How Do You Score a LinkedIn Post Before Publishing?

Paste your draft into Flux's Score My Post page and get a predicted engagement multiplier calibrated against your own last 200 posts — not a generic rubric, not someone else's audience. The model is trained on your posting history, accounts for your growth trajectory, and weights recent posts more heavily — so it predicts how a draft will perform relative to your median, returning a multiplier like 0.7x (below your typical post), 1.4x (solidly above), or 2.1x (a potential standout). Scoring is free and unlimited, so you can revise and re-score until the prediction lands where you want it.

Edge

What you'll learn

How Score My Post works and what the model measures
Walking through the Score My Post page step by step
How to read the score breakdown — what helps and what hurts
The revise-and-re-score loop that turns a draft into a stronger post
Scoring drafts through Claude or Cursor with the MCP tool
How coaches use scoring live on client calls

How does the scoring model work?

Most draft-scoring tools grade against a generic checklist — readability, emoji count, hashtag density. Flux's scoring is different because the benchmark is you. The model trains on your own last 200 posts, applies detrending to account for your growth trajectory, and weights recent posts more heavily so the prediction reflects your current audience, not who followed you two years ago.

The output is not a letter grade or an abstract score out of 100. It is a predicted engagement multiplier against your own median. A 1.0x means the draft should perform about as well as your typical post. A 1.4x means it should land in your top tier. A 0.7x means something is working against you and the draft is worth revising before it goes live.

Because the model is calibrated to your data, the same draft text would get a different score for two different people. A post about hiring strategy might predict 1.8x for a recruiting leader whose audience rewards that topic, and 0.9x for a marketer whose audience does not. That is the point: the score tells you how your audience is likely to respond, not how posts perform in general.

How do I use the Score My Post page?

Open the Score My Post page at /score in the Flux app. The workflow is three steps:

Paste your draft. Drop the full text of your post into the input field. Write it in your normal LinkedIn voice — the model has seen 200 of your posts, so it knows what that sounds like.
Read the results. Flux returns four things: a predicted engagement multiplier, a factor breakdown showing what is helping and what is hurting, a list of similar past posts from your history, and specific improvement suggestions.
Revise and re-score. Adjust the draft based on the feedback, then score again. There is no credit cost and no limit on re-scores, so treat this as an iterative loop, not a one-shot judgment.

The similar past posts are especially useful. They show you which of your published posts the model considers closest to this draft, along with how those posts actually performed. If the model says your draft resembles a post that earned 2x your median, you are on the right track. If the closest match underperformed, the improvement suggestions will point to what to change.

What does the factor breakdown tell me?

The factor breakdown is the most actionable part of the score. It separates the prediction into the specific dimensions driving it up or down, so you know exactly where to focus a revision.

Factors that commonly show up as helping a draft include topic alignment with your high-performing themes, format choices that match what your audience rewards, and hooks that resemble your strongest openers. Factors that commonly hurt include topics your audience has historically ignored, length mismatches (too long or too short relative to your sweet spot), and weak openings that do not match the patterns in your best posts.

The factor breakdown answers a question that a single number cannot: why is this draft predicted to land where it does? A 0.8x score with a clear "topic mismatch" factor tells you to shift the angle, not rewrite the entire post. A 1.3x with "weak hook" tells you the substance is strong but the first two lines need work.

How should I revise and re-score?

Treat the score as a diagnostic, not a verdict. A low multiplier is not a reason to abandon a draft — it is a signal to adjust the specific factors pulling it down.

A practical revision loop looks like this:

Score the draft and read the factor breakdown.
Pick the one or two factors hurting the prediction most.
Revise just those elements — tighten the hook, shift the angle toward a proven topic, adjust the length.
Re-score. Check whether the multiplier moved and whether the factor breakdown shifted.
Repeat until the prediction reflects a post you are confident publishing.

Most drafts improve meaningfully in two or three passes. The free, unlimited scoring means you are not rationing attempts — you are iterating toward the strongest version of the post.

Can I score drafts through Claude or Cursor?

If you use Flux's MCP server with Claude Desktop, Claude Code, or Cursor, you can ask your AI assistant to score a draft without leaving your editor. Ask Claude something like "score this LinkedIn draft for me" and paste the text. The assistant calls Flux's scoring tool and returns the same prediction, factor breakdown, similar posts, and suggestions you would see on the Score My Post page.

This is useful when you are already writing in an AI-assisted workflow: draft in your editor, score inline, revise, and re-score without switching to the browser. Coaches working in Claude can score a client's draft mid-conversation and talk through the factor breakdown in real time.

How do coaches use scoring on client calls?

LinkedIn coaches and content strategists use Score My Post as a live feedback tool during client sessions. The workflow is straightforward: the client shares a draft, the coach pastes it into the Score My Post page (or asks Claude to score it), and the factor breakdown becomes the agenda for the revision conversation.

This replaces subjective "I think the hook could be stronger" feedback with calibrated evidence: the model says the hook is the weakest factor, and here are three of the client's past posts with strong hooks for comparison. The similar-posts list gives the coach concrete examples to reference, drawn from the client's own history rather than generic best practices.

Because scoring is free, the coach can re-score revisions live on the call and show the client the multiplier moving as they improve the draft. That immediate feedback loop is more convincing than any amount of advice.

Frequently asked questions

Does scoring cost credits?

No. Score My Post and the MCP scoring tool are completely free with no credit cost, so you can score as many drafts as you want without affecting your monthly allowance.

How many posts does Flux need before scoring is accurate?

The model trains on your last 200 posts. If you have fewer than 200, it works with what is available, but the prediction becomes more reliable as your post history grows. Profiles with very few posts will see wider confidence ranges.

Can I score a draft for a client's profile, not my own?

Yes. If you manage multiple profiles on an Edge plan, you can switch to the client's profile and score against their posting history. The prediction calibrates to whichever profile is active, so the same draft text will return different scores for different profiles.

Does a high score guarantee the post will perform well?

No. The score is a prediction based on your historical patterns, not a guarantee. External factors — LinkedIn algorithm changes, trending news, time of day — all affect real performance. But a draft that scores well against your own baseline has a meaningfully better chance than one that does not.

What is the difference between Score My Post and suggest_angles?

Score My Post evaluates a finished draft you have already written. The suggest_angles tool is an earlier step — it generates post ideas from your top-performing topics before you start writing. The typical workflow is: use suggest_angles to pick an angle, write the draft, then score it.

On this page