Set-up and run A/B-testing

From CRO specialist with experience in A/B-testing from 2016го
On one of the three most advanced platforms
On your ideas or my practice and research
On the high and low traffic

This is suitable for business websites with traffic from 15K/month. On you want to run A/B or Split URL-testing. If you need help from CRO-specialist with experience in setting up, running tests on the main platforms: Google Optimize , VWOOptimizely , with a bag of successful practices and the ability to form and prioritize hypotheses.

Test algorithm

If there are already formed hypotheses - I see the process of the test in four stages:

1. Pre-test analysis

We agree on KPI - for the evaluation of test results. I estimate the optimal test duration and statistical error.

2. Test setup

I spend setting up the software in the platform for A / B testing. I set targeting, conditions, goals, create test versions of pages.

3. Running test

Immediately check for any conflicts. Then I track test results to assess the dynamics of change.

4. Post-test analysis

I evaluate the test results: the truth / falsity of the test hypothesis, KPI values, statistical error. The report recommends the introduction of a winning option.

Collaboration formats

Choose the option that best suits the level of development and needs of the project:

PRO

Price: upon request

First, search for hypotheses through research (usability, web analytics), then setup.

Standart

Price: upon request

Technical implementation of hypotheses from experience, best practices for similar projects.

Basic

Price: upon request

Only technical implementation (customization, launch) of your ideas for A/B tests on platforms.

Answers to basic questions

Do I guarantee the effectiveness of A/B tests?

It is difficult to guarantee the results of A/B testing, but I offer a 30-day money-back guarantee. If you aren't satisfied with my hypotheses and the results of the A /B tests, you can return your money in full (packages "PRO" and "Standart").

Are there any additional costs?

Not. You only pay for my time involved in: a) setting up and running A / B tests, sometimes b) forming hypotheses. You only need paid access to the testing platform (if there are no free/trial subscriptions).

How long do the A/B tests take?

Much depends on the amount of traffic, macroconversions, microconversions. The more of them - the faster the test. The minimum time for 1 test is 7 days, the maximum is 4 weeks. We will discuss and coordinate all these details at the start.

Do I implement code in the A/B-testing platform?

I implement the main tests myself (I have HTML, CSS, JS skills), if I need to prepare a Split URL version, configure server tests - I'll coordinate your developers. Or for an additional fee I attract my developers.

What to do if there is not enough traffic?

Sometimes even 15K traffic is not enough to correctly and reliably conduct a test. For some, it is A/B tests that are not suitable and I suggest: a) Split URL tests for more tangible changes, b) user tests or c) add PPC traffic for the duration of the test.

What else is important?

Testing is an indicator of project maturity. At this stage, web analytics should be fairly deeply tuned. Not only final macroconversions should be tracked, but also intermediate microconversions, funnels.

Submit a request for A/B-testing

What is most often tested?

Elements : CTA, text, buttons, headers, product name, order form, special offers, links, number of goods or services on a page, navigation, cross-sales, additional sales, product location, etc.

 

Pages : main, popular target, registration, authorization, funnel entry, product page, category, ordering, promotions, subscription.

Money back

If you are not satisfied with my hypotheses and the results of A / B tests, you can fully return your money (Premium and Standard).

Confidentiality

I promise not to shine data about your company / site by default. Only by agreement on the shore.