Creator

CTR Prediction

A Content Spy packaging tool that estimates click-psychology strength for content you have not published yet — hook, title, optional thumbnail description, optional creator history — then returns a 0–100 CTR Prediction Score. 6 AI credits.

Overview

CTR Prediction sits in Content Spy under Packaging & Click Optimization. It is built for unpublished packaging: you describe the hook and title (and optionally the thumbnail and channel history). Open Go Viral → Content Spy (creator account, desktop). Each run costs 6 AI credits. Compare mode supports 2–3 videos: cost = number of videos × 6 (for example 3 × 6 = 18).

This is a relative scroll-stopping / click-psychology score, not a real CTR %, not impressions, and not an A/B test. It does not invent YouTube Analytics.

How to run

  1. 1Sign in as a creator. Open Go Viral → Content Spy (desktop), or open the CTR Prediction form from the tool.
  2. 2Required: content type (Shorts vs long-form), hook / concept, and planned title. Optional: thumbnail description and a short creator-history note.
  3. 3From Content Spy with a loaded public or unlisted video, the tool can prefill title, description-as-hook, and channel name. Private videos are not supported.
  4. 4Run it (6 credits). Credits deduct only after a successful result. Read the overall score band, five sub-scores, then Strengths, Weaknesses, Risks, Improvements, and Platform Fit.
You receive 100 free credits when your token wallet is created. For a true pre-publish test, upload as unlisted and paste the link — or fill the form without a URL. Skipping the thumbnail description caps that sub-score at 55.

What it uses

Unlike CTR Audit, this engine does not pull views, likes, or a thumbnail vision pass. It scores the packaging you describe.

  • Content type: YouTube Shorts (short-form) or long-form video
  • Hook / concept — the opening promise (curiosity gap, surprise, outcome)
  • Planned video title

Optional — but they change the score

  • Thumbnail description: face, contrast, overlay text, colors. If omitted, Thumbnail is capped at 55 and flagged in Detected Issues.
  • Creator history: channel and audience fit. If omitted or “not specified”, Creator History is set to 50 (neutral) and called out in the breakdown.
There is no transcript or YouTube stats feed. If hook and title disagree (title promises X, hook describes Y), the model penalizes the mismatch. Prefill from a loaded video uses title, description snippet, and channel name only — still not Analytics CTR.

Scoring framework

Five integer sub-scores (0–100) combine into the overall CTR Prediction Score. Weights below are mandatory in the model — they are not a separate calculator in the UI.

Hook

35%
  • Curiosity gap, surprise, contrarian angle, concrete outcome
  • Emotional intensity in the opening promise
  • Penalty for generic filler (“hey guys”, no tension)

Title

20%
  • Specificity: numbers, timeframes, a clear benefit
  • Cognitive tension (secret, nobody, mistake) and a front-loaded hook
  • Penalty for SEO-only wording, vagueness, emoji/caps spam, or mismatch vs hook

Thumbnail

30%
  • Facial emotion, high-contrast / tension framing
  • Minimal overlay (about 1–4 words), focal clarity vs clutter
  • If not provided: cap at 55 (incomplete packaging)

Platform fit

5%
  • Shorts: instant hook, kinetic / shock / curiosity fit
  • Long-form: narrative or outcome promise that earns the click

Creator history

10%
  • Audience fit and credibility only if history is substantive
  • Otherwise 50 with an explicit “neutral track-record” note

What you get

The result is formatted for the Content Spy results panel.

  • CTR Prediction Score: XX / 100 plus a band: HIGH (90–100), HEALTHY / TESTABLE (70–89), WEAK / RISKY (50–69), LOW (below 50).
  • Score breakdown: Hook, Title, Thumbnail, Platform Fit, Creator History — each 0–100 with a one-line why.
  • Strengths, Weaknesses, Risks, Improvements, and a Platform Fit Analysis (Shorts vs long-form vs the packaging).
Scores are model judgments on described packaging. They are not official YouTube CTR and not a guaranteed click rate.

Case study

Composite documentation scenario — not a live client report. Scores below show the tool’s result UI; they are model judgments, not YouTube Analytics CTR.

Scenario

A productivity creator has a title and hook for an unpublished “I tested 5 focus apps for 30 days” video. They skip the thumbnail description to see how incomplete packaging is scored before designing the image.

Context (available inputs)

  • Content type: long-form
  • Hook / concept + planned title (required)
  • Thumbnail description: not provided → Thumbnail capped at 55
  • Creator history: not specified → History = 50 (neutral)
  • Cost: 6 AI credits

Tool result

Score breakdown

0–100

69avg
  • Hook84/100
  • Title76/100
  • Thumbnail55/100
  • Platform Fit80/100
  • Creator History50/100
CTR Prediction
Hook84 / 100
84
Title76 / 100
76
Thumbnail55 / 100
55
Platform Fit80 / 100
80
Creator History50 / 100
50
70

CTR Prediction Score

70 / 100

HEALTHY / TESTABLE — strong hook and title, but incomplete packaging: missing thumbnail description caps that pillar at 55.

Strengths

  • 30-day outcome loop in the hook.
  • Title names a concrete test and a count.
  • Long-form choice matches a series payoff.

Weaknesses

  • No thumbnail description — incomplete packaging.
  • Creator history left blank (neutral 50).

Improvements

  • Add a 1–4 word overlay + shocked face description.
  • Front-load the result in the first four title words.
  • Note typical CTR or niche if you have channel history.

Takeaways

  • This scores unpublished packaging you describe — not Studio CTR on a live video.
  • Skip the thumbnail and that 30% pillar cannot go above 55. Fill it before you treat the overall as “testable.”
  • Use CTR Audit on a published (or unlisted) competitor when you need metadata + vision, not a form.

Frequently Asked Questions

Is this my real YouTube CTR?

No. CTR Prediction is a 0–100 click-psychology score for packaging you have not published (or that you describe). It does not read YouTube Analytics and does not invent a CTR percentage.

How is this different from CTR Audit?

CTR Audit scores a published video’s packaging from metadata, spoken hook, thumbnail vision, and authority metrics. CTR Prediction scores hook + title (+ optional thumbnail text and history) for content that is not live yet.

How is this different from Click Probability in Content Lab?

Click Probability is a Content Lab score on a loaded video’s title and thumbnail. CTR Prediction is the Content Spy pre-publish form: you type the packaging, including Shorts vs long-form and optional history.

What if I skip thumbnail or history?

The run still completes. Missing thumbnail caps Thumbnail at 55 and flags it. Missing history sets Creator History to 50 (neutral) and says so in the breakdown.

How much does one run cost?

6 AI credits per prediction. Compare mode multiplies by the number of videos (2–3).


Next Steps