Leave a comment
Get the GH Bookmarklet

Ask GH

Let's say you start working with a client with this shitty website.

Is it better to:

a) first completely redesign it to a point where everything is done according to best practices, and then start running a/b tests?


b) start running tests right from the beginning

Let's say it gets <50k visitors/month

  • SK

    Sean Kirby

    almost 2 years ago #

    I'd take it on a case by case basis.

    How is the site performing? If it's ugly but converts well, then I would definitely test before making wholesale changes.

    What is the client expecting? How did you pitch them? If they are expecting measurable results, then you need to be able to provide that.

    What other variables are in play that could skew results? If there are changes in the way the client is promoting their site, you won't know what effect that has compared to the redesign.

    If the site isn't performing well, the analytics show consistent performance over a long period, and the client just wants to start hitting their goals, then you be able to get some quick wins by just doing the redesign first.

    Just make sure you can justify your decision.