agentsclimarketplace

Ab title test brief

Skill ur-grue/autopunk-media-skills/skills/youtube/analytics/ab-title-test-brief

394 free AI skills and 6 agents for media professionals — journalists, producers, podcasters, YouTubers. Quality-tested. Works with Claude, ChatGPT, Cursor, Codex CLI, and more.

Install
npx -y skills add ur-grue/autopunk-media-skills --skill ab-title-test-brief

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 16 stars16 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generates 4–5 title variants for A/B testing — each on a different click-psychology principle, with a rationale and a test hypothesis.

SKILL.md

7.0 KB, as published. Nobody here has run it

A/B Title Test Brief

What This Skill Does

Generates 4–5 title variants for A/B testing — each on a different click-psychology principle, with a rationale and a test hypothesis.

When To Use This Skill

  • A published video has lower-than-expected click-through rate (CTR) and you want to test whether a title change improves performance
  • You've finished a video and have multiple title directions you're considering before publishing
  • You want to build systematic title-testing habits and need a structured framework for each test
  • You're analyzing what title elements work for your specific audience and need controlled variants to learn from

What You Need To Provide

Required: The video topic and main argument or value proposition; the target viewer's primary reason to click (what problem does this video solve or what curiosity does it satisfy); your current title or working title. Optional: Channel niche and audience description; current CTR if this is an optimization test on a live video; titles that have performed well on your channel; your main competitor titles on this topic.

How the Assistant Approaches This

  1. Analyzes the current or working title to identify what click psychology principle it's using (or failing to use) — search intent match, curiosity gap, benefit promise, social proof, urgency, specificity.
  2. Generates 4–5 title variants, each deliberately testing one distinct principle: one optimized for search (exact-match language), one optimized for social/browse (emotion and curiosity), one with a specific number or data point, one with a "mistake" or "warning" frame if it fits the topic, and one that leads with the viewer's desire rather than the topic.
  3. For each variant: states the principle being tested, rates the estimated CTR potential (relative, not absolute), and notes which audience segment is most likely to respond to it. Ends with a testing hypothesis: which variant to test first and why.

Output Format

Table of 4–5 variants with columns: Title, Principle Tested, Why It Should Work, Likely Audience. Followed by a "Testing Hypothesis" section (1 short paragraph) recommending which variant to run first and what to watch for in the results. Total length: 300–450 words.

Quality Criteria

  • Each variant is genuinely distinct — different words, not just different word order
  • Each variant tests a clearly stated, different principle
  • Variants are all under 70 characters (YouTube title display limit)
  • Testing hypothesis names a specific first variant to test and explains why
  • At least one variant is optimized for search discovery, at least one for browse/suggested
  • Rationale column is specific — not "good for clicks" but "targets viewers already in-market for a solution"

Example

Input

Video topic: How to stop procrastinating on deep work tasks — specifically the psychological technique of implementation intentions Main argument: Writing down exactly when, where, and how you'll do a task increases follow-through by 2–3x compared to just intending to do it Target viewer: Productive people who struggle with procrastination on their most important tasks Current working title: "How to Stop Procrastinating on Deep Work" Channel niche: Productivity science and evidence-based work habits Well-performing channel titles: "Why Your To-Do List Is Making You Less Productive," "The Science Behind Getting Things Done"

Output

A/B Title Test Variants

#TitlePrinciple TestedWhy It Should WorkLikely Audience
1How to Stop Procrastinating on Deep Work (2025)Search intent matchExact-match query; year token adds freshness signal for returning searchersViewers actively searching for procrastination solutions
2The Procrastination Fix Scientists Actually RecommendAuthority + curiosity gap"Scientists actually recommend" implies mainstream advice is wrong; curiosity drives the clickBrowse/suggested viewers; responds to authority framing
3Write This Sentence Before Every Task — Procrastination DisappearsSpecific action + intrigueConcrete, mysterious instruction creates an irresistible "what sentence?" questionHigh social share potential; works in browse and Shorts clip
4Why You Keep Procrastinating on Your Most Important WorkProblem acknowledgmentSpeaks directly to the viewer's frustration rather than offering a solution upfront; validated by channel's top performersSelf-aware procrastinators who have tried other solutions
5I Tested the Science of Implementation Intentions for 30 DaysPersonal experiment + credibilityFirst-person experiment format + scientific term signals depth; "30 days" promises a real testAudience that trusts data and personal case studies over opinion

Testing Hypothesis

Start with Variant 3 ("Write This Sentence Before Every Task"). It creates the strongest curiosity gap of all five options, tests well against the channel's pattern of performing well when titles frame a specific, unexpected action. The phrase "Procrastination Disappears" is a benefit promise that could draw concern about overclaiming — monitor for comment pushback. If CTR improves but view duration drops, the title may be overpromising relative to the content; adjust the video's hook to deliver the "sentence" earlier.

If Variant 3 underperforms after 2–3 weeks, switch to Variant 4 — the problem acknowledgment frame has a direct analog in "Why Your To-Do List Is Making You Less Productive," the channel's strong performer.


Known Limitations

  • YouTube's built-in A/B title testing feature (currently available to some channels via YouTube Studio experiments) is the correct tool for this — it splits impressions between variants automatically. If you don't have access, manual title changes can be tested by switching after 2–3 weeks, but confounding factors (time of year, algorithm changes) affect reliability.
  • CTR estimates in the output are relative, not absolute. A "high CTR" variant for a 2% average-CTR channel looks different from the same variant on a 10% CTR channel. Calibrate expectations against your channel's own baseline.
  • This skill generates title variants, not thumbnail direction. CTR is a product of title and thumbnail together — title changes without matching thumbnail optimization are often incomplete tests.

Related Skills

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.