Tell me more ×
Answers OnStartups is a question and answer site for entrepreneurs looking to start or run a new business. It's 100% free, no registration required.

Is it legally/ethically bad to run a split test on the price point of your product?

Temporary coupons don't test the right things and introduce other variables into the experiment, and changing the set price for an extended period of time wrecks havoc with the P&L towards the end of the quarter.

But if a potential customer sees two different price points, or if her friend purchased the same product on the same day from the same place for a different price, that can lead to a sort of support ticket we'd rather not have to deal with.

So ... accepted ways to test different price points? A good way to run an A/B test on a price point? Thoughts?

share|improve this question

5 Answers

One strategy I have been thinking about, is to have a sales form on the website which the customer fills out. It is worded as if it is a sale and displays the price. The customer clicks a check box and an order button but instead of taking a payment, the site sends an email to me and the customer is told thanks for their interest and we will be in touch shortly to finalise their order.

Depending on your sales volume, you could use this method to split test the price over a short time period. It still tests the intention to buy at a particular price point accurately without the hassle of taking payments from potential customers. You could then offer all respondents a special deal once you had finished testing, so they all end up paying the same price.

I am assuming you are selling over the web.

In the bricks and mortar world, the same stuff is sold at multiple different prices - sometimes in the same shop!

Hope this helps!

share|improve this answer
I like this one a lot -- all's well that ends well. – Jason May 17 '10 at 13:42
2  
Another way to automate this suggestion, is to ask an additional question to everyone "How did you hear about us?" and then 'reward' the higher-priced group with a discount for participating. – JeffO May 17 '10 at 16:44

Whatever you do is going to involve placing cookies on the user's machine. You have to ensure that once you offer a price to a specific customer that customer always gets the same price, whether that customer uses the back arrow, leaves your site and comes right back, or comes by tomorrow to buy. This turns out to be very hard to do in practice.

share|improve this answer

Price discrimination is common in big box retail (location based) and the ideal pricing structure is perfect pricing where a product is sold at the maximum price each customer is willing to spend for the offer. So geographic location is used by the big box and the IP address that is incoming does show where the person is coming from to a degree.

Also many retailers online, use profiles to work toward perfect pricing where they profile of someone with heuristics to identify that this shopper will do x amount of product/price research and is estimated to be x elastic based on the profile of this shopper compared to others with similar characteristics.

So I'd really suggest that the black and white A/B test should be challenged because it depends on the characteristics of what it is being offered and the diversity of the site’s consumer.

Consider the characteristics of what is being offered. So if the offering requires a customer relationship of one time then this may be a much simpler price test that you could do. You could use Google ad words to land people on different prices with the same ad copy reaching different price points for different key words. Also testing could be done with Facebook advertisement, which is not search driven, but demographic driven and provide different price points to get an indication of demographic price elasticity. The whole consistency of price only matters if the customer behavior is to go back and forth to the website. If the customer normally buys or abandons on the first landing then putting in a cookie or profile method is unnecessary and a simple A/B is all that is needed. If the customer does research before pulling the trigger, which search engines have made to be easy, then having a cookie for price behaviors is needed. If the site is a diverse complex offer (Amazon like) where greater profiling and price determination could be done then the situation is far more than just a simple A/B test to maximize profits but customer profile testing to determine like customers and what is their price elasticity for different offerings.

share|improve this answer

This is an interesting question because price is something that is kind of hard to test.

I don't see any legal or ethical issues with this as long as you offer your customers a commitment to honor a lower price if they find it elsewhere.

One method to test price is to have a bid or auction or just have them name their own price (ala Priceline). If you have a big enough sample space and the sampling is random enough, that should tell you the optimum price for your service. It's a bit more complex than an A/B comparison but a lot more instructive.

You could even throw in a fixed price option for a random selection of people to get a control group but you might already have that data from your current pricing model.

share|improve this answer

I thought it was worth adding that almost every international airline and most international software companies offer identical products at different prices in different markets as a course of business, and 'price testing' happens in almost all markets.
Avoid upsetting the customers but carry no guilt for whatever strategy you employ- I don't see this as a question of ethics.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.