The exact automations we run for clients. Free.

Blueprint

1. 1

   Buy a Twilio number, port or forward your business line to it, and enable A2P 10DLC registration (approval takes 3–10 days).

2. 2

   Create a TwiML Bin that dials your real phone with a 20-second timeout and posts the DialCallStatus to a webhook.

3. 3

   In Make/n8n, add a webhook trigger. Filter for DialCallStatus = no-answer | busy | failed.

4. 4

   Look up the caller in your CRM by E.164 number. Create a contact if none exists and tag it 'missed-call'.

5. 5

   Send the SMS through Twilio's Messages API, then log the message on the contact timeline.

6. 6

   Add quiet-hours logic and an opt-out keyword handler (STOP/UNSUBSCRIBE) to stay compliant.

<Response>
  <Dial timeout="20" action="https://hook.make.com/YOUR_WEBHOOK_ID" method="POST">
    <Number>+15550100000</Number>
  </Dial>
</Response>

Messaging automations must follow consent rules (TCPA, A2P 10DLC, GDPR/PECR). Only message people who opted in, and honour STOP immediately.

Blueprint

1. 1

   Point your form at a webhook and normalise the payload (name, email, phone in E.164, source, UTM fields).

2. 2

   Deduplicate against your CRM by email and phone; merge UTMs onto the existing record.

3. 3

   Generate a single-use booking link with prefilled name and email.

4. 4

   Send SMS immediately, email at +2 minutes, then SMS at +1 hour, +1 day, +3 days, stopping the moment a booking webhook fires.

5. 5

   Route high-intent leads (budget field above threshold) to a human via Slack with a one-tap call link.

Messaging automations must follow consent rules (TCPA, A2P 10DLC, GDPR/PECR). Only message people who opted in, and honour STOP immediately.

Blueprint

1. 1

   Write an ideal-client rubric: budget, timeline, service fit, location, authority. Weight each 0–5.

2. 2

   Send the lead payload plus the rubric to the model with a strict JSON output schema.

3. 3

   Validate the JSON. On failure, retry once with the validation error appended, then fall back to 'needs review'.

4. 4

   Write score, reasoning and suggested reply to custom CRM fields. Move scores ≥ 18 to 'Hot' and alert sales.

5. 5

   Log token usage and cost per lead so you can see what qualification actually costs.

{
  "type": "object",
  "required": ["score", "tier", "reasoning", "first_reply"],
  "properties": {
    "score": { "type": "integer", "minimum": 0, "maximum": 25 },
    "tier": { "enum": ["hot", "warm", "cold", "disqualified"] },
    "reasoning": { "type": "string", "maxLength": 400 },
    "first_reply": { "type": "string", "maxLength": 320 }
  }
}

Messaging automations must follow consent rules (TCPA, A2P 10DLC, GDPR/PECR). Only message people who opted in, and honour STOP immediately.

Blueprint

1. 1

   Trigger on 'job complete' or 'invoice paid' in your CRM or field-service tool.

2. 2

   Wait 2 hours, then send a one-question SMS: 'How did we do, 1 to 5?'

3. 3

   Replies of 4–5 get your direct Google review link (g.page/r/…/review). Replies of 1–3 open a private feedback form and alert the owner.

4. 4

   Reminder at +3 days if no reply. Never ask twice for the same job.

Messaging automations must follow consent rules (TCPA, A2P 10DLC, GDPR/PECR). Only message people who opted in, and honour STOP immediately.

Blueprint

1. 1

   Watch a Drive folder for new video files. Extract audio with ffmpeg (mono, 16 kHz) to stay under upload limits.

2. 2

   Transcribe with timestamps. Chunk transcripts over ~20 minutes before sending them to the LLM.

3. 3

   Prompt for: 5 short-form hooks with timestamps, 3 LinkedIn posts, 2 X threads, 1 newsletter draft, in your brand voice.

4. 4

   Write outputs to a Google Doc for approval. On approval, schedule posts through the Buffer API.

Messaging automations must follow consent rules (TCPA, A2P 10DLC, GDPR/PECR). Only message people who opted in, and honour STOP immediately.

Blueprint

1. 1

   Export and clean your list: normalise phones, remove duplicates, drop anyone who opted out or lacks documented consent.

2. 2

   Segment by last interaction, service bought and lifetime value.

3. 3

   Generate a personalised opener per segment with the LLM, then human-approve every template.

4. 4

   Send in throttled batches (e.g. 50 per hour) to protect number reputation, with STOP handling on every message.

5. 5

   Let the AI qualifier handle replies and hand hot conversations to a human or the booking link.

Messaging automations must follow consent rules (TCPA, A2P 10DLC, GDPR/PECR). Only message people who opted in, and honour STOP immediately.