Skip to main content

Split Testing Video Sales Letters (VSLs)

Learn the best practices and alternatives to split test your VSLs

Aldo Peña avatar
Written by Aldo Peña
Updated over 3 weeks ago

What is Split Testing for VSLs?

Split testing (also known as A/B testing) is the process of comparing two or more versions of your Video Sales Letter to determine which performs better. This is done by showing different versions of the VSL to separate segments of your audience and measuring key performance indicators (KPIs) like conversion rate, watch time, and click-through rate.

Why Split Test Your VSL?

Your VSL is often the cornerstone of your sales funnel. Small tweaks—like a different headline, a new call to action, or a change in tone—can dramatically increase your sales. Split testing allows you to:

  • Identify what truly resonates with your audience.

  • Optimize for higher conversions and lower cost per acquisition (CPA).

  • Make data-driven decisions instead of relying on guesswork.

What Elements Should You Split Test?

You don’t have to change everything at once. Focused testing yields better results. Consider these components:

Element

What to Test

Headline (hook)

Different openers, emotional triggers

Length

Short (2–3 min) vs. long-form (10+ min)

Call to Action (CTA)

Placement, wording, or timing

Voiceover vs. No VO

With narration vs. text-only

Visual Style

Animation, talking head, slides

Offer Presentation

Bonuses, urgency, pricing strategy

Testimonial Sections

More or fewer social proofs

Best Tools for Split Testing VSLs

1. VWO

  • Type: Enterprise-level A/B testing platform

  • Best For: Marketers who want advanced targeting and deep insights

  • Features: Multivariate testing, audience segmentation, heatmaps

  • Pros: Professional-grade, scalable testing

  • Cons: Costly for small businesses

  • Type: Enterprise A/B testing and experimentation suite

  • Best For: Product and growth teams needing advanced experimentation

  • Features: Behavioral targeting, real-time data, personalization

  • Pros: Very powerful; integrates with CDPs

  • Cons: Requires technical setup for complex experiments

  • Type: Funnel builder with built-in A/B testing

  • Best For: Marketers and entrepreneurs without dev teams

  • Features: Split test pages, headlines, VSLs, checkout flows

  • Pros: User-friendly, quick deployment

  • Cons: Less advanced analytics

  • Type: All-in-one marketing platform

  • Best For: Building full funnels with automation

  • Features: Video hosting, A/B testing, email marketing

  • Pros: Great for solopreneurs

  • Cons: Slight learning curve for full features

  • Type: Budget-friendly funnel builder

  • Best For: Beginners and budget-conscious marketers

  • Features: Page A/B testing, email marketing, automation

  • Pros: Free plan available, intuitive interface

  • Cons: Basic reporting

Statistical Significance Calculator

Running an A/B test without verifying statistical significance can lead to false conclusions. These tools confirm whether a winning version is actually better or just the result of chance.

Recommended Tool:

The Bayesian A/B Testing Calculator by Dynamic Yield is a free online tool designed to help marketers and analysts determine the statistical significance of their A/B test results using Bayesian inference.

Key Features

  • Multiple Variations: Supports up to 10 different test variations.

  • Input Parameters: For each variation, you can input the total sample size (e.g., users, sessions, or impressions) and the number of conversions (e.g., clicks or goal completions).

  • Probability to Be Best: Calculates the likelihood that each variation will outperform the others in the long term.

  • Expected Loss: Estimates the potential loss in conversion rate if a suboptimal variation is chosen as the winner.

  • Posterior Simulation: Provides a visual representation of the distribution of conversion rates based on the collected data.

How It Works

The calculator employs Bayesian statistics, which allows for continuous updating of the probability estimates as more data is collected. This approach provides a more intuitive understanding of the results compared to traditional frequentist methods, as it directly answers questions like, "What is the probability that Variation A is better than Variation B?" It also allows for more flexible decision-making without the need for predefined sample sizes.

💡 Tip: Don’t stop your test until you’ve reached at least 95% confidence with enough traffic (sample size).

Effective Split Testing

  1. Test One Variable at a Time
    Focused changes help you learn what matters.

  2. Let It Run Long Enough
    Collect enough traffic for valid comparisons—don’t stop early.

  3. Track What Matters
    Go beyond views—measure opt-ins, add-to-carts, and purchases.

  4. Document Everything
    Keep a log of what you tested, how long it ran, and what you learned.


For additional questions, feedback or assistance please feel free to reach out directly to support at [email protected]. 😊

Did this answer your question?