How to Manage DMs During a Launch Without a Team
Launch week does not break your inbox because the volume doubles. It breaks it because the volume doubles at the exact moment every message is worth money, and you are the only person who can answer any of them. Fifty conversations a week is a rhythm. Two hundred over four days, while you are also running lives and fixing a checkout page, is a queue with a leak in it.
The leak is not random either. The messages that get dropped are the ones that came in while you were on camera, and those are the warmest. This is a guide to the arithmetic and the rules that decide what you can actually hold, so you can decide what to cut before launch day decides for you.
TL;DR
- Meta's documentation is explicit that you cannot start the conversation: "Conversations only begin when an Instagram user sends a message to your app user". There is no compliant way to blast a launch offer to past enquirers.
- Group messaging is not supported either. One conversation is one person, so launch week is one to one work by design, not by choice.
- Your real ceiling is arithmetic, not effort: minutes you can spend in the inbox, divided by minutes a real conversation costs. Do that division before you set your cart dates.
- Triage in four buckets and answer them in order. If bucket one does not fit in your day, you are over ceiling and something else has to go.
- Messages sitting in the Requests folder with no activity for 30 days drop out of API calls, so the backlog you promise to handle "after the launch" has a clock on it.
Table of contents
- What actually breaks, and when
- Four documented rules that decide your launch week
- Your launch ceiling, in one line of arithmetic
- The four buckets, and the order you answer them in
- The week before: what to prepare while it is quiet
- Launch days: protecting the first reply
- The message you must not send
- The Requests folder and the thirty day clock
- After the cart closes
- When one person is genuinely not enough
- Frequently asked questions
- What to remember
What actually breaks, and when
A launch does not produce more of the same messages. It produces a different mix. Outside a launch, most of your inbox is people circling: they liked a reel, they had a thought, they will decide in three months. During a launch, a third of the inbox is people with a card in their hand and one unresolved question.
That change in mix is what makes the usual approach fail. Answering in the order messages arrived is fine when every message is worth the same. It is expensive when message forty is "can I pay in three parts" and message forty one is a compliment on your lighting.
The second thing that breaks is your own availability. Launch weeks are full of scheduled obligations: a live, a webinar, an email that has to go out at 6pm. Those are exactly the hours that generate the DMs, and exactly the hours you cannot answer them.
Verdict: plan for a mix problem and an availability problem, not a volume problem. Volume is the symptom.
Four documented rules that decide your launch week
Before any tactic, four facts from Meta's own documentation, consulted on 10 September 2026. They matter because most launch advice quietly assumes the opposite of each one.
| What the documentation says | Where it comes from | What it rules out on launch week |
|---|---|---|
| "Conversations only begin when an Instagram user sends a message to your app user" | Instagram messaging API | Opening a thread with everyone who asked about your programme last year |
| "Group messaging is not supported. An Instagram professional account can only converse with one customer per conversation" | Instagram messaging API | Any form of broadcast inside DMs |
| "Messages sent within the 24-hour window may contain promotional content", and promotional content outside it goes through Sponsored Messages | Messenger Platform policy | Pitching a cart deadline to a thread that went quiet a week ago |
| "Message text must be UTF-8 and be a 1000 bytes or less" | Instagram messaging API | Pasting your full sales page into a reply |
The 24 hour response window and the seven day human agent exception sit underneath all of this. We covered them rule by rule in what Meta actually allows when you automate DMs, so this guide does not repeat them.
One honest note on sourcing. Instagram's consumer Help Center is rendered in JavaScript and returns an empty page to a plain fetch, so nothing here is attributed to it. Everything above comes from the developer documentation, which is served as readable text and which anyone can check against the links.
Verdict: launch week is one to one, inbound only, and short form. Every plan that ignores that is a plan for a restriction.
Your launch ceiling, in one line of arithmetic
Here is the calculation almost nobody does before setting cart dates. It is arithmetic, not research: fill in your own two numbers and the rest follows.
Ceiling = minutes you can genuinely spend in the inbox each day, divided by minutes one real conversation costs.
The trap is in the second number. A conversation is not one message. A buying conversation on a 500 to 5000 offer usually runs four to six exchanges, spread over the day, each one requiring you to reload the context. Six minutes of total attention per thread is optimistic, not pessimistic.
| Minutes in the inbox per day | Minutes per conversation | Conversations you can close per day | Over a 4 day cart |
|---|---|---|---|
| 30 | 5 | 6 | 24 |
| 60 | 6 | 10 | 40 |
| 90 | 6 | 15 | 60 |
| 120 | 8 | 15 | 60 |
Now compare that last column with the number of buying conversations your last launch produced. If your cart generated 90 real questions and your ceiling is 40, you did not have a discipline problem. You had a capacity problem, and no amount of getting up earlier fixes a division.
Two ways to close the gap, and only two: reduce the minutes a conversation costs, or increase the minutes available. Everything in the rest of this guide is one or the other.
Verdict: do this division before you announce your dates, not after.
The four buckets, and the order you answer them in
Triage beats speed. Sorting takes seconds and protects the messages that pay for the launch.
- Card in hand. Asked about price, payment plans, deadlines, or whether it fits their situation. Answer first, same day, every day, without exception.
- One specific blocker. They want it and something is in the way: timing, a partner, a prior bad experience. Answer second, with one question back rather than a wall of reassurance. Our questions that qualify a lead before you book anything work here almost unchanged.
- Warmth. Congratulations, emoji, "this looks great". Real and worth honouring, but batched into one pass at the end of the day, one line each.
- Noise. Cold pitches, unrelated requests, spam. Archive without reply. It costs nothing and it removes them from tomorrow's count.
The decision rule that makes it useful: if bucket one does not fit inside your daily ceiling, you are over capacity today. Cut a bucket three pass, move a piece of content, or shorten the cart. Do not solve it by answering everything more briefly, because the brevity always lands on the wrong bucket.
The week before: what to prepare while it is quiet
The single highest leverage hour of a launch is spent before it starts, writing the eight or nine answers you will otherwise improvise ninety times.
- Write your price answer once, in full, and make it a message you would be happy to see screenshotted. Vague price answers generate three extra exchanges each, and at six minutes a thread that is where your ceiling goes.
- Write the "is this for me" answer as two questions rather than a paragraph.
- Write the payment plan answer, the deadline answer, and the refund answer as separate short messages, each under the 1000 byte limit.
- Decide in advance what happens to bucket two when you are live on camera for two hours. Silence is a decision too, it is just an unplanned one.
If you have never built a first reply that buys you time honestly, what an Instagram auto reply can and cannot do sets out the three layers and where each one stops. For the wording itself, DM scripts that book discovery calls is the closest starting point in this blog.
Not sure your launch is coming out of the right place at all? Booking calls from your DMs is the five step version of the funnel this guide assumes you already run.
Join the waitlist to hear when BookDM opens.
Launch days: protecting the first reply
Speed matters most on the first response and much less afterwards, because attention on a feed is borrowed and gets returned within minutes. The public research on that is old and was measured on phone calls to web forms, which is exactly why we read it carefully rather than repeating the headline in six numbers on lead response time.
Three practical consequences for a launch day.
Answer in blocks, not continuously. Three fixed passes beat sixteen interruptions, and they preserve the attention your live actually needs. Continuous monitoring feels responsive and produces worse answers.
Put the block right after the thing that generates the messages. Ten minutes after your live ends is worth an hour the next morning.
Never let the first reply be a holding message with nothing in it. "Hey, give me a sec" spends the opening and buys nothing. A first reply that answers the likely question and asks one back does both jobs. If your instinct is to automate that first minute, automating DMs without sounding robotic is the honest version of what that costs.
The message you must not send
Every launch produces the same temptation on the last day: sixty threads went quiet, the cart closes at midnight, and a single message to all of them would obviously work.
Two documented reasons it is not available. Group messaging is not supported at all, so there is no broadcast object to send. And promotional content is what the 24 hour window is for; outside it, the documented route for promotional messages is Sponsored Messages, which is advertising with an ad budget, not a free blast. The One-Time Notification token, which people sometimes reach for here, "can only be used once" by design.
What is left is legitimate and often better: answer the threads where the person actually spoke to you today, and let the cart deadline live in your email and your stories, where deadline messaging belongs. What to send when a conversation goes quiet covers the one to one version that stays inside the rules.
Verdict: if your launch plan depends on a mass DM, it depends on something Meta's documentation says does not exist.
The Requests folder and the thirty day clock
Messages from people who do not follow you land in Requests. During a launch that folder fills with exactly the people you most want: new audience arriving from a reel, an affiliate, a podcast.
It also decays. The documentation is specific: "Messages in the Requests folder that have not been active for 30 days will not be returned in API calls". So the backlog you promise yourself you will work through calmly in October is not sitting there indefinitely for any tool you connect later.
The practical version is a five minute pass through Requests once a day during the launch, sorting only into the four buckets above. Not answering, sorting. Bucket one gets pulled into the main inbox, the rest waits its turn.
After the cart closes
The week after a launch is where most of the remaining money is, and it is the week everyone spends recovering.
- Go back to bucket two, the specific blockers, and answer them properly now that you have time. Some of those blockers were real and have since resolved.
- Ask the people who said no what stopped them, in one sentence, with no pitch attached. It is the cheapest research you will ever run and it writes your next launch page.
- Count two numbers: how many buying conversations arrived, and how many you actually answered the same day. The gap between them is your ceiling, measured rather than estimated.
If that gap was large and you want to know what closing it would cost with a human, what an appointment setter really costs a solo operator has the public ranges.
When one person is genuinely not enough
There is a threshold, and most people meet it during a launch rather than in a normal week. Below it, response time is a discipline problem you can fix with blocks and prepared answers. Above it, it is a capacity problem, and discipline just changes which messages get dropped.
You are above it when bucket one alone does not fit in your daily ceiling for three days running. At that point the options are the honest ones: shorten the cart so the surge is denser but shorter, bring in a person for the week, or put something in front of the inbox that handles the first minute so that your minutes go to the conversations that need a human.
BookDM is being built for that last option: an AI setter for solopreneurs who sell in Instagram and WhatsApp DMs. It is not open yet, and the only thing you can do today is put your name down. If you would rather compare the tools that exist right now, our honest review of ManyChat for a coach who wants booked calls is the most useful place to start, and the pillar on what an AI setter actually does explains the category itself.
Join the waitlist, or read more on BookDM.
Frequently asked questions
Can I send one message to everyone who asked about my programme last year? No. Meta's documentation states that conversations only begin when the user messages you, and that group messaging is not supported. There is no compliant broadcast inside Instagram DMs.
How many DMs can one person really handle during a launch? There is no published number, and anyone quoting one is guessing. Do the division instead: minutes you can spend in the inbox, divided by the minutes a real buying conversation costs you. Six minutes per thread is a realistic starting estimate for a 500 to 5000 offer.
Should I automate my DMs for launch week only? Automation inside the 24 hour window is ordinary permitted use, and launch week is when it earns the most. What it should do is handle the first minute, not the sale. The layers and their limits are in what an Instagram auto reply can actually do.
What do I do with the Requests folder afterwards? Sort it daily during the launch rather than after it. Messages with no activity for 30 days stop being returned in API calls, so a tool you connect in November cannot rescue a backlog from September.
Is a lead magnet a good idea during a launch? It changes the mix of your inbox, which is the thing under strain. If you want conversation starters that do not add noise, lead magnet ideas that start conversations covers what tends to work for a solo practice.
What to remember
Launch week is one to one, inbound only, and short form, because that is what the documentation allows. Inside those limits, the only variable you control is where your minutes go, and the only honest way to plan that is the division: minutes available over minutes per conversation. Do it before you pick your dates. Sort into four buckets, protect bucket one, and let the deadline messaging live in email where broadcast is actually allowed.
And write your price answer this week, while it is quiet. It is the single message that will cost you the most minutes if you improvise it ninety times in four days.