Introduction to Experiments

Experiments are Cubb's secret sauce! They allow you to create a library hypothesis-driven CX improvement opportunities that turn your insights into action. 

A crucial part of making customer experience improvements is running CX experiments. Creating customer personas and mapping customer journeys gives you a strong understanding of who your customers are and what they want, but that's not enough - now you need to make real improvements to those experiences. How do you do that? By coming up with a range of ideas for CX improvements, deciding on which ones to test, defining a clear hypothesis, and then running an experiment and gathering data to find out whether your changes had the impact you were hoping they would.


Below are the steps on how to create and implement experiments on your journey map:

1. Open a journey map.

2. Hover your mouse at the top of the step that you want to add an experiment to. You will see a small green test tube icon will appear. 

3. Click on the test tube icon and the experiment tab will appear from the right-side panel. You can add different experiments for each step you have on your journey map.

A sample screenshot below.

 You will need to parameters like the status of the experiment, experiment start date, experiment end date. We provide prompts to make it easy to structure each experiment as a hypothesis-driven activity. When you add details the system saves automatically.

Testing with hypotheses

One of the biggest and most common mistakes made in CX experimentation is the failure to define a clear and measurable hypothesis before going ahead with a test. A test without a measurable hypothesis gives you no valuable insights, because it's impossible to tell whether you achieved what you set out to. 

A test with a clear and measurable hypothesis makes it easy to see the data that you're looking for, and whether or not you were successful in your experiment. 

Here are examples of good hypotheses:

  • Our new landing page design will result in 30 online sign-ups over two weeks.
  • Our improved customer welcome pack will result in a 15% improvement in new customer satisfaction ratings over one month.
  • Our simplified workflow will result in a 2-day average reduction in customer processing time over one month.

You can also view the experiments at the "Experiments" tab at the top.

Running CX experiments

When you're running CX experiments, you need to embrace a lean testing mindset, which means being willing to move quickly, give something a try, and accept failure as part of the learning process. The design of some experiments might be simple, but some CX improvements will require you to think outside the box about how to create a prototype that will enable you to collect the data you need. In addition to these flexible and creative approaches, experimentation requires an analytical and scientific mindset; it's all about collecting accurate and reliable data about what works and what doesn't.

Test, Learn, Repeat

CX experiments aren't one-off events. The idea is not that you test an improvement, evaluate its impact, and then go back to business as usual with that small tweak incorporated. Instead, CX experimentation should become a constant cycle of testing, evaluating, and learning from new ideas and experiments. Some experiments will be instantly successful, and you'll start to incorporate those changes into your standard operating procedures. Some will have mixed results, and you may need to adjust your approach slightly and try again. Some will fail entirely, which will give you important insights and the impetus to choose a new idea and test again.