It's a well known fact that economic and social sciences cannot be the subject of a lab experiment like it happens in phisics or chemistry. Though, when working with variables(clicks,views,sales) that reflect characteristics of real observations(visitors or customers) a 'lab' can make it possible to run experiments

Conversion scientist Brian MasseyBrian Massey calls himself a Conversion Scientist and has the lab coat to prove it. But not the coat is the only one to prove it, but his experience and results that he managed to deliver to his clients. Brian is the founder of Conversion Sciences and the author of 'Your Customer Creation Equation: Unexpected Website Formulas of the Conversion Scientist'.

Brian defines website optimization as 'the science of finding more online revenue without paying to get more visitors'. In the next few lines of the interview you can read about the benefits of conversion optimization and A/B testing.

1.If A/B testing was a company, what would be its unique selling proposition?

Since Conversion Sciences is an A/B testing company, this should be easy. More sales. More leads. Same traffic.

2.What are the most important 5 metrics that you monitor for your clients’ ecommerce websites?

  • Accountant smile rate. This tells us that A/B testing making more money for the business. 
  • Revenue per visit combines conversion rate, which can lie to you, with average order value. You don't want to make conversions go up at the cost of your profit margins.
  • Cart Abandonment tells us how the purchase process is working at each step. It helps us zero in on poorly performing pages.
  • New vs. returning visitors. This isn't a metric per se, but segmentation is a powerful way to find wins.
  • Tests per month. This tells us the velocity and maturity of the testing program.

Other metrics like bounce rate, time on site, pages per visit and exit % are used during our analysis, but are predictive metrics. They can diagnose problems, but don't necessarily predict the health of a testing program.

3.What are the basic steps a marketer shouldn’t miss if his goal is to optimize an ecommerce website for conversions?


1. Setup the digital lab and correlate the analytics to reality. We usually compare revenue reported by analytics to revenue reported by the financial system. See this article from Marketing Land for more details.

2. Develop a long hypothesis list. Using analytics and best practices, list the 50-200 things you think will increase revenue per visit (RPV).

3. Rank the hypothesis list based on the amount of proof you have, the expected impact, the amount of traffic effected and the difficulty of testing.

4. Test in cycles. When you have a winner, use follow up tests to see if you can make the win even bigger.

5. Ask "why" when a treatment wins or fails. Learn from your tests.

4.Does online personalization deserve to be considered an approach to increase revenues?

Yes, but personalization comes after you've begun to understand your visitors through a program of split testing. Testing helps you find your key segments and understand what they need from you in order to buy.

5.What is your advice as a conversion scientist for small ecommerce businesses?

Start by looking at your competitors. Use a tool like Ghostery to see if your competitors have installed analytics and testing tools. If they have, you are in trouble. If they haven't, you can create a serious advantage over them.

Then call someone that will teach you the ropes on A/B testing. We offer a free consultation at