A/B Testing your Prices: Don't A/B Test Yourself to Death

If your company sells online, you can do A/B testing. If you are not familiar with the concept, it means that different visitors to the website automatically are directed to different pricing or purchase pages, and you measure which price sells the most. Unfortunately, it is not that simple. If you sell something that does not have a strong reference price, something new, something differentiated, you may end up making your A/B testing in completely the wrong price-band or interval. Like in this example: 

This company had the view that their educational services should have a price around $250, and they did A/B testing around $250 to find out the exact price. But they did not see much change in volume and testing a slew of different price points had taken them about three months with no conclusive result. Pricing research that only took two weeks showed that willingness to pay for the company's services was much higher, and the right price was $399, with a sharp price wall at $400. The company simply fished in the wrong pond. Setting the right price increases sales volume some 50%, and revenue more than doubled. Revenue, the company, left on the table during A/B testing…

I spoke to one of the founders of Uber the other day. In his new company, he is certainly not going to rely only on A/B testing to set prices. To paraphrase what he told me: “I’ve just seen too many companies A/B testing themselves to death!”