A/B Testing

Identify the changes that will make the most impact to your conversion rates and revenue

Address identified problems and test new ideas before you build them, to increase conversion rates and revenue from the people already visiting your website.

Our goals with A/B testing are to make more effective changes, reduce the duration (and cost) of each test, and improve the overall speed of your experimentation programme, to have the greatest possible impact on revenue.

Experimentation, Strategy

Our Promise

We only ever have somebody running a project who has been doing CRO for a long time.

Website experimentation that drives continuous growth

About A/B Testing

A/B testing is one of the fundamentals of CRO, but everyone has their own process. Here at Good Signals, we perform proper controlled A/B/n testing with the help of leading tools like Optimizely and VWO and follow the scientific method. We use strict stopping rules for tests to ensure trustworthy outcomes and to avoid imaginary lifts.

Before testing, we always derive tests from data-backed hypotheses that are then mapped out by priority. All tests run for a minimum of 2 business cycles (usually 2 to 4 weeks) and we make sure to never call them early.

We ideally want to test as many variations as the traffic allows as there’s never just one right solution to a problem. We come up with a bunch of ideas to tackle the problem, then based on our own experience, common sense, intelligence and what we know about the industry we pick some of the better ideas and create treatments.

Our main goals when testing are to make sure the tests we do are the most effective options, reduce the duration and cost of optimisation (without sacrificing effectiveness) and to improve the speed of experimentation – the process includes everything from turning data into a hypothesis, a hypothesis to designing a treatment, to coding it up, to launching the test – which can involve the sign off process, brand, compliance, the CEO, etc.