If you’ve ever wondered if a different price on your product page would boost sales—or tank them—this guide is for you. You’ll learn how to use Google Optimize to run honest, real-world pricing experiments on your site. No theory, just a step-by-step walkthrough, including what works, what to skip, and why you shouldn’t trust your gut alone.
Why Test Pricing? Isn’t That Risky?
Before you fiddle with numbers, it’s fair to ask: is it really worth testing prices? The answer: yes, but only if you do it right. Humans are bad at guessing what price will work best. Customers often react to tweaks in ways you can't predict. Testing means you can stop guessing and start making decisions based on actual data.
If you’re worried about scaring off customers or “cheapening” your brand, don’t be. You can run controlled experiments with only a portion of your traffic. The goal isn’t to race to the bottom on price—it’s to find out what your customers will actually pay, and how that changes conversion.
Step 1: Decide What, Exactly, to Test
Testing “pricing” sounds simple, but there are a few ways to approach it. Do you want to see if a higher price hurts conversion? Or if a lower price brings in more buyers but lowers your margin? Or maybe you want to test bundles or discounts.
Pick 1–2 variables for your first test. Here are some common options:
- Straight price change: Try $39 vs $49.
- Discount messaging: Show “$49 $39” vs just “$39.”
- Bundling: Main product vs. product + add-on for a single price.
- Price anchoring: Show original price crossed out next to sale price.
Don’t try to test everything at once. Keep it simple, or you’ll never know what actually made the difference.
Pro tip: If you have a high-traffic site, you can get fancier. If not, keep experiments basic or you’ll wait weeks for results.
Step 2: Set Up Google Optimize
Google Optimize (free, with a paid version that most folks don’t need) lets you run A/B tests and see which page variant performs better.
What You Need
- A Google Analytics account
- Access to your site’s code (or a developer who can help)
- Google Optimize account
How to Get Started
-
Sign up and link to Google Analytics. When you set up your Optimize container, you’ll be prompted to link it to your Analytics property. This is how you’ll track conversions.
-
Install the Optimize snippet. You’ll need to add a small script to your site. This is what lets Optimize swap out your test versions. For most setups, it’s as simple as pasting a line of code into your
<head>tag. -
Check that everything’s working. Use the “Preview” mode in Optimize to make sure your site loads and the snippet is firing. Broken tags mean broken tests.
Pro tip: If you’re using a platform like Shopify or WooCommerce, look for plugins or apps that make this easier. But always double-check—these integrations are famously finicky.
Step 3: Build Your Pricing Experiment
Now for the fun part—setting up your test.
Choose an Experiment Type
For most pricing tests, you’ll use an A/B test (aka “A/B/n” if you want more than two variants). Don’t mess with “multivariate” or “redirect” tests unless you’re sure you need them.
Create Your Variants
- Original: Your page as it is now.
- Variant(s): The new price, bundle, or message you want to try.
You can use Optimize’s visual editor to swap out prices right on the page—no coding required for simple changes. If you want to test something more complex (like bundles), you might need a developer.
Watch out for:
- Currency formatting. Don’t forget to update all instances of the price on the page—not just the headline.
- Price in the cart/checkout. If your test changes the product page but not the cart, you’ll confuse people and mess up your data.
- Inventory and fulfillment. If you’re selling at a discount, make sure you have the margin and stock to handle a sudden spike.
Step 4: Define Your Success Metric
Don’t just test for clicks or “engagement.” Pricing tests should be about real business outcomes. Pick a metric that matters:
- Conversion rate: % of visitors who buy
- Revenue per visitor: Total revenue ÷ number of visitors (this accounts for both conversion and average order value)
- Average order value: Can be useful if you’re testing bundles or upsells
Set this up in Google Optimize by connecting to your Analytics goals or events. If you’re not sure, conversion rate is the safest bet for most product pages.
Pro tip: Revenue per visitor is the gold standard if you care about profit, not just sales.
Step 5: Split Your Traffic
Choose what % of your visitors see each version. For a first test, a simple 50/50 split is fine. If you’re nervous, you can send only 10–20% of traffic to the test version—but it’ll take longer to get results.
Don’t:
- Run a test for just a day or two. You need enough data to see real patterns (usually at least a week, often longer).
- Peek at results and end the test early. It’s tempting, but you’ll fool yourself with bad data.
Do:
- Let the test run until you have at least 100 conversions per variant, or longer if you can. Small numbers = misleading results.
- Check for weird technical issues (like coupons not working or prices not updating in the cart).
Step 6: Launch, Monitor, and Wait
Hit “Start Experiment.” Now, go do something else. Don’t obsess over daily numbers.
What to Watch For
- Technical issues: Check your analytics for sudden drops or spikes—could mean the test is broken.
- Customer confusion: Watch for an increase in support requests or abandoned carts. If people are emailing “Why did the price change?” you might need to rethink your approach.
- Statistical significance: Google Optimize will show you when a result is “significant”—but don’t trust this blindly. If the change is tiny or the sample size small, keep running.
Step 7: Analyze and Act On the Results
When enough data is in, look at your primary metric. Did the new price help or hurt? Here’s how to read what you see:
- Higher price, same conversion: Congrats, you’re leaving money on the table at the old price.
- Lower price, higher conversion but lower revenue per visitor: You’ve boosted sales but maybe cut profit. Decide what matters most.
- No clear winner: That’s fine. Sometimes small price changes don’t matter. Move on to another test.
Don’t chase “statistical significance” for its own sake. If the difference is only 1% and your business isn’t huge, it probably doesn’t matter.
What NOT to Do
- Don’t announce price changes to your whole email list mid-test. You’ll contaminate your experiment.
- Don’t make permanent changes based on a fluke week (holidays, big promo, etc).
- Don’t ignore customer feedback if a change blows up your support inbox.
What Works, What Doesn’t, and What to Ignore
Works: - Simple A/B price or discount tests - Testing bundles or add-ons for higher order value - Using real traffic, not “friends and family”
Doesn’t Work: - Running tests on low-traffic pages (you’ll wait forever for results) - Changing too many things at once - Stopping tests early because you’re impatient
Ignore: - “Best practices” blogs that say “$39 converts better than $49.” Test on your site, with your customers. - Overly complex multivariate tests unless you have thousands of daily visitors.
Keep It Simple. Iterate.
That’s it. No magic, just honest experimentation. The best pricing strategy is the one that works for your actual customers, not what some guru says.
Start small, keep your experiments clean, and don’t be afraid to call a test a dud. The more you test, the more you learn—and the more likely you are to stumble on a price that actually works.
Now, go run a test. Your customers will tell you what works. All you have to do is listen.