How We Make Clicky Thumbnails Using Ai and Psychology.

Main ideas & lessons

• Progress in YouTube thumbnails

  • The creator explains that they’ve been making thumbnails for 4 years, showing improvement from their first video to their 100th video and current thumbnails.
  • Since 2023, the biggest change has been AI—though they stress that psychology remains a major factor in thumbnail effectiveness.

• Psychology-based thumbnail design process

  • The creator claims thumbnail performance improves when design aligns with how the viewer’s mind works and how attention patterns are formed.
  • They reference that 99% of YouTube creators follow two main thumbnail layout rules, also commonly used in filmmaking/cinematography:
    1. Rule of Thirds
    2. Symmetry Rule (they note it is “kept in layout”; only 9% use it, per the creator)

• Rule of Thirds (attention capture)

  • Place the main subject around where the grid lines intersect.
  • The eye naturally notices those intersection points first.
  • Example concept: a faceless element can trigger curiosity (“no eye/mouth/ears” → unusual → stops scrolling), and then the viewer connects the image with the title to understand the story.

• Left vs right meaning (viewer vs creator perspective)

  • The creator describes a directional bias:
    • People typically scan left to right (reading/writing habit).
    • So the left side is treated as the viewer’s perspective (emotion/expressions/story elements).
    • The right side is treated as the creator’s perspective (what the creator intends to show to motivate the click).
  • They say you can flip left/right, but this is what they believe works best psychologically.

• Text and wording strategy

  • Amount of text: keep thumbnails to 2–3 words maximum (it may be more if the audience already understands the context).
  • Even if you write no text, the thumbnail can still work if the image carries the story.
  • Choice of words matters more than generic messaging.
  • Example emphasis:
    • Ensure the thumbnail text matches/justifies the video’s premise.
    • “Legal clickbait” example: they claim “10 lakhs earned” is not their claim, but is justified by the video showing how someone else achieved it.

• Color focus / contrast

  • Thumbnail “poppiness” depends heavily on color selection.
  • They recommend high-saturation/contrasting bright colors because:
    • Black/gray can remain dull even with contrast.
    • Vibrant colors can be enhanced in post-production to stand out more.
  • They note that challenge/reaction thumbnails often use 4–5 main colors for this reason.

---

Methodology / step-by-step workflow (detailed)

Step 1: Thumbnail planning with imagination + psychology

• Start with imagination, but only after understanding thumbnail psychology and the viewer’s perspective.
• Apply layout rules:
  • Rule of Thirds: place the main focus near intersection points.
  • Symmetry Rule (if chosen): create a mirrored/equal balanced layout.
• Consider left vs right:
  • Put story/emotion elements on the left (viewer perspective).
  • Put creator-intent/motivation elements on the right (creator perspective).
• Plan thumbnail text:
  • Limit to 2–3 words (generally).
  • Ensure wording strongly fits the video title/promise.
  • The image should still “tell the story” even with little/no text.
• Use color strategy:
  • Prefer bright, saturated colors so they “pop” after editing.

Step 2: Generate an AI thumbnail image (example concept + prompt structure)

• Example video imagination given:
  • “I am shaving my beard after 80 months to see family reaction”
• Concept compositions suggested:
  • Symmetrical option: split visual (beard vs no-beard), blur for curiosity, place family in the middle to show reaction.
  • Alternative option: show a man shaving with the family’s reaction behind.
• Prompt guidance (as described):
  • The prompt’s first line will always be something like:
    • “Make a cinematic, realistic portrait image of a…”
  • Then you specify the rest based on your idea.
• Set aspect ratio for thumbnails:
  • Use 4: aspect ratio = 16:9 (as stated)

Step 3: Put the creator’s face onto the AI image (Discord bot workflow)

• Install a bot/tool mentioned as “insidFace bot” / inside face bot (exact name as heard).
• Follow the described steps:
  • Save ID
    • Use a clear, good-quality photo of the creator.
    • Save the creator’s name/id.
  • Swap ID
    • Upload the AI-generated image.
    • Select/attach the saved face id so it matches the creator’s face.
    • Confirm by pressing enter.

Step 4: Enhance image quality in Canva

• Upload the AI/face-matched image to Canva.
• If quality degrades, improve it:
  • Use Effects/App panel
  • Install/download an image upscaler (via apps panel, if needed)
  • Choose an upscale factor: 2x / 3x / 4x / 8x (they select a level)
• Optional finishing edits:
  • Remove background
  • Add frames
• Download the thumbnail with highest settings.

Step 5: Color grading with Lightroom (specific suggested settings)

• Use Adobe Lightroom (mentioned as working on iOS and Android).
• Workflow:
  • Select the image to color grade.
  • Adjust:
    • Light
      • Exposure: ~0.44
      • Contrast: +15
      • Highlights: +2
      • Shadows: -8
    • Color
      • Vibrance: +33 (or suggested range 28–35)
      • Saturation: around 5–11 (they mention keeping at 10 in the example)
    • Effects
      • Vignette color: -48 (depending on desired corner darkness)
      • Clarity: +32 (range suggested 28–38)
      • Texture: ~8 (range suggested 5–15)
    • Detail
      • Color noise: 8 (range suggested 5–15)
  • Convert settings into a preset for reuse.
  • Share/save the preset (they mention: share → save copy to device)
• End goal: produce the finished thumbnail.

---

Sources / speakers featured (as stated in subtitles)

• The video narrator/creator (main speaker; name not provided in the subtitles)
• AI tools/platforms mentioned
  • Discord (used to run/install a face-swap bot)
  • Inside face bot / “insidFace bot” (face swap bot name as heard)
  • Canva (for upscaling and editing)
  • Lightroom (Adobe Lightroom) (for color grading and preset creation)
• Conceptual source referenced
  • Filmmaking / cinematography (where the rules also appear)
  • “Rule of Thirds” and “Symmetry Rule” (concepts; no single author cited)