your AI assistant

View prompt

Start wherever you are

Bring something you’ve noticed or explore what to look for. Your agent helps you decide whether to act, wait, or skip.

Start with your AI assistant. Connect tools when needed.

Fictional example

You

An agency I spoke with asked me to come back after their website launch.

Agent

What do we know about whether that agreed follow-up moment has arrived?

You

I’ve only seen a launch announcement. I haven’t checked the original thread yet.

One possible decision after separating the signal from the evidence

Wait · check the original conversation

A reason to check, before a reason to send.

The announcement may connect to an agreed follow-up moment. It does not yet establish when to contact the person.

Possible signal: The agency announced a website launch after previously asking to reconnect after launch.
Decision now: Wait. Check the actual launch date, original request, right contact, and any newer conversations or contact restrictions.
Reach out when: The event and agreed follow-up are confirmed, the contact is eligible, and a useful message has been reviewed. Skip duplicates or withdrawn requests.

Reported by the user

A launch announcement and a recollection of a request to reconnect.

Still unverified

Original wording, actual event date, current relationship context, and whether someone has already followed up.

Next useful step

Read the original thread and current CRM context before drafting a message.

Context & open questions

Why the distinction matters: Publication date and event date can differ. An announcement is not proof of buying intent.
How to revisit it: This is a one-off review. Choose an owner and review condition if the decision stays on hold.
If a routine would help: Agree sources, frequency, owners, costs, notifications, and review rules first. Call it active only after setup is confirmed.

Your result will reflect your conversation. Keep it in your assistant or a document you choose.

How the tools fit together

01
Signal source
What happened and when
02
Capable + conversations
What the relationship tells us
03
Your agent + you
Reach out, wait, or skip
04
Chosen tool, if useful
Carry out the reviewed step

Choose tools together. Execution depends on the connections and actions available in your assistant.

Read the full guide

Start wherever you are

Bring a situation or work out what makes timing meaningful for your offer. No finished ICP or prospect list is required.

Figure it out together

Choose a meaningful signal
Explore what could make a conversation useful, then agree sources and evidence.
Check event and context
Verify identity, actual date, and freshness. Review conversations, ownership, active outreach, and restrictions.
Reach out, wait, or skip
Make the reason clear. Review outreach before it happens, or agree a condition for revisiting a wait decision.
Choose whether to repeat
Keep it one-off or agree a supported review routine. A saved brief does not activate monitoring.

Tools and connections

Your signal source: Provide event evidence. Check coverage, original dates, access, and costs.
Your AI assistant: Separate facts from inference, compare context, and shape a useful next step.
Capable: Provide relationship context and supported records for decisions, tasks, and confirmed activity.
Your outreach tool, if needed: Handle approved messages. Check current restrictions and actual stop rules before sending.

What you’ll keep

A useful result and a copyable working brief with your decisions, reasons, evidence, assumptions, actual status, pending actions, and next question. Bring it back when you’re ready to continue.

A useful first outcome

You can explain the evidence, why the signal matters, and whether to act, wait, or skip. Any action or routine has an honest, verified status.

When to revisit it

Review when signals become noisy, sources change, or your offer evolves. Recheck current context before acting on an old brief.

View the setup prompt

Copy into your AI assistant to begin

Coach me through deciding when a conversation could be useful, checking evidence, and taking a reviewed next step through supported tools. Help me choose whether to reach out, wait, or skip.

Check supplied context first. Reuse a situation, signal definition or working brief; confirm changes and ask the next unanswered question. If no direction is clear, ask: "What would make now a useful time to start a conversation?" Ask one useful question at a time, reflect answers and explain options. I may bring a situation or need help choosing signals. No ICP, list, campaign or customers are required. Produce an early provisional decision or signal brief with assumptions labeled.

Explore whom I could help, what would be useful, and the next step. Compare signals such as an agreed time to reconnect, a relevant role/company change, or a stated need. Choose the signals and number together without fixed thresholds or a mandatory checklist. Events, opens, clicks and engagement are not proof of buying intent.

For each signal agree source, relevant people/accounts, evidence, freshness and expiration. Separate event date from publication and discovery dates. Prefer original sources and verify identity. Flag missing dates, contradictions, uncertain matches, stale evidence and repeated reports. Republished news is not a new event. Show facts separately from inference about why they matter.

Before acting, check Capable context and actual outreach tools: prior conversations, promised follow-ups, ownership, customers/open opportunities, recent or active outreach, duplicate signals, bounces, opt-outs and do-not-contact status. CRM membership alone does not forbid contact; agree relationship scope while honoring suppressions. Use supported queries and complete relevant pagination; explain gaps. Hold uncertain identity or suppression cases. No sourcing or paid enrichment for existing CRM people without a separate explicit request.

Agree reach-out, wait and skip rules with reasons, and owners and review dates/conditions where agreed. A valid signal may still be irrelevant. Define duplicate checks or cooldowns. Try relevant, stale, duplicate and ambiguous examples before relying on the rules.

If contact is useful, choose channels and count, message, next step and follow-ups together. Do not assume email, a sequence or automatic enrollment. Ground drafts in verified facts without invented familiarity or implied knowledge of a problem. Agree sender, timing and time zone, reply owner, limits and cross-tool stop rules. For waiting, prepare an agreed review step only through supported tools; for skipping, keep the reason where supported.

Prefer existing tools. Choose signal/monitoring tools after understanding need, coverage, access and budget. Inspect capabilities and schemas before connecting or saving. Use official authorization, never pasted credentials; agree paid scope and limits first. Capable provides context, supported records and confirmed activity. Verify signal/task fields and agree another location for fuller evidence if needed. Capable does not send outreach or store message drafts. Its pause or tracking state does not prove an external sender stopped.

Before contact, scheduled sending or enrollment that can send, show exact recipients, reason, message, channels, sender, schedule, costs and stop rules for explicit approval. Recheck suppressions and active outreach before execution. Verify actual state afterward, distinguishing draft, scheduled, confirmed activity and pending manual steps. Accepted requests are not delivery. Inspect uncertain outcomes before retrying.

Ask whether a one-off check, review task or recurring routine is useful. For recurring work, agree sources, scope, frequency, time zone, owner, notifications, costs, expiration, duplicate handling, stop controls, and whether candidates or drafts are produced. Automatic sending is not the default. Configure only explicitly agreed supported behavior and test it. A signal, task, prompt or brief is not an active monitor. Promise future checks only after setup and status are confirmed; otherwise give a clear manual handoff.

Leave a reusable working brief with purpose; signals, sources and freshness rules; exact decisions and reasons; evidence versus inference; any reviewed draft; actual action/routine status; tools and limits; pending steps; next useful coaching question; and owners or review conditions only where agreed. Keep it here or in my chosen destination without implying automatic saving or cross-chat memory. On return, check changed context and current evidence before acting. Learn from relevance and responses without promising replies or revenue.

How did it go? Optional feedback

Reach out when the timing is right

Start wherever you are

A reason to check, before a reason to send.

Reported by the user

Still unverified

Next useful step

Example working brief

Signal source

Capable + conversations

Your agent + you

Chosen tool, if useful

Start wherever you are

Figure it out together

Tools and connections

What you’ll keep

A useful first outcome

When to revisit it