Follow-up Automation

What follow-up means in this project

Follow-up automation in OrcaPulse is the part of the workflow that keeps outreach going after the first capture, qualification, or routing decision. It is how the product turns one lead event into a staged sequence of messages, calls, waits, and retries.

In the current project, follow-up is not a separate subsystem. It is built directly into workflow steps, execution scheduling, recall settings, and lead execution state.

Follow-up building blocks

The workflow model already supports the main outbound actions that follow-up needs: SMS, email, WhatsApp, AI call, and wait steps, with CRM sync and webhook sync available when downstream systems should be updated as part of the sequence.

These actions only work well when their required resources are already connected, such as templates, email providers, WhatsApp accounts, assistants, and phone numbers.

Waits, scheduling, and time windows

Timing is one of the most important parts of follow-up automation in this project. Execution delay spaces out the first action, wait steps pause the workflow between actions, and day and night mode can push execution into allowed business-hour windows.

The scheduler resumes pending leads when their execution time arrives, so a follow-up path can unfold over minutes, hours, or days rather than running everything immediately.

• Execution delay: controls when the first follow-up action should start after scheduling.
• Wait steps: hold the lead in a pending state until the next step should resume.
• Business-hour windows: day and night mode can shift execution into allowed hours.
• Timezone-aware behavior: scheduling logic uses the lead timezone so follow-up timing stays realistic.

Recall and retry logic

Follow-up automation is not limited to messaging. The workflow model also includes recall settings with maximum attempts and call-outcome conditions such as voicemail, not available, customer busy, and reschedule request.

This means AI call follow-up can continue intelligently instead of retrying every lead the same way. Different outcomes can map to different delays, assistants, or retry behavior.

• Recall enabled: workflows can reschedule contact attempts after certain call outcomes.
• Max attempts: follow-up should stop once the configured retry limit is reached.
• Outcome-based timing: voicemail, busy, no-answer, or reschedule-style results can shape the next attempt.
• Assistant-aware retries: recall settings can point follow-up calls to a chosen assistant.

Visibility and control

Follow-up automation is visible in the current project through lead execution state, scheduled execution time, recall state, call records, and timeline events. This is what lets teams understand whether a lead is still waiting, actively being contacted, completed, failed, or manually stopped.

Review Leads and Lead Hub are important here because follow-up often spans time. Teams need to see not only what happened once, but what is still scheduled to happen next.

• Lead state: pending, in progress, completed, failed, stopped, and recall-specific states show where follow-up stands.
• Timeline history: records sends, waits, calls, resumes, failures, and completion.
• Manual intervention: operators can stop execution or cancel recall when automation should not continue.
• Review surfaces: use Review Leads and Lead Hub to inspect long-running follow-up paths.

How to design the first sequence

The best first follow-up workflow in this project is usually short, paced, and easy to inspect. Start with one clear post-qualification sequence instead of trying to automate every channel and retry rule at once.

A practical starting point might be one wait step, one outbound message, and one AI call or CRM sync, followed by a simple recall rule only if the voice path matters.

• Keep the first version narrow: a short sequence is easier to test and tune than a large branching follow-up tree.
• Use pacing on purpose: add waits only where timing improves response quality.
• Avoid channel overload: do not fire SMS, email, WhatsApp, and call steps all at once unless that is truly intended.
• Review outcomes before scaling: confirm timeline events and lead states match the designed sequence before expanding it.

Next steps

After follow-up automation is stable, the next layer is deciding how much autonomy OrcaPulse should have in choosing actions or adapting the path dynamically instead of just following a fixed sequence.

That naturally leads into agentic automation and richer decision-making across the workflow lifecycle.