Rewarding reach
A post that gets seen and noticed is worth more than one that does not, and paying for the difference is the hardest part of an earned tier to get right. Four decisions, each of which can be gamed if made carelessly.
Views say how far a post went. Likes, replies and reposts say where it landed.
What do you measure?
| Signal | What it tells you | What faking it costs |
|---|---|---|
| Views | How far the post travelled | Pennies per thousand |
| Likes | A tap of approval | Pennies each |
| Replies | Someone typed something back | One account, replying to itself |
| Reposts and quotes | Someone carried it to their own followers | Needs accounts with followers |
| Saves | Someone meant to come back to it | Rarely worth anyone’s trouble |
| Follower count | Who the poster already is | Cannot be faked, and rewards nothing the post did |
The public signals a platform shows on a post, and what each one is worth to a faker.
Every one of these can be bought on its own, and every one says something the others do not. Reward a single metric and you have named the thing to buy. Reward several and you have to decide how they combine, and whether the rule still means anything on the next platform, where the signals differ.
Measured against what?
A thousand likes on a post about a product nobody has heard of is a different event from a thousand likes on a post about one everybody has. A fixed bar is out of reach for the first and routine for the second. Measure against follower count and you are paying for who the poster is.
Whatever you choose, two people need to predict it. The poster wants to know the bar before posting; the operator wants to know the cost before the month starts. Those pull against each other.
When do you look?
Measured in the first hour, a bonus rewards timing the algorithm. Measured a month later, it pays the poster after they have stopped caring. Platforms set a wall in between: X's public search reaches back seven days and no further. And the poster must be paid for the same window however long your verification took, or your own latency starts deciding what a post earns.
What is the most one post can cost?
The operator should know the worst case before it happens. But a hard cap on a linear rate makes a cliff, where a post seen ten times as widely earns the same, and a curve toward the cap is a second thing to explain and tune. Uncapped generosity gets clawed back after the first surprise; capped generosity survives success, if the cap is the right shape.
What does the poster have to understand?
The reward has to fit in a sentence, before posting and after. A formula reads as a trick, and a bar that moved between the post and the payout reads as a broken promise. The operator faces the mirror: a number they cannot price without data they will never have is a number they will adjust forever.
What it rewards, when it is right
A small account whose corner of the internet actually notices out-earns a big account posting into silence. Nobody is paid for who they are; everyone is paid for what the post did. That is why the reach bonus is half of the post's price rather than a gimmick: the flat claim value buys the post, and the bonus buys the part the poster cannot fake alone.