Gustoso
MethodologyBlogPricingSign Up
← Back to blog
Guides

How to Test an Ad or Product with Gustoso AI

Darron Jennings

Get purchase intent scores and real consumer feedback on your creative — in minutes, not weeks. Whether you're A/B testing ad variants or validating a product concept, Gustoso's Quick Evaluation gets you there fast.

Best for: A/B tests, concept checks, creative pre-testing, packaging comparisons

Quick Start

Here's how to run your first ad or product test in under 5 minutes.

1. Choose "Test an Ad or Product"

From your dashboard, click New Study. You'll see two options — select Test an Ad or Product and hit Start.

2. Name Your Study

Give your study a descriptive name (like "Summer Campaign A/B Test") and optionally pick a product category. The category helps calibrate result confidence for your industry.

3. Set Up Your Questions

You'll start with a default purchase intent question on a Likert scale. This is the core of every quick evaluation — it measures how likely someone is to buy. You can add more questions if you want, but one is enough to get started.

4. Define Your Audience

Pick a demographic template (like "US General Population") or build a custom audience by selecting age ranges, income levels, gender, and regions. Templates are the fastest way to get going.

5. Upload Your Creative

Drag and drop your ad image, product mockup, or text description into the upload zone. For A/B testing, upload multiple variants and label them (A, B, C).

6. Set Sample Size and Review Cost

Choose how many AI personas will evaluate your creative (100-500). The default of 300 gives you solid demographic coverage. You'll see the credit cost calculated in real time.

7. Review and Launch

Check your configuration one last time, accept the terms, and hit Launch Study. That's it!

8. Get Your Results

Results arrive in about 1-5 minutes. You'll see purchase intent scores, Likert distributions, demographic breakdowns, and AI-generated qualitative feedback.

Full Detail Guide

Ready to get the most out of your quick evaluation? Here's everything each step offers.

Understanding Question Templates

The default question — "How likely are you to purchase this product?" — uses a 5-point Likert scale from "Very Unlikely" to "Very Likely." This is the industry standard for measuring purchase intent. You can also add questions from these templates:

  • Appeal — "How appealing is this product to you?" Measures emotional reaction.
  • Clarity — "This concept is clear and easy to understand." Tests messaging comprehension.
  • Uniqueness — "This product is different from others I've seen." Gauges differentiation.
  • Open Feedback — A free-text question that captures qualitative insights you might not have anticipated.

Choosing the Right Audience

Templates are pre-configured demographic profiles designed for common use cases:

  • US General Population — Ages 18-65, all income levels, balanced gender, all regions. Your go-to for broad market testing.
  • Urban Millennials — Ages 25-34, mixed income, urban areas. Great for trendy consumer products.
  • Budget-Conscious Families — Ages 35-54, lower-middle income. Ideal for value-oriented products.
  • Affluent Professionals — Ages 25-54, high income, urban. Perfect for premium products.
  • Gen Z Early Adopters — Ages 18-24, tech-forward. Best for new tech and innovative products.

Pro tip for A/B testing: Use the same demographic settings across all your variants. This ensures differences in scores come from your creative, not your audience.

Asset Tips

  • Supported formats: PNG, JPG, PDF, and plain text files (max 10MB each, up to 10 per study)
  • For A/B tests, upload each variant and label them A, B, C using the variant labels. Results will be broken down per variant.
  • Text descriptions work too — don't have a visual yet? Upload a text file with your product description or ad copy.
  • Image quality matters — use production-quality images when possible. Low-res placeholders may affect how personas perceive your product.

Understanding Credit Costs

Credits are calculated as: Base cost (personas x 1 credit each) + Evaluation cost (personas x questions x assets). For example, 300 personas with 1 question and 2 assets = 300 + (300 x 1 x 2) = 900 credits.

Reading Your Results

Your results page shows:

  • Overall Score — The average Likert score across all personas (1.0-5.0). A score of 3.5+ generally indicates positive purchase intent.
  • Response Distribution — A bar chart showing how many personas chose each Likert option.
  • Demographic Breakdown — See how different age groups, income levels, genders, and regions responded.
  • Statistical Confidence — Confidence intervals and sample size impact analysis.

What's a Good Score?

  • 4.0+ — Strong purchase intent. Your creative is resonating.
  • 3.5-4.0 — Moderate intent. Consider tweaking your message or targeting.
  • Below 3.0 — Weak intent. Major revision recommended.

Ready to test your next campaign? Create your first study