How do I use Make.com to automate AI calls?
You automate AI calls with Make by building a scenario that watches your lead source, cleans the data and sends each lead to the agent through an HTTP module, plus a second scenario that receives the post-call webhook. Outpero dials within 30 seconds of the lead arriving and returns the outcome, summary and captured fields for Make to route onward.
Why teams pick Make
Make, formerly Integromat, prices by operations and shows each scenario as a visual flow, which makes branching easy: different agents for different products, different routes for different cities, error handlers that retry or alert. For businesses sending a few thousand leads a month, it is often the more economical choice.
Scenario 1: lead to call
- 01Trigger module for your source: Facebook Lead Ads, Google Sheets, a Custom webhook for Google Ads or your website, or your CRM's watch module.
- 02Text functions or a Set variable module to normalise the phone number to +91 format.
- 03A Router if different leads should go to different AI employees, such as Hindi-speaking leads from Lucknow and Tamil-speaking leads from Coimbatore.
- 04HTTP, make a request: POST JSON with name, phone, record ID and context to the agent's lead webhook.
- 05Run once with a test lead carrying your own number.
Scenario 2: result back
- 01Webhooks, Custom webhook. Copy the URL into Outpero as the result webhook, and choose every call, completed calls only or voicemail only.
- 02Make a real call so Make learns the data structure.
- 03Route by outcome: qualified leads to the CRM and an alert, not interested to a suppression list, callback requests to a follow-up sheet.
- 04Add an error handler on CRM update modules so a failed write sends an email instead of silently dropping.
Scheduling matters
Webhook triggers run immediately. Polling modules run on the scenario schedule, which can be every few minutes or longer depending on your plan. For speed to lead, prefer webhook-based triggers wherever the source supports them, and use polling only for sources that offer nothing else.
Language routing example
A Router is the simplest way to match language to lead. If your form asks for preferred language, or your campaigns are city-targeted, route Chennai leads to a Tamil employee and Pune leads to a Marathi one. Every agent still switches to English mid-sentence if the caller does, so a wrong guess is not fatal.
Want to hear an AI voice agent handle this on a real call?
Follow-up questions
Is Make or Zapier better for AI calling?
+
Both do the job. Zapier is quicker to learn and has a larger catalogue of app triggers. Make is often cheaper at higher volumes and handles branching, routers and error handling more visually. If your team already uses one, stay with it. The Outpero side is identical either way, because both talk to the agent through webhooks.