Why the like counter lies to you
A like is the cheapest action on X. It costs the reader nothing: no thought, no words, no time. A thumb moves and the post scrolls away. That is exactly why the number tells you almost nothing about whether the post was any good.
The like counter is also lagging. It reports what already happened after a post spread, so it explains nothing you can change on the next one. And because likes are easy to earn, optimizing for them pulls you in a specific direction: toward posts that are pleasant, safe, and agreeable. Agreeable posts get a reflex tap and then get forgotten. Nobody screenshots them, nobody replies, nobody remembers the account behind them a week later.
The four numbers to watch instead
Each of these is honest for a different reason. Read the column that matters most: the trap, because every one of these can be misread the same way the like counter is.
| Number | What it tells you | How to read it | The trap |
|---|---|---|---|
| Impressions per post | Whether the first line landed | The only line guaranteed to be seen is the first one. Low impressions usually means a weak opener, not a weak idea. | It is heavily shaped by who already follows you, so watch your own trend, not a raw number. |
| Reply rate | Whether the post reached a person, not just a thumb | A reply costs effort. Someone had to stop and type. That is the strongest cheap signal you get. | Bait questions inflate it while teaching people to comment and never care. More on that below. |
| Profile clicks | Whether they wanted more of you | A profile click is a reader deciding whether you are worth following. It rises when your posts stay in one clear lane. | One viral off-topic post spikes clicks from people who will never fit. Judge it across a week. |
| Posts published | The only input you fully control | Every other number here is downstream of this one. Impressions and replies cannot happen on a post you never wrote. | Publishing more filler does not help. This number only leads the others when the posts are worth reading. |
Notice what is not on the list. Likes, and follower count. Both are lagging, both move for reasons you cannot trace, and neither tells you what to write next.
Why 12 replies beat 200 likes
This is the part most people get backwards, so it is worth being blunt about. A post with 12 replies almost always did more for you than a post with 200 likes. The mechanism is simple. A like is a reflex. A reply is a decision. To reply, a reader had to stop scrolling, form a thought, and spend the effort to type it out. That only happens when a post reached them enough to provoke a reaction.
The post of mine that started 3 real conversations this week had 14 likes.
The one with 220 likes started zero.
Guess which one I'm writing more of.
Replies are also where everything downstream begins: relationships, recommendations, the reader who clicks your profile because your answer was worth reading. A wall of likes leads nowhere. A handful of real replies is a handful of people who now know you exist.
The ten minute weekly review
Numbers only help if you look at them on a schedule and then do something. Ten minutes, once a week, same steps every time. You are not hunting for a trend line. You are looking for two posts that beat the rest and one honest sentence about why.
- 1Compare your posts only to your own postsNever measure yourself against a stranger's numbers. Account size is a confound you cannot see: a post with 500 replies on a huge account may be a flop for them, and 12 replies may be a hit for you. The only fair comparison is your posts against each other.
- 2Find your top two by repliesSort your week by reply rate, not likes. Pull the two posts that started the most real conversation. Look at the topic, the format, and the first line they opened with.
- 3Find your bottom two by repliesPull the two that landed flat. These matter as much as the winners, because the contrast is where the lesson lives.
- 4Write one honest sentence about what separated themNot a paragraph. One sentence. "The winners made a specific claim in the first line and the losers warmed up for three lines first." That sentence is next week's plan.
The vanity dashboard trap
There is a strong pull toward dashboards packed with charts: sentiment scores, best-time heatmaps, follower growth curves, engagement rates broken out by hour. They look serious. Most of them are the like counter wearing a suit.
Here is the test. For any number on the screen, ask what you would do differently this week if it doubled or halved. If the honest answer is nothing, it is not insight, it is decoration. A metric you cannot act on this week is not worth tracking, no matter how good the chart looks.
- Would change what I write: reply rate on my last five posts, so I write more like the two that connected.
- Would change nothing: my average engagement rate to two decimal places, which I will stare at and then post the same way anyway.
- Would change nothing: a sentiment score that cannot tell me which sentence to cut.
Watch replies because they signal a real reaction, not because a big number feels good. The moment you start manufacturing the number, it stops meaning anything, exactly like the likes you were trying to escape.