Browse AI Tutorial

Step 1: Sign Up and Set Up Your Account

Head to browse.ai and click 'Start for Free'. I recommend signing up with Google; it's the fastest and integrates seamlessly later. You'll land on the dashboard. What surprised me was how quickly they throw you into the action—you can create a robot immediately. Don't be overwhelmed. First, go to your account settings (top-right corner). Here, connect your Google account under 'Integrations'. This is crucial for exporting data later. Then, check your 'Credits' in the left sidebar. The free plan gives you 50 credits monthly. One credit equals one page load by a robot. For a simple price monitoring task on a single product page, that's plenty to start. In my experience, doing this setup first prevents annoying interruptions later when you're in the flow.

Step 2: Navigate the Dashboard

The dashboard is clean but has key areas. On the left, 'Robots' is your library of all created scrapers. 'Monitoring' shows the results from robots you've set to run on a schedule—this is where your data lives. 'Credits' is your usage meter. The big '+ Create Robot' button is your starting gun. I tested the pre-built templates heavily, and while they're a good conceptual start, I found creating from scratch gives you more control and understanding. The main workspace, where you'll train your robot, is a simplified browser window. The magic happens in the right-side panel where you'll see 'Select Data' and 'Configure Robot' tabs. Spend two minutes just clicking around these areas to get oriented. It's less complex than it looks.

Step 3: Create Your First Monitoring Robot

Click '+ Create Robot'. A new window opens. Paste the URL of your target page. For this tutorial, use a simple product page like a book on Amazon. Click 'Start Training'. The page loads inside Browse AI. Now, the core action: you need to teach the robot what to extract. Hover over the product title—you'll see a green highlight. Click it. A box appears asking what this data is. Label it 'Product Title'. Click the price, label it 'Price'. What surprised me was how well it handles lists. If you were on a search results page, clicking one item often auto-detects all similar items. After selecting 2-3 data points, click 'Done Selecting' in the right panel. You've just built the brain of your robot.

Step 4: Customize and Refine Your Results

Now, configure your robot. In the right panel, go to the 'Configure Robot' tab. Give it a clear name, like 'Monitor [Product Name] Price'. Under 'When to run', keep it on 'Only when I trigger it' for now. The advanced settings here are powerful. 'Paginate' lets you scrape multiple pages. 'Scroll' is essential for modern JavaScript-heavy pages—enable it if your data loads as you scroll. 'Wait for element' is my secret weapon for dynamic sites; tell the robot to wait until a specific selector (like the price element) loads before capturing. Run a test by clicking 'Save & Run'. Check the output in the 'Monitoring' section. If data is missing, go back to training and click more elements. Iteration is key.

Step 5: Save, Export, and Share

Your robot is live. In the 'Monitoring' section, click on your robot's name. You'll see the latest captured data in a clean table. To automate, click 'Schedule' and set a frequency (e.g., daily). This is where Browse AI shines. To export, click 'Export Data'. I always choose 'Google Sheets'. It will create a new sheet with your data and a live link that updates on each scheduled run. You can also get an API endpoint or webhook URL here for developers. To share, use the 'Share' button on the robot's page. You can give view or edit access to team members. In my experience, the Google Sheets integration is rock-solid and the best way for beginners to use the data.

Step 6: Explore Advanced Features

Once comfortable, dive deeper. The 'Recipes' library has pre-configured robots for sites like LinkedIn, Twitter, and Crunchbase—great for inspiration. Explore 'Input Parameters': this lets you create one robot that can scrape different URLs by passing a variable (e.g., different product IDs). The API is robust; you can trigger robots from other apps like Zapier. I tested the 'Change Detection' feature extensively; it can email you only when your extracted data *changes*, which is perfect for price drop alerts. While the UI is beginner-friendly, these features offer enterprise-level automation. My stance is that mastering input parameters is the single biggest leap in efficiency you can make.