Possible Duplicate:
How do you persuade an organization to value UX?
How can I justify UX processes?

How are you justifying UX effort e.g. redesigning a page to be less cluttered or just visually better appearing against numbers like click through rates and business KPIs?

An example in more detail: You want to redesign a start page und you have good evidence from user tests that people are going to like it and that it would be more usable. But you don't get the resources in your team or company to get it done. The answer you get is simply something like: "Why should we do this? It really works good and the numbers are also good. We are making good money."

Is the only way of choice to build an A/B test which will probably show that also click rates get better through an improved design.

But what about things which are simply not measurable? What if your new design is just more competitive or reflects state of the art design?

When is the point where the head of the company just has to support UX and has to say: We want to make it, even if it will not give us immediate boost in numbers?

  • 1
    Similar question asked before: ux.stackexchange.com/questions/4567/…
    – Sheff
    Oct 31, 2012 at 11:41
  • don't think so, because this one is very specific. You could call it Data Driven vs. User Centred Design ;-)
    – FrankL
    Oct 31, 2012 at 15:31

3 Answers 3


But what about things which are simply not measurable? What if your new design is just more competitive or reflects state of the art design?

  1. Surely if it's more competitive then, by definition, that is measurable?

  2. Why should the business care about state of the art design? If I cannot explain why in ways the business values - that is the problem.

When is the point where the head of the company just has to support UX and has to say: We want to make it, even if it will not give us immediate boost in numbers?

Personally, and this may seem harsh, that point never comes.

If I cannot justify and explain my decisions in a way that the business can value then they shouldn't just do what I say.

If the new doesn't doesn't move the numbers that the business cares about why should they do it? Why not spend the money elsewhere on something that will bring value to the organisation?

I'm not in the business of producing art. I'm in the business of making people's products and organisations great. If something is not providing value - why am I doing it and why should the business care?

So - what are the real issues:

  • The business doesn't trust the UX folk. They don't understand the value that UX can bring. The best way I've found to fix that problem is by a series of small wins. Do something that makes something the business values better. Fix that niggling problem that shows up in the usability test that increases conversions by 1%. Or whatever. You can then use the small win to argue for more resources for the slightly bigger win. Repeat until you build trust to do the things that need doing.

  • The business has only a short-term tactical focus. They're only thinking about the short-term wins and missing the longer term advantages that changes might bring. Fixing this probably means dropping out of the design/ux comfort zone since these are not UX problems. Only having tactics, not strategy, is a business problem and fixing those issues involves a whole different chunk of skills and techniques from many of the UX ones.

  • The business values don't match the designer's values. The designer who values premium-brand, high-value, user-focussed values is never going to be happy at Wallmart. Because Wallmart are not in that business. They will never be. They fill a different niche. The need design, excellent design, but it won't be the kind of design that somebody who aspires to work somewhere like Apple will want to do.

  • Sometimes the "designer" is in the wrong. They want to be an artist and use the company's products as their medium of free expression. They, fundamentally, don't care about the consumers or the business. They care about their art. They should go find a new industry.

  • 1
    Thanks, really good points. Your Wallmart comparison is also a good example. And you are right: If you can't explain it and if it doesn't help the numbers, maybe you should not build it. But it is difficult to foresee some numbers to justify what you want to do. There is always to possibility to AB test. But larger changes are not so easy to build in parallel to daily business and so it is hard to find some number. But maybe the most important point is the lack of a clear strategy. That makes it really hard to work towards a common aim.
    – Rainer
    Nov 1, 2012 at 9:49
  • best you try to convince product management for ux goals and build strategy together. by definition PM is setting goals for the product
    – FrankL
    Nov 1, 2012 at 11:32
  • I don't 100% agree with your second point; there's also a role for "tactical UX"; making things recognizable instead of leanable, following conventions instead of making a unique identity, appealing to new users. Those can be UX choices too and can be designed around. The differences is whether UX is being ignored or whether goals are different.
    – Ben Brocka
    Nov 1, 2012 at 13:20
  • FrankL, right you are! But UX should also be involved in setting goals for the product, really. Everything you decide should be based on product/company vision and strategy (of there is one). But you cannot split PM / product decisions from UX decisions because the go hand in hand.
    – Rainer
    Nov 1, 2012 at 13:23

There is a wealth of studies looking at subjective things like aesthetics and their effects on the user that you may be able to use.

The Effect of Aesthetics on Web Credibility argues that the look of the website can make you look more or less legit.

Attention web designers: You have 50 milliseconds to make a good first impression! says that it only takes 50ms for people to make up their mind, hence the importance of aesthetics.

Don Norman even argues that beautiful is sometimes more impactful than usable: Emotion and Design: Attractive things work better. This has to do with perceived usability instead of actual usability, as explained in What is beautiful is usable.

  • Thanks! Will dig into these articles. Some of them are new to me and I appreciate your help!, really
    – Rainer
    Nov 1, 2012 at 9:53
  • Good links, thanks for putting them together and sharing them!
    – Rainer
    Nov 2, 2012 at 18:44

If you are encountering resistance, it is likely due to some other underlying situation that is being masked by the KPI conversation. However, presuming that your question is indeed all that's going on, I'll try to answer it.

By definition, if the change matters, it will be reflected in the KPIs. That's what KPIs are for.

In practice, it is rare to (a) select the right KPIs and (b) be able to differentiate the effect of a particular change from noise or from other changes.

It may well be that you are not justified in wanting to redesign the start page--they may be right, it may in fact work well enough. There may be more bang for the buck in doing something else.

Or it may be that the change will have a big impact. If it increases user satisfaction, does that in turn impact a KPI like user retention, resale rate, or recommendations to others? If it increases user efficiency, likewise?

Changes to things like user retention will not be felt for some weeks or months. Does your business track trends over those time periods? Why not?

  • Yeah, also right. Being able to differentiate the effect of a particular change from noise or from other changes is definitely a challenge. User retention is a very good example for things you cannot really predict beforehand. Of course you can (and should) test IxD and Visual. But in the end, nobody knows really and you -if ever- can only be sure by AB testing it. But isn't there some point where you just have to realize that you will be no more if you don't catch up with competitors at least or even better, be ahead of them by doing something new (and that needs some courage)?
    – Rainer
    Nov 1, 2012 at 13:22

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