The basic concept is to prove things in tests before betting the farm. Several rounds of rigorous testing at various levels are standard prior to the release of any drug. Drugs can obviously be life or death situations, but the point of building confidence that something works before fully implementing it is important. Many companies are now planning new prices for 2018, and within each of those companies there are undoubtedly people who think prices should be higher, and others who think they should be lower. Testing multiple price changes can help settle the debate.
A/B testing is a simple method you can use to evaluate different price points and determine which is superior. For a short but reasonable time offer the same product to similar customers, but at different price points. At the end of the test period, measure the number of customers who bought the product or service at each price point, along with the total revenue, margin, and operating income. Then use the price that delivered the best results.
To make your test valid (and therefore worth relying on), there are a few principles that must be incorporated:
- Do your test on a sample of customers, not all customers. The more customers you include, the bigger the bet you are making. The whole reason for testing pricing strategies is to build confidence in your pricing decisions by obtaining evidence of what works in modest bets.
- Separate the customers into like groups. If you think customers in one location are more price sensitive than in another, and then you offer the lower price to the price-sensitive location, you are biasing the test. Make sure the groups are similar, but the assignment of which group(s) get which price should be random.
- It is critical that you test one variable at a time. If you are testing prices, everything must be the same except the price. If you make any other changes, you won’t know if your results are driven by the price difference or the other changes you made.
- Tests must run long enough to get a reasonable number of transactions that can be evaluated. If you run your test for too long, you will gain confidence in your result but suffer lower returns on the transactions occurring on the inferior price. However, if the test is too short, you will not have statistically valid results. The proper duration will vary from situation to situation. The more transactions that can occur in a time frame, the shorter the time frame needs to be.
- You can run multiple price tests at one time by creating multiple groups. All the same principles still apply, but you are offering different prices to multiple groups of similar customers.
Testing can be used to make choices on a variety of strategic and tactical decisions other than pricing. For example, you could offer beta products to small groups of customers to see how they are used and how well they are adopted before committing to full-scale production. You can also develop tests to determine if promotions make a difference in sales by offering the promotion to some customers but not to others. In any of those examples, the point of testing is to move beyond guessing and obtain data to make better decisions.
Data-driven pricing strategies are far superior to those based on hope, gut feel, or whoever shouts the loudest. Simple tools like A/B testing can help you develop those strategies.
Leave a Reply
Want to join the discussion?Feel free to contribute!