In creating responsive rich internet applications we've begun to run into a lot of little UI issues during our testing phases and we're looking to develop a strategy to navigate this period quicker. One idea is to create tiers of what we find acceptable on certain devices. This is a very rough idea, but I was looking for input from the community.

While our analytic reports indicate almost 500 different devices, there are clear leaders that may deserve more attention than others. As depicted in the image below, iPad and iPhone users are the clear majority in our user base, but all of the other devices added up still equal 48% and throwing out those users' experience completely would be an error on our part. Nor can we suggest that our users change their device, as was common place in the desktop browser realm.

enter image description here

  1. Do you think a tiered approach, with different levels of acceptability based on the amount of users is a good approach? Are all your users' experiences created equal?
  2. Would there be other criteria to consider beyond, how many people use a given device?

Rough sketch of the idea:


download bmml source – Wireframes created with Balsamiq Mockups

Of course in a perfect world, we would want to create the perfect experience on every device, but I believe it is not always worth the cost. There will always be a minority who feels some UX pain based on their device/browser/OS etc... (ask any linux user).

3 Answers 3


There's no blanket answer to this. It really comes down to what specific UX decisions are being made for which particular interactions and functionality per device.

The intent of a responsive site is to be as device agnostic as possible, but of course, there are always outlier devices that will cause you some problems.

At some point, a business decision has to be made. Is the cost of tweaking the UX to accomodate particular device discrepancies worth it for that device? One needs to take into consideration the amount of effort to create work arounds, the effort it will take to support and maintain it over time, as well as factors like how long will that device still be viable on the market.


Tiered approach based on number of users possessing a certain device is too holistic, further segmentation is still needed to understand user acceptance trends. In chart, on " Goal Column", instead of using subjective terms, for instance, "Perfect", you may mention that iPhone has the highest penetration in market. User acceptability is very subjective multi-faceted measure and there are many influencing and moderating factors driving user decision making process. The psychological models such as Technology Acceptance Model (TAM), Theory of Planned Behavior(TPB) and UTAUT model describe various factors responsible for user decision. You can refer to these approaches for more detail.


In terms of helping to communicate design considerations to the business units, I think it is quite a good idea to be able to specify things more clearly than to just try to cater for everything or only certain things. Also, in terms of communicating specifications between businesses or clients I think it is also quite useful. If you define different tiers for UI/UX then at least you can provide an unambiguous 'level of design service' that can be compared across different types of software products and services.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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