There are three main sections to a good hypothesis.
These are:
- An identified or presumed problem
- The solution that is proposed to fix the problem
- The result that is expected as a result of the change
The problem can be identified in a number of ways, through analytics, research, feedback etc. In many ways the problem is separate to the hypothesis, so we can have a problem that says “Customers are finding the language we use on the site too complicated and it is causing anxiety and friction". This would be followed by an hypothesis that says “We want to simplify the language used on the site. This will increase CTR and lead to greater engagement".
The solution is all about understanding the cause of the problem. This can use personas and psychology to understand the impact on the customer. In the case of the technical language, it causes confusion and doubt – this leads to the proposed solution of “simplifying language".
The result anticipated is that the customer will have less anxiety and be more likely to progress through the process, leading to a higher CTR and increased engagement.
To build a working hypothesis, first fill out the form below
KRAKEN DATA is software agnostic, we work with your A/B testing software of choice. Amoungst the ones we support and recommend are:
- Webtrends Optimize
- Optimizely
- Adobe Target
- ABTasty
- VWO
To discuss your optimisation or analytics programs, please contact us.