Inside ThinkPost: How Our AI Coach Actually Scores Hooks

A transparent look at the methodology behind the hook scoring system — what signals we measure, why those signals predict performance, and what no scoring system can ever tell you.

When we built the hook scoring system into ThinkPost, we made a deliberate decision: we would explain exactly how it works. Not in marketing language, but in the actual detail. Because if you're going to use a score to decide whether to publish something, you should understand what that score is actually measuring.

This post is the full explanation.

Why hook scoring matters

The first line of your LinkedIn post determines whether anyone reads the second line. LinkedIn's feed shows roughly the first 200-250 characters of any post before cutting to "see more." That means your hook — your opening sentence or two — is the entire post for the vast majority of people who scroll past it.

Most creators have an intuitive sense of whether a hook is strong, but intuition is hard to act on systematically. When you've written a post and you're deciding whether to publish it or rewrite the opening, "my gut says this could be better" isn't actionable. A score with an explanation is.

The ThinkPost hook score is designed to give you a directional signal — not a definitive verdict — on whether your opening line is likely to earn engagement. Here's how we calculate it.

The five signals we measure

1. Specificity (0-20 points)

Specific hooks outperform vague hooks by a wide margin in our analysis of high-performing LinkedIn content. We measure specificity by looking for the presence of exact numbers, named entities, specific timeframes, and concrete situations versus abstract concepts.

"I've hired 23 salespeople in 6 years" scores higher than "I've hired a lot of salespeople over the years" because the specific numbers create credibility that the vague version can't. Our model recognises numerical specificity, named people or companies, geographic specificity, and temporal precision as markers of a high-specificity hook.

2. Tension or Curiosity Gap (0-25 points)

The most powerful hooks create a gap between what the reader knows and what they want to know. They imply that there's more to understand, that something unexpected happened, or that the conventional wisdom on a topic is wrong.

We score this by looking for structural markers of tension: counterintuitive juxtapositions (two things that shouldn't go together), unresolved questions, implied revelations ("here's what I found"), and contrast patterns ("everyone says X, but Y").

A hook that creates genuine curiosity — one where the reader needs to know what comes next — consistently earns more engagement than a hook that merely states an interesting fact.

3. Relevance to Professional Context (0-20 points)

LinkedIn is a professional platform. Hooks that signal professional relevance — that indicate the reader will learn something applicable to their work or career — outperform hooks that are interesting but not professionally actionable.

We measure this by assessing whether the hook signals a clear professional context, a specific role or challenge that the reader might recognise themselves in, or a business outcome the reader would care about.

4. Emotional Resonance (0-20 points)

The most-shared LinkedIn content consistently triggers an emotional response — not necessarily a dramatic one, but a genuine one. Recognition ("I've been there"), surprise ("I didn't expect that"), admiration ("that took courage"), or curiosity ("I need to know more").

Hooks that trigger recognition — where the reader thinks "I've experienced exactly this" — are particularly powerful because they create an immediate personal connection before the reader has even committed to reading the post.

We measure emotional resonance by looking for vulnerability markers, relatable professional situations, direct address of reader pain points, and aspirational outcomes.

5. Structural Clarity (0-15 points)

A hook can be specific, curious, relevant, and emotionally resonant and still be hard to parse in 1.7 seconds. The structural clarity score rewards hooks that are easy to read quickly — short sentences, familiar syntax, and a clear subject-verb-object structure.

We penalise for passive voice, overly complex clause structures, jargon, and hooks that require context the reader doesn't yet have.

How the scores map to performance

Based on our analysis of post performance across the ThinkPost user base and our broader dataset of LinkedIn content:

80-100: High-performing hook. Posts with hooks in this range consistently earn above-average engagement for their creator's typical performance. These hooks create strong curiosity gaps and contain specific, emotionally resonant language.

60-79: Solid hook. Will earn average engagement. May benefit from a single revision — usually to add more specificity or strengthen the curiosity gap.

40-59: Weak hook. Likely to underperform. Common issues: too vague, no tension, reads like a headline rather than an opening line, or requires context the reader doesn't have.

Below 40: This hook is actively hurting the post. The post may have excellent content but the opening won't earn enough attention to get the reader there. Rewrite before publishing.

What the score can't tell you

We're transparent about the limitations because they matter.

The score doesn't know your audience. A hook that scores 85 for a business strategy audience might score 60 for a technical developer audience. The signals we measure correlate with engagement across a broad LinkedIn audience, but every creator has a specific audience with specific preferences. Your historical data is the best guide to what works for your specific followers.

The score doesn't measure authenticity. You can write a technically high-scoring hook that doesn't sound like you, and your audience will notice. A hook that's a 70 in your genuine voice will often outperform a hook that's a 90 but sounds manufactured.

The score doesn't predict virality. Virality on LinkedIn depends heavily on timing, platform algorithm changes, whether a high-follower account comments early, and factors that no pre-publication scoring system can anticipate. The hook score predicts average engagement, not outlier performance.

The score is a signal, not a verdict. We built it to help you make a faster, more informed decision about whether to revise your hook before publishing. It's not a gate. Posts with scores of 55 can outperform posts with scores of 85. Use it directionally, not definitively.

How to use the score effectively

The most effective way to use the hook score is as a revision trigger, not a publication gate. If your score is below 60, treat it as a signal to spend 5 more minutes on the hook before publishing — not to abandon the post. Often, a single change makes the difference: replacing a vague phrase with a specific number, restructuring the sentence to put the tension first, or shortening a complex opening to its essential claim.

Over time, the most valuable thing the score does is calibrate your intuition. After you've seen the scores on 50 of your own hooks, you develop a faster sense of what a strong hook feels like. The score becomes less necessary because you've internalised the signals it measures. That's the goal — not dependency on the tool, but a better writer who uses the tool as a check rather than a crutch.