Automations
Someone comments a keyword on your post; they get a direct message with the thing you promised. That is the core of the product, and everything else on this page is a detail of it.
How one works
An automation is bound to one post. It watches that post's comments, tests each new comment against your keywords, and for the ones that match it creates a lead and queues a direct message to the commenter.
Automations start as drafts. Nothing is watched and nothing is sent until you activate one, which is deliberate: the first time you use a new template you should read what it produced before a stranger does.
- Each comment is only ever acted on once, by comment and by person.
- Every message goes through the same governance layer as everything else — caps, do-not-engage, de-duplication and your autonomy setting.
- The automation's own counters — comments seen, matched, messages sent, replies, leads — are on its page.
Keyword matching
Keywords are matched on word boundaries and without regard to case or accents. "GUIDE" matches "Guide" and "guiden"; it does not match "guidelines", which is the difference between a lead and a stranger receiving a message they did not ask for.
Negative keywords veto a match in every mode. "Send me the guide" fires; "no guide needed for me" does not.
| Mode | What it does |
|---|---|
any
|
One positive keyword anywhere in the comment is enough. This is the default and it is what almost everyone wants. |
all
|
Every positive keyword must appear. Useful for a two-word trigger like "free" and "guide". |
regex
Advanced
|
One regular expression that you write. Powerful, and the only mode where you can build a pattern that matches more than you meant. |
Templates and their placeholders
The message you send is a template with placeholders. They are filled from the commenter's profile and from whatever the automation is attached to.
| Placeholder | What it becomes |
|---|---|
{first_name} |
The commenter's first name. |
{last_name} |
Their last name. |
{full_name} |
Their full name as LinkedIn shows it. |
{headline} |
Their LinkedIn headline. |
{company} |
Their current company, when we have it. |
{title} |
Their job title, when we have it. |
{public_id} |
The last part of their LinkedIn profile URL. |
{lead_page_url} |
The lead page attached to this automation. |
{link} |
The tracked link attached to this automation. |
{keyword} |
The keyword their comment actually matched. |
{post_url} |
The post they commented on. |
Single or double braces both work, and the names are case-insensitive. A placeholder we cannot fill is removed along with the space around it, so a missing company name leaves a clean sentence rather than an empty bracket in a stranger's inbox.
Replying to the comment in public
An automation can also post a short public reply under the comment — "Sent, check your messages" — which does two things worth having: it tells the person to look, and it shows everyone else scrolling past that the thing is real.
The public reply spends the comment budget, not the message budget, and it is subject to the same pacing as anything else you post.
Follow-ups
You can attach one follow-up message, sent after a delay you choose, and only if the person has not replied. If they replied, it is cancelled — nothing is more obviously automated than a nudge answering a conversation that is already happening.
A follow-up is a message like any other: it spends a slot in the daily direct-message budget, it respects your sending hours, and it is refused if the person has since asked you to stop.
Auto-connect when you are not connected
LinkedIn only lets you send a free direct message to a connection. If the commenter is not one, an automation can send a connection request instead, with a note, and message them once it is accepted.
Connection requests are the most restricted thing on LinkedIn and they carry a weekly ceiling as well as a daily one. Turn this on knowing that it converts more slowly and costs from the tightest budget you have.
How many, and how often
Each automation can carry its own daily cap, and every message it sends also draws from the profile's shared daily budget. There is no separate allowance for automations: an automation, an agent and the API all spend from the same counters.
| Plan | Automations | Highest daily DM cap |
|---|---|---|
| Starter | 10 | 150 |
| Pro | 100 | 200 |
| Operator | Unlimited | 250 |
| Agency | Unlimited | 250 |
How quickly it happens
Comments are polled about every 10 minutes, jittered by up to 3, and a queued message then waits for its pacing gap and for your sending hours. In practice a reply lands minutes after the comment, not seconds.
That is a consequence of running in your own browser rather than a queue we can make faster. If your Chrome is shut, it lands when you open it.
Every cap, limit, timeout and tool name on this page is read from the running product when the page loads. If the software changes, this page changes with it.