Title: Quick Start
Locale: en
URL: https://sensorswave.com/en/docs/experiments/quick-start/
Description: Complete your first A/B experiment in 20 minutes
With this tutorial, you will complete your first A/B experiment in 20 minutes, experiencing the full workflow from creating an experiment, integrating code, to analyzing data.

## Prerequisites

Before you begin, make sure you have:

- Completed SDK integration (this guide uses the [JavaScript SDK](../data-integration/client-sdks/javascript.mdx) as an example)
- Enabled the A/B testing feature (e.g., set `enableAB: true` in the JavaScript SDK)
- Understood the basic concepts of [User Identification](../data-integration/user-identification.mdx)

## Example Scenario

**Business context**: You operate an e-commerce website and want to improve the click-through rate of the "Add to Cart" button. The design team has proposed two options: keep the current blue button, or switch to a more prominent red button.

**Experiment Hypothesis**: The red button can increase click-through rate by 10% compared to the blue button.

**Experiment design**:
- **Control**: Blue button (current design)
- **Test Group** (Treatment): Red button (new design)
- **Allocation**: 50% vs 50%
- **Duration**: 2 weeks (to ensure sufficient sample size)

**Target metrics**:
- **Primary Metric**: Button click-through rate
- **Secondary metrics**: Add-to-cart conversion rate, product detail page dwell time

## Step 1: Create the Experiment

Create an experiment in the Sensors Wave console:

### 1.1 Navigate to the Experiment Management Page

1. Log in to the Sensors Wave console
2. Click **A/B Experiments** in the left menu
3. Click the **New Experiment** button in the top-right corner

### 1.2 Fill in Basic Information

**Experiment Key**:
```
cart_button_color_test
```

**Display name**:
```
Cart Button Color Experiment
```

**Experiment description**:
```
Test whether a red button attracts more user clicks than a blue button, improving add-to-cart conversion rate
```

**Hypothesis**:
```
Red button can increase click-through rate by 10%
```

### 1.3 Configure Variant Groups

**Control**:
- Variant name: `control`
- Allocation: `50%`

**Test Group** (Treatment):
- Variant name: `treatment`
- Allocation: `50%`

### 1.4 Configure Dynamic Variables

Add a variable to control the button color:

- **Variable name**: `button_color`
- **Variable type**: String
- **Control value**: `blue`
- **Test Group value**: `red`

### 1.5 Save and Launch

1. Click the **Save** button
2. After verifying the configuration, click the **Launch** button
3. The experiment status changes to **Running**

## Step 2: Code Integration

Integrate the experiment code on the product detail page:

```javascript

function ProductDetail() {
  const [buttonColor, setButtonColor] = useState('blue'); // Default blue

  useEffect(() => {
    // Get experiment variables
    async function loadExperiment() {
      try {
        const experiment = await sensorswave.getExperiment('cart_button_color_test');
        const color = experiment.get('button_color', 'blue');
        setButtonColor(color);
      } catch (error) {
        console.error('Failed to get experiment:', error);
        // Use default value on failure
        setButtonColor('blue');
      }
    }

    loadExperiment();
  }, []);

  // Handle button click
  const handleAddToCart = () => {
    // Track click event with experiment info
    sensorswave.trackEvent('AddToCartClicked', {
      experiment: 'cart_button_color_test',
      button_color: buttonColor,
      product_id: 'PROD-001',
      product_name: 'iPhone 15 Pro'
    });

    // Execute add-to-cart business logic
    addToCartLogic();
  };

  return (
    
      iPhone 15 Pro
      价格：¥7,999

      {/* Dynamically set button color based on experiment variable */}
      
        加入购物车
      
    
  );
}

```

## Step 3: Verify and Monitor

After the experiment is released, verify that the integration is correct and continuously monitor data.

### 3.1 Verify Exposure Logs

Check exposure logs in the Sensors Wave console:

1. Navigate to **Insights** > **Segmentation**
2. Select Event: `$ABImpress`
3. Add filter: `Experiment name = cart_button_color_test`
4. View data for the last 1 hour

**Verification checklist**:
- Are exposure Events being reported normally
- Does each user have only one exposure log
- Are the Control and Test Group exposure counts roughly equal (50% vs 50%)

### 3.2 Check Group Distribution

Verify that the Allocation is even:

1. In Segmentation, select Event: `$ABImpress`
2. Group by **Variant name**
3. View user count for Control and Test Group

**Expected results**:
- Control: ~50%
- Test Group (treatment): ~50%

If the group distribution has a large deviation (e.g., 60% vs 40%), there may be a configuration error or insufficient sample size.

### 3.3 Monitor Key Metrics

Regularly check the trend of core metrics:

**Primary Metric**: Button click-through rate

1. In Segmentation, select Event: `AddToCartClicked`
2. Group by **Variant name**
3. View event count and user count for each Variant

**Calculate click-through rate**:
```
Click-through rate = Clicking users / Exposed users
```

**Secondary Metric**: Add-to-cart conversion rate

1. Create a Funnel
2. Define funnel steps:
   - Step 1: Product detail page view (`PageView`)
   - Step 2: Click add to cart (`AddToCartClicked`)
   - Step 3: Successfully added to cart (`AddToCart`)
3. Group by **Variant name** to compare conversion rates

## Step 4: Analyze Results

After the experiment has run for 2 weeks (or reached sufficient sample size), analyze the results.

### 4.1 Wait for Sufficient Sample Size

**Minimum sample size requirements**:
- At least 1,000 exposed users per group
- At least 100 converted users per group

**Risks of insufficient sample size**:
- Unstable experiment results, easily influenced by random factors
- Statistical significance testing becomes invalid
- May lead to incorrect conclusions

### 4.2 Compare Key Metrics

Use Segmentation to compare data between the two groups:

**Example data**:

| Variant | Exposed Users | Clicking Users | CTR | Lift |
|------|----------|----------|--------|---------|
| Control (blue) | 5,000 | 1,200 | 24.0% | - |
| Test Group (red) | 5,000 | 1,400 | 28.0% | +16.7% |

**Analysis**:
- Test Group CTR: 28.0%, Control: 24.0%
- Relative lift: (28.0% - 24.0%) / 24.0% = 16.7%
- Hypothesis validation: Actual lift of 16.7% exceeds the expected 10%

### 4.3 Check Statistical Significance

Use Sensors Wave's statistical testing feature (or third-party tools) to calculate the P-Value:

**Criteria**:
- **p 
      iPhone 15 Pro
      价格：¥7,999

      {/* Use red button directly */}
      
        加入购物车
      
    
  );
}
```

## FAQ

### Q: How long should an experiment run?

At least 2 weeks to ensure a complete cycle (weekdays and weekends). If traffic is low, more time may be needed to reach sufficient sample size.

### Q: What if the sample size is insufficient?

There are several approaches:
- Extend the experiment Duration
- Increase experiment traffic (e.g., from 50% to 80%)
- Lower the Minimum Detectable Effect (MDE)

### Q: What if the experiment result is not Significant?

Possible reasons:
- Insufficient sample size: extend the experiment or increase traffic
- Incorrect Hypothesis: the new solution has no Significant difference from the old one
- External factor interference: holidays, marketing campaigns, etc.

Recommendation: Extend the experiment Duration. If results remain insignificant, the new solution has limited impact and you may want to abandon this optimization direction.

### Q: Can the configuration be modified during the experiment?

Not recommended. Modifying configuration (e.g., variable values, Allocation) affects the reliability of experiment results. If changes are necessary:
- Stop the current experiment
- Create a new experiment
- Re-collect data

## Next Steps

Congratulations on completing your first A/B experiment! Next, you can:

1. **[Core Concepts](core-concepts.mdx)**: Gain a deeper understanding of how A/B experiments work
2. **[Experiment Design](experiment-design.mdx)**: Learn how to design scientifically rigorous experiments
3. **[SDK Integration](sdk-integration.mdx)**: Explore more SDK usage methods
4. **[Metrics and Analysis](metrics-and-analysis.mdx)**: Master data analysis techniques

---

**Last updated**: January 29, 2026