Guides and tips for price optimization with FutureMargin Shopify app

Finding the most profitable price for a product

In this tutorial, we will analyze historical profit conversions and learn how to modify prices to achieve greater profit.


The approach takes several steps and the process will cover a several week long pricing test.

  • Select product
  • Launch pricing test
  • Evaluate results

1. Select product

Select a product with enough historical data. This means that the product should be a popular item which receives at least a couple of orders per day. For this test, we assume our product is a special branded camp t-shirt.

Our purchase price (cost of goods) for this t-shirt is $12. We would normally sell this t-shirt for $20, but we don't know if this is an optimal price for maximizing profit. Thus we will test 2 price options: $20 and $25.

2. Deploy pricing test

Our test will consist of two pricing periods. We will set the t-shirt price to $20 for two weeks, and we will set the price to $25 for another two weeks.

Keep in mind, that the aim of price testing period is to collect enough data - sales and visits to be able to meaningfully compare the conversions of different prices. If enough data is collected, we can use a statistical test to compare the results.

Here are some tips for optimal price testing:

  • Timing - Pick two time intervals, where you expect similar amount of traffic and similar conversions
  • Isolation - Don't AB-test other product parameters during the pricing test
  • Duration - Generally try to test each individual price point for at least two weeks

3. Evaluate results

After testing of all desired price points is complete, we can find out which price is more profitable. Go to Profit Analytics page in FutureMargin dashboard and select the correct time range. In our case, the time range spans 28-days.

To meaningfully compare the results, it is important to normalize total profit with number of visits for each given period. FutureMargin automatically does this for all key metrics. Finally, we compare profit per visit for the two tested prices and we see that profit per visit is greater for the price point of $25.

We can now also observe actual visit counts and various conversion metrics and graphs for the two periods.

We also check the value of statistical confidence. In this case, the value is high enough. If it is not, we should increase the duration of each testing period, to collect more data for evaluation.