Comment-to-DM against link-in-bio
Two ways to get an interested stranger from a post to an offer. One of them is what nearly everyone does and the other is what the platforms reward. The gap between them is not a matter of taste, and it is worth about as much as anything else on this page.
The two routes, side by side
| Link in bio | Comment-to-DM | |
|---|---|---|
| What you ask for | Leave the app | Type one word |
| Steps for the reader | Four or five | One |
| Effect on reach | Suppressed — the post sent people away | Boosted — the post generated replies |
| Time to delivery | However long they take to get there | Seconds |
| What you end up with | A click you cannot follow up | An open conversation |
The row that decides it
Reach. A post that sends traffic off the platform is a post that cost the platform a session, and distribution is trimmed accordingly — so the harder a link post sells, the fewer people ever see it. Replies point the other way: they are the signal most strongly associated with something worth showing to more people.
Which means the two routes do not merely differ in conversion. They differ in whether success shrinks or grows the audience it came from, and that difference compounds every time you post.
Where link-in-bio still wins
Two cases, and they are real. When the audience is already yours and already warm, the extra steps cost little. And when whatever you are sending cannot be delivered in a message — a booking, an application, anything requiring a form — the detour is the point rather than a leak.
For cold reach on a new account, neither applies.
Picking the trigger word
Judged against the failure it prevents rather than by any rule:
| Choose a word that is | Because otherwise |
|---|---|
| Short and spelled one way | Half the triggers never fire |
| Tied to what arrives | The thread reads as noise instead of demand |
| Rare in normal speech | You message people who were only chatting |
Automated against manual, at low volume
At fifteen comments a post there is nothing automation improves, and answering by hand teaches you the actual questions people bring — information no tool reports and every later piece of writing depends on. Build the automation when the manual version has become impossible, and not one post before.
Setting up tooling for volume that does not exist is one of the more convincing ways of avoiding the work.
Automated against manual, once it opens
The volume problem lives entirely in the first message. Automating past the first real answer adds no leverage and costs the only thing that made the exchange work, so the sensible boundary is: automated up to the handover, a person after it.
A slow reply is forgiven. Realising mid-exchange that the other side was a script is not, and it turns a warm contact cold in a way nothing recovers.
What you send, against how you sent it
The mechanic delivers to somebody who raised their hand, and that is the whole of its contribution. Whether it was worth raising a hand for is decided entirely by what arrives, and no amount of clever triggering compensates for a thin thing with a pitch stapled to it.
The version that works ends with one open question about their circumstances — specific enough that answering means describing where they actually are. Without it the whole apparatus is a delivery service. With it, it is the start of something.
The affiliate route this powers · All six routes · Starting with no audience
Checked 24 September 2026.