Best prompts for YouTube titles using video analytics interpretation

You will be given a compact analytics block labeled {ANALYTICS} containing: top search queries (with impressions and CTR), primary channel category, video length, and current title. Analyze which search terms have the highest impressions and CTR, then produce 8 title variants ranked by predicted CTR. For each title include: (1) predicted CTR uplift (%) vs current, (2) targeted primary keyword, and (3) a 6-word rationale. Keep titles <= 70 characters, put the primary keyword near the start, and avoid clickbait. Output only a JSON array of objects: {title, predictedUplift, keyword, rationale}. Example analytics will follow.

Given {ANALYTICS} containing average view duration (AVD), retention graph key drops (timestamp and % remaining), top audience segments, and current title, produce 6 title variants focused on improving watch time. For each title, add a one-sentence on-screen promise (what viewers get in the first 60 seconds) and a 20-character suggested first-line description. Prefer clarity over mystery. Output plain text list numbered 1–6 with the three fields per item.

Input {ANALYTICS} with top viewer countries, age brackets, language, and top-performing thumbnails. Produce 3 localized title variants for each of the top 3 countries. Keep each title culturally appropriate, include a localized keyword if available, and mark language. Return a JSON object keyed by country with an array of 3 titles (max 65 characters). Also provide a 1-line reason for each country's top choice.

Using the {ANALYTICS} block (current CTR, top suggested traffic %, top keywords, recent CTR trend), generate 6 title variants grouped into two A/B tests (A1–A3 and B1–B3). For each variant provide: headline, which test group, predicted CTR rank within group (1–3), and one-line reason. Finally, recommend which group to run first and why. Keep titles <= 75 characters and include top keyword in at least two variants per group.

Given {ANALYTICS} (current click-through on mobile vs desktop, top keywords, average title character length), output two lists: Short Titles (<= 50 chars) of 5 items optimized for mobile thumbnails, and Long Titles (60–100 chars) of 5 items optimized for search. For each title include one-line tradeoff note (why short helps or why long helps). Return a JSON object with keys 'short' and 'long' each containing arrays of {title, note}.

Use {ANALYTICS} including top thumbnail that had highest CTR, current title, and top-performing color/word cues. For the video, create 7 title and thumbnail-caption pairs. Each pair should be: Title (<=70 chars) and Thumbnail Caption (3–4 words). Also include a one-line note explaining why that pairing should increase CTR. Output numbered list 1–7.