The automation platforms on the table, compared honestly — and the one that fits this project best.
Juan Felipe Ramos · July 2026
| Make.comVisual automation platform | ZapierMost popular automation tool | Lindy.aiReady-made AI assistant | n8nOpen-source automation | Custom botTailor-made code | |
|---|---|---|---|---|---|
| Cost at our volume | ~$17–25/mo | ~$69+/mo (per-task pricing) | $49–99/mo (credits) | ~$24/mo cloud · free self-hosted | ~$6/mo AI only, free VM |
| Ease of setup | Medium — visual, needs care | Easiest of the four | Easy — chat-style setup | Technical — semi-code | Built already, but code to maintain |
| AI reads & replies to emails | ✓ Gemini module, full prompt control | ✓ AI step, good control | ✓ built-in, but generic tone | ✓ any model, full control | ✓ full control + custom empathy |
| Gmail + Calendar depth | ✓ deep native modules | ✓ good coverage | ✕ basic booking only | ✓ full API access | ✓ direct Google API |
| Timezones + fair rotation | ✓ possible, hands-on formulas | ~ limited branching | ✕ not supported | ✓ full, but needs code | ✓ already built in |
| Execution logs / audit | Excellent — step by step | Good | Limited | Good | Custom dashboard built |
| Servers to maintain | None | None | None | Own server if self-hosted | Own VM + token upkeep |
| Main pro | Best balance of power, price and visibility | Mature, huge app library, very reliable | Fastest to start, AI-native | Most flexible, cheapest at scale | Cheapest, fully tailored, already working |
| Main con | Detailed hands-on configuration | Per-task pricing punishes our volume | Covers only part of this workflow | Needs technical skills + own infrastructure | One person maintains it — no vendor behind it |
| Verdict | Best fit | Solid but costly | Partial fit | Powerful but heavy | Cheapest, less trust |
Costs estimated at V2I's volume: ~900 replies/month (3 campaigns × 300). On mobile, swipe the table horizontally.
A visual automation platform: you connect apps like Gmail, Google Calendar and AI models in a flowchart called a scenario. Each step is a module; each time a module runs, it consumes one operation.
Used by 500,000+ organizations. SOC 2 certified, GDPR compliant.
The whole workflow is a diagram anyone on the team can open and read.
No servers to maintain. Make runs it, logs every step, and retries failures.
Answers the campaign email
Gmail module checks for new replies
Reads intent, tone, proposed times
Sends the email down the right path
Check team calendars → book the meeting → create Google Meet → send confirmation.
Send a warm reply with the answer and ask for their availability.
Cancel the calendar event if one exists, confirm kindly, close the loop.
Every run is saved in Make's execution history — full visibility of what was received, decided and sent.
This is the recurring operating cost — the Make.com subscription plus AI usage — paid monthly by V2I. ~10 operations per reply.
USD · one-time payment
for research, setup and testing
A one-time build fee, separate from the monthly operating cost above — which V2I pays directly, with no intermediary.
Study Make, connect Gmail & Calendar, build the main flow
Timezones, booking, fair rotation between the team
All branches finished, edge cases handled
Real emails through every path before go-live
The final week is fully dedicated to testing — already included in the timeline and in the fee.
Make is visual, but it is not plug-and-play. Every branch, filter and formula is set up by hand — and some problems only appear when the scenario runs.
"1 PM EST" must become the correct hour for each host. In Make this is chained date formulas — one wrong format books a meeting at the wrong time, silently. Every case needs manual testing.
Sharing meetings evenly across the team needs a Data Store with counters plus extra modules and filters. More modules mean more operations used and more points of failure.
Scenarios look fine while you build them. Real problems — a strange email format, a full calendar, an API hiccup — only show up in execution, and debugging means reading run logs afterwards.
At 300 replies × 3 campaigns per month, ~90% of the Core plan is used. Retries and edge cases eat the remaining margin — a strong month can force an upgrade to Pro.
Juan Felipe Ramos · Venture 2 Impact · July 2026