As with all scientific testing, we are constrained by the requirements of needing enough data to be able to make statistically relevant decisions.
- Whilst we do see uplifts in a surprisingly short time, it normally doesn't happen overnight.
- The greater the volume of visitors being tested and the longer the period of time dedicated to testing, the higher your likelihood of long term success with Ezoic.
A normal “pattern” of optimization for a site in the Ezoic system is 12 weeks. Thereafter, improvements are produced as the system tests new layouts and gets smarter about what your users like best. What may work wonderfully on one site, may not work for your users, so we need to test and find out.
Discovery (Weeks 1-3)
During this time period, Ezoic is gathering data about your site and the performance of several major layout changes.
Performance during this time period is normally anywhere from slightly below to slightly above the performance of your old website. Often, you’ll see a slight improvement over this time period.
Optimization (Weeks 4-12)
Ezoic has learned a little bit about your website and its visitors, but it still doesn’t know quite as much as it needs. Once Ezoic learns a little more, you’ll often see a sharp improvement in results, followed by 6-8 weeks of steady improvement.
Perfection (Week 12+)
The Ezoic platform is always testing and always learning. Most websites see gradual improvement month-over-month for eternity. You get the benefit of continual testing and the improvements that an entire team of engineers and data scientists deliver each month. You should mentally prepare yourself to give Ezoic at least 12 weeks to demonstrate its full potential. We speed up the process as much we can by using statistical clustering, etc, but we are at the mercy of the math in that the variation of results is too high on smaller data samples. Because of this, in an analysis of past publishers, we've determined that the number one factor of a site's success or failure on the Ezoic platform is time - the lower the time the system is given to produce a result, the lower the likelihood of success.