0

have a webapp where I make http post requests and I wanted to improve the form where users post advertisements. Now when I improved it, how do I actually measure or know that I made an "improvement"? The load time is quicker, the http requests are fewer, but how to I measure usability?

The prior version of my ad posting page looked like this and was successful.

enter image description here

The new version looks like this and gets better results in tests.

enter image description here

Should I accept my own redesign already, or should I improve it more?

  • 5
    The redesign looks much better (although I'm not sure about the massive margin between labels and inputs? and the alignment of the whole page looks a bit off). If you're asking about usability testing though, you're probably better asking on User Experience – Cai Jun 29 '16 at 14:16
  • Is your question "which is more usable" or "how can you determine which is better for users"? – Zach Saucier Jun 29 '16 at 14:17
  • I know you didn't ask for a critique and this is nit-picking here, but the progress bar was a little confusing that it didn't show the progress until you hit the submit button (if that was the purpose of that bar). – AndrewH Jun 29 '16 at 14:21
  • @AndrewH The progress bar is an interesting problem. I'm really glad that you mentioned it. I acknowledge that I have a problem with it and I'm working on it but the response time seems to be divided into both uploading the data and waiting for my server code to respond, so there could be two progress bars, one for uploading and one for waiting for the server. Do you agree? – Niklas Jun 29 '16 at 14:44
  • 1
    @Programmer400 Ok, that makes more sense. If the function of that bar is to show the response time between the send and the server then I would hide that bar until the user hits the submit button. Other suggestions: Show that bar completely empty with a percent #, and turn green when 100% to show successful submission with written text. I don't think a progress bar for the form itself is necessary unless the form is multi-page but that is just my opinion. – AndrewH Jun 29 '16 at 14:49
2

So I am not sure what it is exactly you are looking for so I am just going to take the opportunity to give you feedback on the design.

  • I think the labels are too far away from the input boxes, this can me it confusing for users to keep track of what they are filling in.

  • The social media buttons are far more prominent than the submit button. Not online in color but also in size. Either create more space or make the submit button more prominent. Plus I think that content like social media and copyright belongs in a footer, not right under your content.

  • My OCD can not handle the fact that your content is wider than the content in your header bar.

I threw together a (not so pixel perfect) example; example image of improvements

  • Thank you. I was looking for this kind of feedback, because I'm not a designer of graphics. – Niklas Jul 2 '16 at 17:57
4

To measure the success of two or more website versions you can use A/B Testing coupled with an analytics service to compare a wide range of metrics such as time spent on page, click-through-rate (CTR), buttons/links interacted with most and much more.

Once you compare those metrics with a large enough data set you should be able to see what's working well, what needs to change, and either choose one version, or chop and change parts to bring together a version that combines the successful elements of each version.

Large companies use this method for both small and large differences, from whole different versions to just a different call-to-action.

One well trusted A/B Testing service is Optimizely where you just include a snippet of code on your page and it injects the different versions of the websites.

This is undoubtedly one of the most logical ways to decide between two or more website versions.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.