How to Connect Instagram Keyword Triggers to Make.com

Been getting a few DMs asking how I set this up, so figured I’d just write it out properly instead of explaining it one message at a time.

Quick context: “keyword triggers” means someone comments or DMs a specific word (like “PRICE” or “GUIDE”) and it automatically fires off a reply, a DM, or whatever action you want — without you touching anything. That’s the whole magic trick behind half the “comment DEALS below” posts you see.

Here’s how I actually wired it up in Make.com.

Step 1: Connect your Instagram account first

Before touching scenarios, go to Connections in Make and add your Instagram (Business) account. It has to be a Business or Creator account linked to a Facebook Page — personal accounts won’t show up as an option. If you skip this and try to build the scenario first, Make will just keep asking you to authenticate mid-build, which is annoying.

Step 2: Create a new scenario and pick the right trigger module

Search for “Instagram” in the modules list. You’ll see a few options — you want either “Watch Comments” or “Watch Messages” depending on whether you’re catching comments or DMs. A lot of people grab the wrong one here and then wonder why nothing fires.

If you’re doing the classic “comment a word, get a DM” flow, you actually need both eventually, but start with Watch Comments.

Step 3: Add a filter for the actual keyword

This is the part people mess up the most. Don’t try to get clever with the filter logic right away. Just add a simple filter after the trigger:

Text contains → your keyword

Keep it lowercase and check “case insensitive” if Make gives you that option, because “PRICE,” “price,” and “Price :woman_raising_hand:” are all going to come through and you want to catch all of them, not just the exact match.

Step 4: Add the action — usually “Send a Private Reply” or “Send Message”

Once the filter passes, add your action module. For comments, “Send Private Reply” is the one that actually slides into their DMs — this is what makes the whole thing feel automatic and slick. For direct keyword DMs, it’s just a straightforward “Create a Message” type module.

Write your reply like you’re actually replying to a person. I’ll be honest, my first version sounded like a terms-and-conditions page. Nobody wants that in their inbox.

Step 5: Test it before you trust it

Comment the keyword yourself from a second account, or ask a friend to. Watch the scenario run in real time in Make’s execution log. This step feels tedious but it’s saved me more than once — the first time I built this, my filter was accidentally case-sensitive and it just… never triggered for half the comments.

Step 6: Turn on scheduling and let it run

Once it works, flip the scenario to “Immediately” (real-time) if your plan allows it, or set a short interval like every 1–5 minutes if you’re on a lower tier. Don’t leave it on manual run — that defeats the whole point.

A couple things I’d tell past-me before starting:

Instagram’s API has a 24-hour messaging window, so if someone doesn’t respond to that first automated DM, you generally can’t just message them again later without a specific tag. Plan your flow around that instead of finding out the hard way.

Also — start with one keyword, not five. Get it working clean, then expand. I tried to launch with a whole keyword list on day one and spent more time debugging filters than actually talking to leads.

If anyone’s stuck on a specific part of this setup, drop it below — happy to help troubleshoot since I’ve probably already hit the same wall.

Hey, not sure if I am missing something, but there is no “Create a Message” module… Only Create a Reply to a Comment. Can you elaborate?

handling instagram comment and direct message keyword triggers requires strict error boundaries around the twenty four hour messaging window. if an automated reply fails silently, users get confused and retry multiple times, triggering rate limits.

for our multichannel social distribution pipelines, routing drafted messages through blotato helped keep outgoing response queues organized across platforms. having scheduled buffers prevents accidental concurrent webhook bursts during peak engagement spikes.

are you adding a router module to tag repeat commenters before dispatching the response payload?