I started working as a UI/UX Designer for a large company with a big website.

Since the developers are busy with developing backend functions, the resources are limited right now.

Should I start designing good UI even though they wont be able to change it globally (for example, forms, modals, etc.)?

Or should I design with the old, bad UI so the users feel consistency until we manage to make an overall redesign?

  • Use modular design that you can replace site wide. I.e. create a new form class then apply to all forms on the site. It will be consistent but better and you can reuse after the overall redesign because its modular. – DasBeasto Jul 11 '17 at 18:13

As lead designer in the startup I work at I have to handle the same situation quite often. I don't think there's one out-and-out solution. That depends on your situation. However, my experience in handling this situation might help you.

We have a clear definition of versions. A version contains a set of features that have to be built. At the start of the development period we sit down together to discuss what the developers and designers have to do to complete the upcoming version.

Lets say the current backend features developers are working on are part of 'version 1'. Every time they ask you for a design that has impact on that specific set of features, you design it as defined for version 1. Even though you might know you want to redesign it for version 2.

Now, in the mean time you're working on the design for the upcoming 'version 2'. In this version you can implement new and improved design elements. Make a clear list of all the improvements you'd like to see for the next version and be sure you can discuss them at the next developer meeting where they talk about upcoming plans.

In my experience, you have to be very clear about these things. We've wasted some time before because I wanted to have better design right away. Having a complete set of features is worth more than having an incomplete set of features with good design. But then again, that's how we have set up our communication between development and design.

Maybe the best thing to do is talk to your developers and try and set up something to work together.

  • out and out did you mean cut and dried? – cat Jul 11 '17 at 19:09
  • 1
    @cat The wording "out and out" works fine here I think.According to dictionary.cambridge.org it means "complete or in every way" – Grumpy says Reinstate Monica Jul 11 '17 at 20:22
  • @cat "out and out" means 'in every respect; absolute.' That's what I meant. I think it's correct. – Nick Groeneveld Jul 11 '17 at 20:29

To be honest, I'm not sure what do you mean, but whatever it is, the answer will always be: do it right

There really isn't a single scenario in which doing something wrong will help anyone at anything. And the whole idea of UX is to IMPROVE experiences and processes. And that includes saving development times.

You say resources are limited. If you do things wrong, you'll increase the need for more resources, leaving only 2 paths:

  1. spend more time/resources to solve something that should have never existed
  2. Launch something incomplete or wrong

As you can see, it's a lose-lose situation.

On the other side, if you do things right from the start, developing the tasks in a modular way, you'll be able to do your work as expected and developers will already know what to be prepared for. Additionally, if they are in developing stage, chances are they will be ready to do whatever they are told to do, including.... proper UX.

Read item 1 on this page (extract below)

Developers’ time is extremely valuable. In a perfect world, developers would spend 100% of their time building awesome new products and features. In reality, an estimated 50% of engineering time is spent on doing rework that could have been avoided. What’s more, fixing an error after development is up to 100 times as expensive as it would have been before.

Final thought:

how do you do things wrong on purpose? One thing is failing trying something we think will work, but it doesn't. We all have gone through that (hence why we test so much). Another different one is to do it wrong on purpose. Personally, I'd need to think hard on how to do things wrong on purpose. I imagine it would be something like "theory and experience tells me things should be like X. Then do something else randomly", but even then it would be something difficult to grasp (conceptually speaking, it's like saying: "I'll unlearn what I know")


Make sure your definition of Good (new) UI is aligned w/ the current paying customers needs. That can give you leverage with management to move forward and devote more resources towards new design (and development).

If you're working for a large established company, I'm assuming they've been doing some things right. That said, it sounds like you've assessed they could use some serious improvements.

You say resources for development are limited; can the same be said for research and testing?

If you can devote 20% to building a case for new design, you might get more leverage to move forward. In the meantime, paying customers will have managements attention.

If you get buy in from the influencers at your organization, @NGAFD has outlined some good points here on approaches to implementing a new version of your product.


I currently help maintain close to 50 web apps.

You should go with the small, incremental changes.


  • resources are limited
  • Big changes would cost a lot of time to the developers, you dont know if they have it
  • Testing might increase dramatically, not only would they need to test for the new stuff, but also the old one that you changed in your new UI.

And I will agree with Mike M answer, your changes might not be something the client asked for, if they break ANY old stuff your entire team takes unnecessary blame.

Big drastic changes usually happen when applications are in old technology and are being moved to a newer one. Or when there are drastic changes in the requirements.

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.