Server-side A/B Testing

Server-side testing is an approach to A/B testing where variations of the experiment are returned directly by the back-end development stack. It varies from typical client-side experimentation or conversion rate optimisation since it does not typically rely upon front-end JavaScript tags.

Learn more about what we do

Our Partners
Our Partners
Our Partners
Our Partners
Our Partners

What is the benefit of server-side testing?

The main benefits of server-side testing are performance, elimination of the flickering problem and better alignment with product teams. It does however bring with it a unique set of challenges.

FAQs about server-side testing

We've received a lot of questions about server-side testing over the years from customers, clients and stakeholders just like yourself. Some of these came up in the recent DigitalEliteDay 2020 virtual conference. Here's a few of them.

How does server-side testing work?

Server-side testing is a method of A/B testing that allows you to leverage back-end technologies to handle processing before anything is sent to the browser (i.e.: in the backend technology stack via some sort of server-side SDK type approach.

What are some of the server-side tools out there?

Optimizely Full Stack, LaunchDarkly, VWO Full Stack and SiteSpect are just some of the server-side testing tools that are available on the market, but there are many many more.

What programming languages does server-side testing support?

Most vendors have SDKS for a plethora of programming languages including Java, C#, Go, Objective-C, PHP, Python, React, Ruby, Node and Swift.

If you are using something a bit more obscure or low-level, it's likely not that difficult to come up with a new SDK that does not currently exist to support it. Most SDKs are fairly small and easier on the portability scale.

What are some of the potential SEO impacts of server-side testing?

They're less well known or agreed upon in the experimentation community right now. In the old days, cloaking was a black-hat SEO technique in which the content presented to the search engine spider is different from that presented to the user's browser.

Theoretically, if you are not careful about the way your server-side tooling tracks and measures (or more accurately, should exclude entirely) robots and crawlers -- your server-side experiments (if large enough in nature which they often are) might unintentionally cause your actions to be misinterpreted as cloaking by major search engines.

While Matt Cutts commented back in 2012 that testing is mostly safe or 'approved', he was referring to JavaScript-based changes which are very different to the nature of server-side experiments. The IAB/ABC International Spiders and Bots List is a good starting point for sane exclusion lists.

How can I track events with server-side testing? Can I use my goals from Google Analytics?

Unfortunately, it is generally not very straight forward to re-use the same existing goals or events that may exist within your web tracking tooling. If you have a well constructed data layer (client-side), this may not be easily accessible from the server-side so thinking about the overall application architecture is important.

If you'd like to explore some integration POCs, let us know.

How can I distinguish between logged in users and non-logged in users with server-side testing?

It depends upon the architecture and structure of your application (and what language it's written in). Generally, to make a server-side testing engagement go smoothly requires some degree of ensuring alignment between marketing team and product/development teams. If you'd like to understand more about how we may be able to assist, get in touch.

A few companies we’ve helped generate massive optimisation ROI

(... we help smaller companies with high traffic too, but you have to have at least 100,000 unique visitors monthly for it to make sense)

People walking
Logo Cathay

We are very, very happy with our collaboration with Web Marketing ROI, and greatly appreciate their partnership.

View the case study
Logo BetEasy

Web Marketing ROI are experts in optimisation and helped us to achieve a regular rhythm of experimentation and continuous improvement

View the case study

Want to discuss your conversion rate optimisation, UX or A/B testing program today?