Playbooks · 5 min read
Launch Day on Instagram: Automating the Flood, Not the Post
A launch is the one day your Instagram behaves differently. Normal weeks arrive at a rate you could handle by hand if you had to. A drop does not — the comments land in a burst, they are almost all the same three questions, and the people asking will decide within the hour.
One clarification before the setup, because it decides what this post is for. Nothing here posts anything. We do not publish, schedule, or write captions. You post the launch yourself, at the time you chose, with words you wrote. What this handles is everything that arrives afterwards.
The week before: build it against last time
Do not build the launch automation on launch morning. Build it now, against the comments on your last drop, because those comments tell you the exact words people will use again.
- One keyword, short, in the caption.
DROP,LINK,PRICE. Whatever you put in the caption is what people will type, so pick it before you write the caption rather than after. - Match loosely. Case and punctuation are ignored already. Beyond that, expect
drop pls,Drop!!,dropp. Match on the word appearing anywhere in the comment. - Negative keywords, seriously. Launches attract the loudest opinions of the year.
scam,overpriced,worstshould mean do not reply, and you will want to have decided that in advance rather than at 10pm. - A public reply that is a receipt. "Check your DMs." Not the price, not the link. The public reply is a signal to everyone else that this post answers people.
- A private reply that answers and asks. Price, link, and then a question. The question is what earns you a second turn, because their reply restarts the twenty-four hour clock.
- Test on yourself. The whole pipeline runs and lands on your own account. On a launch there is no second chance to fix the wording.
The three settings that decide whether launch night goes well
The hourly cap. This is the one. A drop that travels produces comments faster than any sane sending rate, and without a cap the replies go out as fast as the API accepts them. A few hundred people get a near-identical DM inside an hour, some of them report it, and the account picks up a delivery problem that outlasts the product.
With a cap, the same burst becomes a queue. The person who commented at the peak gets their DM a while later, still inside the seven day comment window, and nothing about the account changes.
Quiet hours. Most drops go up in the evening, which means a chunk of the queue would otherwise send at 2am. Held until morning is better: the reply lands when they are awake, and it does not read as a machine working the night shift.
First-time senders only, off. This is the exception to normal practice. On a launch the people you most want to answer are often your existing regulars, and the setting that stops repeat greetings on daily content will suppress exactly the wrong people on drop day. Turn it back on afterwards.
The model on launch night, and the queue that catches it
The high-volume questions on a drop are templates — price, link, sizes, delivery. Free and unlimited, no credits, and they will handle most of what arrives.
What the model earns its credit on is the leftovers, and a launch generates unusual ones: "will you restock the tan one", "is this the same fabric as the last one", "I ordered the previous version, is this an upgrade". Those are real questions with your-catalogue-specific answers, and a template cannot anticipate them.
Replies the model is not confident about wait for a person instead of going out. On a normal day that is a safety feature. On launch night it is a live report on your own copy, and the single most useful hour of the evening is spent reading it.
The restock question deserves its own handling. A RESTOCK keyword that captures to Leads turns a sold-out disappointment into a list you can email when stock lands — which is the correct channel for that follow-up, because the twenty-four hour DM window will be long closed by the time the goods arrive.
Sold out is not a dead conversation. It is the highest-intent lead capture you will get all quarter, and it expires unless you write it down.
Day two: read the blocked list
The morning after tells you more than the night itself. Open Activity, filter to blocked, and expect this shape:
- Rate budget hits. The cap worked. This looks alarming in a log and is the good outcome.
- Duplicates. Instagram redelivering the same comment. Ignore.
- Closed windows. Comments on your older posts, aged past seven days. Normal.
- Takeovers. You answering by hand mid-flood. Fine, and if the count is large your automation should be narrower.
What you are hunting for is the fifth category: replies that were stopped and should not have been, or worse, replies that went out and should not have. Fix those before the next drop rather than during it.
After the launch
Export the leads. Everyone who typed your keyword is in there with the keyword attached, which means your drop produced a segmented list as a side effect. That list is the input to the next launch, and it is the part of launch day that keeps paying.
Then put your settings back — first-time senders on, cap where it was — and leave it alone. If you are building keyword triggers for the first time, the matching rules are worth reading properly. For a drop announced through stories rather than the grid, the story trigger feeds the same inbox on a different clock, and for the wider scope of what any of this does for a store, start with the overview.
