Skip to main content

Step by Step Guide to Setting Up A B Testing in Optimizely

Step by Step Guide to Setting Up A/B Testing in Optimizely​

We were sitting in our moss-green bean bags—those delightfully absurd cushions that threaten to swallow you whole—sipping on lukewarm matcha lattes that were a little too bitter, when it hit us. That epiphany, as sudden and bold as a splash of cold water, about the true magic of A/B testing. It wasn't merely about statistics and metrics; it was an art. A beautifully chaotic dance between two versions of a dream. One summer evening, my friend Marcus had a dazzling idea for our fledgling startup's website, a veritable spark of digital grandeur. But was his idea "the one"? This quest to answer that question led us into the labyrinth of A/B testing with Optimizely. Here's our winding journey through the steps, wrapped in our tales of triumph—and occasional despair—as we embarked on the quest for the optimal.

Step 1: Enter the Optimizely Rabbit Hole​

Entering Optimizely was like unveiling secret blueprints to a universe hidden within our screens. I remember that moment clearly—it was a lot like entering Willy Wonka's factory, minus the chocolate rivers. We were greeted by a clean interface, promising the power to test ideas without imploding our website. Let's get started:

  1. Sign Up or Log In: Navigate over to Optimizely and sign up for an account if it's your first foray into this digital playground. For the seasoned mavericks among us, log in.

  2. Create a New Project: Once in, click on the "Create New" button. This opens a gateway to a new project. Name it something whimsical (or practical—your call!) and choose your primary goal metric that echoes with your project's ambitions.

  3. Install Optimizely Snippet: Ah, the snippet. That tiny piece of code, much like an incantation, will bind Optimizely to your site. Place it in the header of your site. For my fellow code wizards, insert:

    <script src="https://cdn.optimizely.com/js/YOUR_PROJECT_ID.js"></script>

    Savor this moment, because it’s the start of the magic.

Step 2: The First Experiment - A Canvas of Possibilities​

The sun was setting outside our window, the sky painted in hues of indigo and pink, when we embarked on setting up our very first test. Our excitement was as palpable as the smell of fresh coffee from the downstairs café. Here’s what we did:

  1. Create the Experiment: Optimizely allows you to craft experiments, much like a digital Mad Hatter's tea party. Click on "Experiments" and choose "New Experiment."

  2. Define Variations: This is where the creative juices flowed—set your baseline, which will remain untouched like that last slice of pizza no one claims, and then concoct your variations. Variation 1, Marcus decided, would have a bold color palette that screamed innovation.

  3. Target Your Audience: Optimizely lets you aim your arrows—your variations—at specific demographic targets. Choose wisely, young archer.

Step 3: Define Goals - What Victories Shall We Measure?​

Marcus leaned back in his bean bag, a quizzical expression on his face. "But how do we know if we're winning?" he mused. A pause, then enlightenment: metrics! We need to measure, not by teaspoon, but by grand bucket.

  1. Set Performance Goals: Each test must have a yardstick. The success of our website variations danced on metrics—conversion rate, time on page, clicks, you name it. Define these goals in Optimizely.

  2. Add Secondary Metrics: Like a sneaky detective keeping tabs on background characters, set secondary metrics to capture the subtleties of user behavior. This helps reveal the hundred shades of visitor interaction.

Step 4: Launch the Test - Unleash the Kraken​

The moment of truth. Clicking 'Launch,' we felt akin to mad scientists turning on their machines for the first time, unsure if we’d created a wondrous marvel or a terrifying beast. The thrill was intoxicating.

  1. Review and Launch: Do a once-over on your settings, ensuring no detail has been overlooked (like the time Marcus forgot his grandma’s birthday, oops!). If all looks well, press that launch button with the grandiosity of a fireworks director on the fourth of July.

  2. Monitor Real-Time Results: It’s alive! Monitor the test’s progress in real time. The Optimizely dashboard becomes a living palette, painting an ongoing story of user interaction.

Step 5: Analyze Results - Deciphering the Da Vinci Code​

Soon enough, data began to flow in, whispering secrets of user preferences. It was like an archaeological dig through the layers of user behavior and interaction.

  1. Gather Data: Watch Optimizely as it compiles user interactions. The results are charted like delicate strokes of a Monet masterpiece.

  2. Compare and Decide: Here comes the climactic decision! Compare the data from your baseline and variations. Optimizely even highlights the statistical significance, like a star student eagerly raising their hand in class.

  3. Iterate: With results in hand and a gleam in our eyes, Marcus and I decided to take the learnings forward into further experiments. After all, when it comes to A/B testing, there’s always another layer to delve into or a hypothesis to prove.

Conclusion - Onward and Upward​

Reflecting back, those nights in our tiny office, filled with hopeful dreams and lukewarm matcha, etched themselves into our story. Each test taught us lessons—not just about our users, but about the courage to question and explore. Optimizely became more than a tool; it was our canvas, and A/B testing, our brush. So, as you embark on your own journey, whether it's your first experiment or your fiftieth, remember this: each test is a step in the dance of discovery, a delightful waltz of insight and innovation.

May your tests be insightful and your outcomes ever surprising!