How To Use Split Testing for Pricing Optimization

Price testing allows companies to find the ideal price point by balancing customer satisfaction and company profits. If you lower your prices, you’ll almost certainly make more sales—but will the additional sales outmatch the revenue you were earning at the higher price point?

As a business owner, one way to experiment with different price points is to use a technique called split testing. It’s a way to determine price sensitivity among potential customers and find a competitive pricing structure to deliver maximum revenue. 

Here’s an explanation of split testing for pricing, with key tips for applying it to your own business.

What is split testing?

Split testing is a pricing strategy in which you offer a product or service at different prices and see how they impact conversion rates and overall profits. It’s a form of A/B testing—you compare two different versions and see which performs better. 

Here’s how the split test pricing model works in a real-world scenario: Let’s say you have an ecommerce store that specializes in skin care, and you currently sell a jar of moisturizer for $29.99. To see if you can increase overall revenue without significantly impacting your sales volume, you A/B test your pricing. This means setting up two test groups:

• Group A. This is a control group. They continue to see this product’s pricing at $29.99.

• Group B. This is your test group, or experimental group. They see the same jar of moisturizer listed at a higher price, $34.99.

You’d then run a pricing split test for a period of time—usually weeks or months, not days—and measure revenue collected at each price point. Perhaps the higher price will bring in more revenue by boosting your average order value (AOV). Maybe you’ll find the lower price maximizes revenue by boosting the total number of orders. Keep experimenting until you find the actual price point that brings in the most money without diminishing customer satisfaction.

Are there any downsides to split testing?

Split testing gives you a concrete, data-driven way to determine how much consumers will pay for your products and services. However, this approach comes with risks and complexities. Here are a few worth considering:

• Potential customer alienation. You could alienate your existing customer base if you start offering lower prices to some of them and higher prices to others. Customer loyalty is a valuable asset, and you may not want to sacrifice relationships over small pricing decisions.

• Lack of statistically significant data. For a split testing price experiment to offer statistical significance, you need a large enough sample size and sufficient traffic. This can be challenging if your business currently has a limited customer base or sells niche products.

• Uncertain findings. A split test may not account for long-term customer behavior or seasonal trends. What works now might not scale as your target market expands.

• Implementation difficulties. Managing multiple price versions can complicate your inventory, sales management, and performance tracking systems. The process may strain your operational costs, especially if you need outside vendors to oversee the process.

Alternatives to split testing

Split testing isn’t the only way businesses find the optimal price for their offerings. Consider these alternative price testing methods:

• Customer surveys.Well-designed surveys can provide statistically significant insights into customer preferences while avoiding the risks of live price variations. Most notably, they can provide quantitative data on consumer willingness to pay at different price points. You can also ask how customers perceive specific features or benefits and what they think they’re worth. 

• Customer interviews. Compared to surveys, interviews are more personal but harder to conduct at scale. You can speak to existing users of your products as well as new ones who fall within your target audience. You can ask direct questions about pricing levels and let the responses help shape your strategy.

• Competitor analysis. While not directly involving your customers, analyzing competitor pricing—core rates, discount strategies, subscription offers—can provide a market benchmark and help you establish the ideal price for your own offerings.

• Price sensitivity management (PSM). The Van Westendorp Price Sensitivity Meter is a market technique businesses leverage to directly ask customers about their perceptions of different price points. Specifically, customers rate prices as cheap, expensive, too cheap, or too expensive. This approach can help you achieve price optimization without the upheaval that may come with actual split testing.

How to split test your pricing

1. Define your objective
2. Test one variable
3. Segment your audience
4. Run your test
5. Analyze your results

Here’s a simple five-step method for running a split test to determine the ideal price for your merchandise or services:

1. Define your objective

Begin by identifying the purpose of your pricing test. Are you seeking increased revenue, higher conversion rates, or improved customer lifetime value (CLV)? Each of these objectives might call for different pricing strategies, so it’s crucial to know your objective before you start testing.

For example, you could decide to test two different monthly subscription prices for your software-as-a-service (SaaS) company’s premium plan—$89 versus $99—to see if you can charge more without putting a dent in your conversion rate. 

2. Test one variable

Unlike multivariate testing, which measures many variables at once, A/B split testing focuses on one variable at a time. If you want to test more than one pricing variable—such as one-time purchases and recurring subscriptions—run those tests separately.

For instance, you could plan a test that offers a percentage discount to people who subscribe to an annual plan (e.g., a monthly $9.99 subscription goes down to $100 flat for a full year). However, you wouldn’t concurrently test another variable, like making the cheaper plan offer fewer features.

3. Segment your audience

Split your audience evenly and randomly so each group sees a different set of prices. If you’re A/B testing using pay-per-click ads, you could direct each group of users to a different landing page. Group A could be sent to a pricing page with old rates, while Group B could be sent to a different page with your new rates. You could even create two sets of individual product landing pages to focus your split test on specific products.

4. Run your test 

Tests must run for a long enough duration to gather statistically significant data. Long-running tests help account for variations in traffic and purchasing behavior. For instance, people’s online shopping habits may be different on weekdays than on weekends. Some products experience robust sales demand in the summer, while products in the same category sell slowly in the winter. An A/B testing software solution like Intelligems or Shoplift can facilitate this part of the process.

5. Analyze your results

Once the test concludes, analyze the key metrics for each pricing variation. Determine which price point led to the best outcome based on your initial goals. Perhaps you’ll discover that a dynamic pricing strategy, where pricing changes based on short-term market indicators, boosts short-term revenue. You might separately find that a price hike coupled with discounts for repeat purchases can boost customer loyalty. Consider running multiple tests until you arrive at the right price to maximize your profit margin while retaining your customer base.