Playbooks · 5 min read
Course Launches Expire. Your Automation Does Not Notice
On this page
- A course business runs two clocks and most accounts build for one
- What the automation actually knows
- Build the launch automation like something you are going to throw away
- Build the evergreen one like something you will never look at again
- The window runs from the comment, not from the post
- The quarterly read-through
- What this will not do for a launch
Somebody comments on your reel in April. The reel went up in January, for a cohort that started in February and closed enrolment on the 9th. Thirty seconds later they get a DM that says enrolment closes Sunday, use the early-bird code, seats are limited.
Nothing malfunctioned. The template said what you typed into it and it will keep saying that until you go and change it. The automation has no idea what month it is, and it has no idea your cart shut nine weeks ago.
A course business runs two clocks and most accounts build for one
There is the launch, which is a few days long, high-volume, date-shaped, and full of numbers that stop being true. And there is the evergreen inbox, which never stops, is much slower, and asks the same eight questions forever.
Those are different automations. They live on different posts, use different trigger words, and one of them has an expiry date written on it. Building a single automation to do both is how a January price ends up quoted in July.
The other structural problem in this inbox — that people who already bought and people still deciding arrive in the same place and look identical to a keyword matcher — is a separate argument and it is written up properly here. This post is only about time.
What the automation actually knows
It knows a word matched, under a post you named. That is the whole of it.
There is no integration with your course platform. No cart state, no enrolment count, no seats remaining, no coupon validity, no payment link, no calendar. We do not make posts either — no scheduling, no captions, no hashtags, no analytics. What exists replies to comments and DMs, and it replies with the string you wrote, on any day of any year, until somebody switches it off.
A template is not a fact about your business. It is a photograph of your business on the day you typed it.
Build the launch automation like something you are going to throw away
Because you are.
- Scope it to specific posts. Under Advanced you can restrict an automation to particular posts. Use it. The launch automation should fire on the three pieces of launch content and nothing else, so the rest of your account is never quoting a closed cart.
- Give it its own trigger word. Not
INFO. Something you only ever used in this launch — the cohort name, the batch number. When you switch it off, no other post is depending on it. - Put everything in the one reply. A DM thread stays open to you for twenty-four hours and only their next message restarts it. There are no sequences here and no broadcasts. The deadline, the price, the syllabus link and the way to pay all go in the first message or they mostly do not arrive.
- Write the switch-off date in your calendar the same day you build it. Not after the launch. The day you build it, before you have forgotten which templates carry dates.
Build the evergreen one like something you will never look at again
Which, honestly, is also true.
The evergreen automation answers the questions that have the same answer in every season: is it live or recorded, what language is it taught in, how long is access, is there a certificate, how much time per week, can I see the syllabus.
Notice what is not in that list. No dates. No price. No batch number. No "next cohort starts". Every one of those is a fact with an expiry, and an evergreen automation is precisely the thing that will outlive it.
The window runs from the comment, not from the post
This is the detail that decides whether an evergreen automation is worth building at all, and it is usually explained wrong.
The seven-day comment window is seven days from that comment, not seven days from the post. Somebody who finds your January reel in April and comments today is answerable today. The reel does not age out. Its comments do, one at a time, each on its own clock.
So the evergreen automation is not a decoration. It is the thing that keeps working on a reel that is doing quiet numbers eighteen months after you made it — which is exactly why a stale price inside it is a live problem rather than a historical one.
The quarterly read-through
Twice a year, open every live automation and read the private replies as a stranger would. You are looking for one thing: a sentence that was true when you wrote it.
The usual offenders, in order of how often they turn up. A price. A batch date. A discount code. The words "starting next month". A link to a page that now redirects. A mention of a bonus you stopped including. And the one that catches almost everyone — "currently taking enrolments", on a course you quietly retired.
Keyword matching folds case and punctuation, so PRICE? and price are the same trigger. It matches whole words, so a trigger of price does not fire on pricing. It does not fold Indic vowel signs, so दाम and दम are different words to it. Worth knowing when you write the trigger; worth knowing again when you retire one.
What this will not do for a launch
It will not message anybody who has not messaged you first, which rules out the thing every launch wants on cart-close night. No broadcasts, no nurture sequences, no WhatsApp. No payment link, no cart, no course-platform integration of any kind. Captured people land in Leads and leave as a CSV that you take wherever your enrolment actually lives.
Templated replies are free and unlimited, so both automations above cost nothing to run. A credit is spent only when the AI writes a new reply. Free tier is 100 credits with no card, packs start at ₹499.
Takeover means that the moment you reply by hand in a thread, the automation stands down there permanently. During a launch that is how you should expect to work: the template opens, you close.
If you are picking trigger words, that decision has its own post. If you want the map of which course-business problem is written up where, start with the overview.
