The Waiting Game Is Over: Why Modern Inbound Needs an Outbound Engine

For a decade the inbound playbook had one moving part you did not control: the prospect. You published, you optimized, and you waited for somebody to book a demo. Everything before that moment was invisible to you, and everything after it was late.

On LinkedIn the waiting is now optional. The people who care about the problem you solve rarely arrive as demo requests. They arrive as a comment under a post, or a reaction at eleven at night, and then they scroll on. The comment is the hand-raise; the reaction is a quieter version of the same thing. It is subtler than a form, it happens far more often, and most teams let it pass because nothing in their stack is built to notice it.

This post makes one argument: modern inbound is not passive, and it is not a separate motion from outbound. It is a stream of small public signals that needs an outbound engine behind it — something that watches the engagement, works out which engagers fit your ideal customer, and starts a conversation while the signal is fresh. The mechanics below are the ones Sendpilot's inbound campaigns run.

Where Inbound Actually Happens Now

A demo request is the last event in a process you did not see. Before it, the buyer read a few of your posts, reacted to one, commented on another, and compared you with two alternatives. The old model met them only at the end. The new model meets them at the first public signal, and on LinkedIn that signal is engagement on your own posts.

Run the arithmetic on your own feed. Say a post that does its job draws 40 engagers — fourteen comments and the rest reactions. In our experience most of them are peers, friends and habitual commenters, and somewhere between a handful and a dozen match the customer you actually sell to: right seniority, right industry, right company size. Multiply by two or three posts a week and a quarter produces a long list of named, ICP-fit people who came to you, with a topic already attached and no list to buy.

Why a demo request is the wrong thing to wait for

• It is rare. For every person who books, many more read, react and say nothing — and some of them buy from whoever reached out first.
• It is late. By the time someone asks for a demo they have usually shortlisted. Your window to shape the criteria closed weeks earlier, somewhere in the comments.
• It says nothing about topic. A form says "talk to me." A comment says which problem, in the buyer's own words, in public.

The Gap Is the Handover, Not the Attention

Most teams are not short of engagement. They are short of a mechanism that turns it into a conversation. The intention to work the comments is universal; the follow-through is not, and the failure is structural rather than a discipline problem.

• A notification feed is not a queue. No unworked state, no owner, no deadline. Yesterday's engagers have scrolled out of view by the time you look.
• Triage costs more than sending. Opening each profile to check seniority and headcount runs a couple of minutes a person. Forty engagers is over an hour before a single message is written.
• Non-connections stall. A good post travels beyond your network, so many engagers are second- or third-degree — and LinkedIn only lets you message your connections. They need a different first touch: a reply in the thread that asks them to connect, then the message once they do. By hand, that is the step that gets skipped.

This is the point where inbound needs an engine, not a reminder. The engine has four parts.

What an Outbound Engine Behind Inbound Looks Like

Most of it is machine work. Your part is the post, the Action Words, the three templates — the thread reply, the non-connection reply that asks them to connect, and the DM — and accepting the connection requests that second reply brings in. The templates should read like a person wrote them, because a person did, once.

1. Capture every engager, continuously

You point an inbound campaign at one of your LinkedIn posts. From then on it watches that post and captures everyone who comments on it or reacts to it, as they do it — not as a one-off export you run on Friday. An engager at nine on Tuesday morning is in the list at nine on Tuesday morning.

Action Words are the trigger for everything that follows. You set them when you build the campaign, the post asks for them — "comment PLAYBOOK" — and a comment containing that word is what starts the automation. Nothing else starts it: not a reaction, not a view, not a mention. So write the ask into the post, and pick a word that suits it: a resource name for a lead magnet, something like MORE for the long version of an opinion piece.

Everyone else is still captured. Reactors, and commenters who never typed the word, land in the campaign's lead list and get scored like the rest — they simply do not receive the automatic reply or the DM. That is the useful division of labour: the Action Word decides who gets answered today, and the score decides who is worth a sequence this week.

2. Answer in public, then in private

The reply template answers the Action Word comment in the thread. Reply templates are randomizable, so a thread with thirty replies from you does not read as thirty copies of one sentence — the fastest way to tell every silent reader that the engagement was transactional.

Which reply someone gets depends on one thing: whether the two of you are already connected. LinkedIn only delivers a DM between connections, so the campaign splits there. A commenter you are connected to gets the standard reply and then the DM. A commenter you are not connected to gets the non-connection reply, which asks them to connect with you first.

That second template is doing more work than it looks. It turns a public comment into a connection request you can accept, and the moment you accept it, the DM goes out — so the step people skip by hand happens on its own, in the thread, while the post is still live. It is also a filter. Someone who bothers to send the request wanted the thing enough to ask twice, and you end the week with a larger network as well as a conversation.

The DM itself carries first-name and full-name variables and nothing cleverer, which is the right amount of personalization for a message that goes out minutes after the comment — or minutes after the connection is accepted. Do not pitch in it. Deliver what the post promised, or say something about what they wrote, and stop. Their answer lands in Unibox alongside every other conversation, so nobody has to check a second inbox to see who replied.

3. Score every engager against your ICP

Engagement tells you what someone cares about. It does not tell you whether they can buy. That is the job of ICP Scoring, which reads the whole LinkedIn profile — title, seniority, industry and company size, plus skills, endorsements, network, experience and recent activity — and returns a 0–100 fit score for each captured engager.

Hold the two signals apart, because the engine depends on both. The post they engaged with is intent: a topic, a moment, a reason to write. The score is fit: whether that person and that company are the ones you sell to. A high-fit engager on a relevant post is the best lead you will see this week. A low-fit engager on the same post is a pleasant conversation and nothing more, and the score is what stops you spending an hour on it.

4. Transfer the high-fit ones into an outbound campaign

This is the step the title of this post is about. Engagers above your fit threshold are transferred into an outbound campaign — the same sequence builder you use for cold lists, now fed by people who already know your name. A sequence for a warm engager is shorter than a cold one: a connection request that names the post, a message a few days after acceptance that picks up the thread, and a Voice Notes step for the ones you most want to talk to. A voice note is recorded once and sent as a step in the sequence, so the highest-fit engagers hear a human voice without you recording forty separate notes.

The sequence can also include a profile view or a post like — small, low-pressure touches that keep you visible between messages. Every reply, from the first DM to the last voice note, arrives in Unibox.

A Week in the Engine

Here is the loop on a single post, with illustrative numbers. Substitute your own.

Notice where the human time goes: the post, the templates, the connection requests you accept, and the conversations that follow. Nothing else.

The Guardrails

An engine behind inbound has to be quieter than a cold campaign, not louder. Three rules.

• Sameness is the risk, not volume. Follow-ups from one post are a handful of sends against the limits Sendpilot enforces per account — about 25 connection requests a day, and no more than 75 messages a day, which is our recommendation rather than a LinkedIn-published figure. Use the full set of reply templates rather than one, so a long thread does not read as the same sentence pasted down the page.
• One follow-up, not five. This person engaged with a post, not a lead form. A five-step cadence gets you muted. A connection request, a message and, for the best fits, a voice note is enough.
• Suppress the obvious. Customers, competitors, colleagues, and anyone already in a sequence.

On account safety: every connected account in Sendpilot gets a dedicated, geo-matched proxy, automatic warm-up and enforced per-account limits. Used on its own inside those limits, it keeps accounts safe. Restrictions come from connecting one account to several tools at once, or from sustained abuse. LinkedIn has no official API for any of this and endorses no vendor; it acts on behaviour, which is exactly what the limits control. The connection request limits guide has the week-by-week ramp for a new account.

Content Becomes the Targeting

In the old model content was a magnet: it pulled traffic, and the form did the rest. In this model content does two jobs, and the second matters more.

1. 01It attracts the right people. A post about the problem you solve draws the people who have it. That was always true.
2. 02It labels them. The post someone engaged with tells you what they care about before any conversation starts. Someone who commented on your post about LinkedIn follow-up does not need the whole platform pitched at them. They need the answer to that one problem.

The loop closes on the editorial side. Read your inbound campaigns per post: the ones with the highest ICP-fit share are telling you what to write next, and the outbound conversations they start hand you the objections and the language for the post after that. That feedback is the whole point of running inbound and outbound as one All-Bound revenue engine rather than two teams sharing a logo. For the message-level detail — what the thread reply should say, and how the DM reads for someone who has just connected — the comment to DM playbook goes deeper.

The Bottom Line

Inbound still builds the trust. What has changed is the assumption that trust announces itself with a demo request. On LinkedIn it announces itself in the comments, dozens of times a week, and the teams turning that into pipeline are the ones with an outbound engine behind it: capture the engager, answer the comment, get the connection in place, score them against the ICP, and move the fit ones into a sequence while the post is still on their mind. Stop waiting for the hand to go up. It already did, in the comments.

Where Sendpilot Fits

Sendpilot runs the whole loop on LinkedIn from one workspace. Inbound campaigns watch a post and capture every commenter and reactor. A comment carrying one of your Action Words triggers a randomizable reply in the thread, plus the DM if you are already connected; if you are not, the non-connection reply asks them to connect first and the DM follows once they do, with first-name and full-name variables. ICP Scoring scores each engager from their full profile, and the high-fit ones transfer into outbound campaigns for connection requests, messages and Voice Notes, with every reply in Unibox.

Every plan includes inbound campaigns, ICP Scoring, outbound campaigns, Voice Notes and Unibox, priced per LinkedIn account — never per user. ICP Scoring draws on your plan's monthly credits — 800 per LinkedIn account, so 800 on Launch, 4,000 across Growth's five accounts and 20,000 across Agency's twenty-five — the same balance that covers enrichment, the Lead Database and the Lead Extractor, so count the engagers you expect to score when you pick a plan. A solo operator posting from one profile fits Launch ($79/month, or $66/month billed annually). A small team running the engine across several profiles fits Growth ($329/month, or $274/month billed annually, five accounts included), which also adds the API, webhooks and the MCP server — the route for handing high-fit engagers to your CRM or email tool, or to Zapier or n8n if that is where your hand-offs live. Agencies running it for clients fit Agency ($699/month, or $583/month billed annually, 25 accounts included). Sendpilot is LinkedIn-only, so keep your email tool for email. Start a free trial, or compare the plans on /pricing.

Frequently Asked Questions

Does this need website traffic or a form?

No. The signal is engagement on your own LinkedIn posts: comments and reactions are captured, and a comment containing an Action Word is what starts the automation. Sendpilot does not identify anonymous web traffic or watch what people do on your site. Its inbound is LinkedIn post engagement, captured and acted on inside LinkedIn.

What counts as engagement?

Capture and trigger are two different things. The inbound campaign captures both comments and reactions, so everyone who engaged with the post becomes a scored lead you can work. What starts the automation is narrower: a comment containing one of your Action Words. A reaction is not a trigger, and neither is a view or a mention. Treat a reaction as the weaker signal when you set your fit threshold — it is relevance, not a question — and let repeat engagers across several posts count for more.

Can the engine message people I am not connected to?

Not until they are connections — LinkedIn only delivers DMs between connections, and the campaign works with that rather than around it. A non-connection who comments gets the non-connection reply in the thread, which asks them to connect with you first; when they send the request and you accept it, the DM goes out. If you would rather make the first move, the captured engager transfers into an outbound campaign and its connection request step does the same job from your side, naming the post. Messages and Voice Notes follow once you are connected either way. On a post that travels beyond your network, teams we work with typically find a good share of the best conversations start on this path.

How does ICP Scoring decide who is high-fit?

You define the profile — industry, company size, seniority, titles and so on — and Sendpilot reads each engager's whole LinkedIn profile against it: title, seniority, industry and company size, plus skills, endorsements, network, experience and activity. The result is a 0–100 fit score. The score measures fit; the post they engaged with supplies the intent. Pick a threshold, run it for a month, and adjust it based on who actually replied.

How much does this add to my daily sending?

Very little. Follow-ups from a single post are typically a handful of DMs and a few connection requests, against Sendpilot's enforced limit of about 25 connection requests a day and a recommended ceiling of 75 messages a day per account. If you also run cold outbound from the same account, the engager sequence shares those limits, so give it priority. A warm engager is worth more than a cold name.

Do I need outbound campaigns to use inbound campaigns?

No, but you want them. An inbound campaign on its own captures every engager, replies to the Action Word comments, DMs the ones you are connected to and scores all of them. The transfer into an outbound campaign is what turns a captured engager into a sequence — the connection request for the people who never connected, the follow-up message, the voice note — and both campaign types are on every plan, so there is no reason to stop at the DM.