Q2 2026: BLOCK 3

Q2 2026 - QUARTERLY WORKSHOP

BLOCK 3: Q2 Creative Scaling System

1. Q2: High CTR or die
2. Q2: What’s the avatar looking for?
3. Q2: Winning creatives prediction
4. Creative Diagnosis Checklist
5. How to develop a winning ad (iterations and hypoth.)

Q2: High CTR or die

• ROAS is derived from CPC and RPV
• CPC going up will usually decrease our ROAS
• So if we assume CPM is about to go UP
• CTR must go UP as well

General drivers of CTR:

1. Adapt the message to Q2 / Summer desires
2. Image -> High color contrast
3. Image -> Clear headline with relevance
   “I understand what they are selling”
   “I understand this is for me”
4. Video -> Thumbnail and Hook
5. Engagement in the ad
6. Tension / Reason to click
   No tension = Ah nice product = Scroll past
   “This week only”, “Last Days”, “Until Sunday”, …
   “Last Units”, “Sold out 3x”, “Last chance”, …

Q2: What’s the avatar looking for?

Context

1. Summer is coming
2. People will spend more time outside / Social activities
3. Their body will be more exposed, less clothes
4. Oily skin, sun exposure
5. Summer holidays are being booked in advance
6. More light at home, less darkness

So the big DESIRES are:

1. Confidence
2. To look better with LESS effort
3. To not feel behind others (comparison)
4. To be ready for key moments
5. To enjoy life more (winter sadness, sun is coming)

Asymmetric value function (Loss Aversion)

• We value gains and losses differently,
  with losses being felt more intensely than equivalent gains
• 
  
• Copywriting tip: Position your product as the solution to avoid a repeated pain in Summer
  
  Skincare: “Another summer dealing with acne?”
  Fat loss product: “Another summer avoiding the beach because of this?”
  Jewelry: “SS26 | From basic to standout with one piece”

Negative angles / Fix a problem

Q2: Winning creatives prediction

Early Summer adaptations

• Backgrounds
• Text: “SS26” or “Summer”
• More outdoors than indoors
• How the UGC model is dressed

Before/After by Season or Result By Summer

• POV: You have a Summer Glow Up
• The Winter Me vs the Summer Me
• Start today and get (desire) before Summer
  
• Transformation ads:
  January: Start                       By February

Suggestion: Creative research

→ Looking for Summer ads, how are your competitors adapting

EVERGREEN AD CREATIVES

What’s working best for me right now:

• Cosmetics:
  VSL 1-2min ads that focus on problem/solution
  (Show the pain in black and white, then show the desire, then show the mistakes people make with X emoji, then show the desire again in 3 different avatars, …)
  
  And whitelisting creators, explaining their experience with the product after 1 month, in the bathroom, 1-2min
  
  Easy way:
  → Send partnership request on their IG
  → Some accept
  → Don’t run both identities, “only first identity”
  
• Sleep client:
  Multiproduct images with a strong headline
  
• Fashion:
  40% of the sales coming from DPA carousel
  Then split screens: Model UGC + Product
  Then multiproduct images with small text
  

WHITELISTING AND PARTNERSHIP ADS

1. LOW TRUST: The brand is selling me something
2. MID TRUST: Manel and Brand
3. HIGH TRUST: Manel

UGC 5 videos: 500€

Add 100€ → Run partnership ads

• Followers

Creative Diagnosis Checklist

I wanted to add this extra because none of you is starting from zero, all of us have 10-100 ad creatives in the ad account.

With time, we accumulate old creatives, sometimes launch new stuff, etc… and we get distracted.

It’s inevitable.

So I’ve made a small checklist to easily self-diagnose your ad account:

https://docs.google.com/spreadsheets/d/1GPxv_2ugsJXqqHCxeXCGo_cgvL4TSg8e9jsImU3TgW4/edit?usp=sharing 

How to develop a winning ad (iterations and hypoth.)

• Iterate forever = no breakthroughs
• Test random ideas = no compounding

1. To develop, of course start with finding a winning ad:

• More than 10 purchases at good ROAS
• Above account average CTR
  
  
  
  
  

2. Deconstruct WHY it works
   Desire: Bright Skin

Angle: Transformation with certainty (97%)

Concept: Before/after image
Extra: Statistics
Model: She has a certain problem, skin looks good

ITERATIONS

• Model Iterations
• Keep all the same
  Change model → Younger
• Keep all the same
  Change model → Older
• Keep all the same
  Change model → White
• Keep all the same
  Change model → Someone with a stronger problem
  
• Desire Iterations
• Keep all the same (Or adapt the UGC transformation)
  Change the desire: Get rid of acne scars in X days
• Keep all the same (Or adapt the UGC transformation)
  Change the desire: Goodbye dark spots, Hello Glow
• Keep all the same (Or adapt the UGC transformation)
  Change the desire: Fade hyperpigmentation or get your money back
  
• Keep the Angle: Transformation with certainty (97%)
  But change the concept, model or format
  

In general, for iterations we change the angle/desire or ad format, but if we change the concept (from split image to UGC video) then that’s a new hypothesis.

So again,

When we start “Concept Iterations” in reality we are trying new things, it’s actually a new hypothesis.

Example: Would before and after work in a video, where a person explains how the transformation happened?

I’m pulling the learning from the previous winning ad, but it’s not really an iteration:

Example: Listicle Video

The 3 videos on the right are from June 18th, and the other 2 on the left are from July.

The 45+ and 55+ are a very clear iteration, just changing the avatar.

LAYER 1: Just iteration → Super small change

LAYER 2: for example switch from before/after image to video (pull the learning from the winner and build on top of it)

LAYER 3: Complete new hypothesis → I see our competitors running these ugly ads, we’ve never tried, let’s shoot 3.