Thomas Ardal

Entrepreneur and founder of

Lean Startup on - Split testing

This is the fourth post in a series about Lean Startup on Check out the other posts here:

  1. Introduction
  2. Minimum viable product
  3. Continuous deployment
  4. Split testing
  5. Metrics
  6. Pivot

Split testing or A/B testing is one of the practices from Lean Startup that I really enjoy. Making a few adjustments and seeing conversions or similar increase, is very enjoyable as a business owner. I split test everything from button colors on the website to pricing of the paid subscriptions.

For website split testing, I’ve used Google’s AB test tool (no longer available) in the past. A few years ago, a colleague showed me Optimizely and I was sold. If I remember correctly, the lowest plan on Optimizely was still quite expensive previously. But today, you can use most of the features for free. Optimizely offers a nice real-time editor of your site and it lets you create multiple variations. You can use it to test locations of buttons, colors, flows and much more and I really recommend for you to check it out.

Converting users into paying customers is pretty much the most important aspect of running a successful SaaS business. Email marketing is probably still the best converting channel for acquiring paying customers. I spent the first years thinking that emails were an awful and old-fashioned way to attract customers and did everything I could not to use that channel much. Intercom changed my perspective on this and offers a nice product to send emails to both trial users and paying customers. There’s built-in support for split testing some of its features like auto-email, which I use to experiment with new email templates, email content and more to increase conversion from emails. The split testing feature in Intercom works pretty well, but lacks some reporting and support in all parts of Intercom for sure.

Experimenting with prices and the length of the trial period is something I’ve tried out as well. started out with a free plan, but after numerous attempts to increase the conversion from free to paid, I decided to switch to a trial based approach. The length of the trial period have been between 2 and 4 weeks and I’ve currently settled on 3 weeks. Similar attempts have been tried out with the price. I started out with prices as low as $9 per month. I currently offer five different plans, but will start experimenting with fewer later this year. When I started working on 4 years ago, $9 per month was pretty much what people would pay for SaaS. Luckily, it seems like people have begun to understand what is required to run a SaaS solution with near 100% uptime.

Show Comments