There has been a lot spoken about user experience design for interactive products and services, but is there some standard or guideline to capturing user feedback for non-interactive products that has a significant visual design element to it?

I think the difficulty is how to evaluate the different elements of design that contribute to the aesthetic appeal/quality, so the indirect way to measure this is to understand what the users think. Has anyone tried to create some systematic way to evaluate this?

This article discusses some of the principles of visual design, and I can see elements of it being adopted for a survey or questionnaire focusing on each aspect of:

  • Consistency: Do you notice things that don't fit well together?
  • Alignment: Do things feel nice and neatly grouped?
  • Proximity: Do things feel logically grouped together?
  • Contrast: Are things easily noticeable and easy to read?
  • Hierarchy: Are things in the order that you would expect to find them?

Update: I saw this paper that goes into some detail about the study of website design appeal which I think some people might find interesting. It points out some general trends between different demographics that are worth noting:

  • females liked colorful websites more, and colorless websites less, than males.
  • both genders reached their peak appeal at a similar low to moderate complexity level, but females disliked simple websites more.
  • adults aged 41 years and above liked websites with a higher colorfulness and complexity than younger age groups.
  • negative correlation between education level and colorfulness, as well as between education level and complexity. Independent of age, highly educated users prefer less complex and less colorful websites than others.
  • a user’s geographical location is an additional factor influencing appeal.
  • Why do you want to be systematic about it? Design is instinctual creativity... otherwise known as art... and best done by those that "get it". Design-by-committee is negatively and mercilessly rocked because it's never been shown to work well.
    – Confused
    Commented May 16, 2014 at 18:51
  • 3
    If you have to pay an agency or a designer lots of money to produce a logo or a brand identity, you would probably want to know whether it was designed based on good principles and whether people like it or not wouldn't you? Commented May 16, 2014 at 23:20
  • It's a faith based exercise. Regardless of the money.
    – Confused
    Commented May 17, 2014 at 1:13
  • An interesting thing for you to think about Michael, is that some of the greatest products in history weren't designed for a test group. A designer that "gets it", finds a problem in the world, and creates a beautiful and effective solution for it. I've never seen a single beautiful, effective product that did not succeed. Find yourself a great designer. Maybe a good question for you is "How can I tell if a designer 'get's it'?" Commented May 17, 2014 at 20:29
  • @JonathanTodd So how do tell if a designer gets it? Because I don't think they'll let me ask that question here... Commented May 17, 2014 at 22:47

3 Answers 3


Surveys and test marketing have been around since at least the 1930s, for good reason. Any time a company is about to spend a large amount of money on a marketing campaign, movie release or product launch, there will be surveys and testing. The bigger the investment, the bigger the risk, the greater the detail of surveys and/or extent of the tests. Poor or no surveys = poor or no results and a waste of money. There are many books on the subject, and any marketing course or text has extensive information on surveying.

Whether it's usability testing, test screenings of movies (with survey of the audience afterwards), reaction surveys, focus groups or in-person interviews, the only way to know how people will react to a given piece is to survey (if it's something new) or find out what has already been shown to work for that audience, preferably both. Sometimes you'll find the client's marketing or sales staff have a wealth of information about what works for their audience. Never ignore that; they know their public much better than you do.

Graphic design never exists in a vacuum. It always has a purpose, and most of the time that is a marketing purpose. Aesthetic appeal far from the only criterion. Those hideous-looking used car ads you still see in newspapers and online work when they're done like that. They don't work when they're aesthetically designed. (You may take a moment to recover from the horror of that before continuing. We understand.)

Before you can design something, far less decide what survey questions to ask or how to test, you have to know what the design is supposed to do. Your own evaluation of what works or doesn't is the first cut, then you go into the field and find out, using surveys or testing, how your audience reacts or responds.

If it's packaging, put it on a shelf with competing items and ask people which they would be likely to buy and why. If it's an ad, let people look at it for a few seconds, then ask what they got out of it. If you're typesetting a book, set up a half dozen variations and ask people which ones they find easier or more difficult to read. On the web, A/B testing is a standard way to test the effectiveness of an email campaign or a landing page.

Probably the only rule in all of this is: When you're testing a design, don't ask people what they "like" unless it's a purely personal item, like a birthday card or a wedding invitation. Asking what people like in a design focuses attention on the piece itself, not on what it's supposed to do, and you won't get the information you need.


Has anyone tried to create some systematic way to evaluate this?

Of course. There are lots of methods to analyse user interfaces and interaction systems.

This post might be helpful.

The thing is: graphic design does not - as Alan points out - exist in a vacuum. There is always context. Methods for user testing includes stuff like "quick and dirty methodology", you do not have to spend thousands of dollars/pounds/kroner/yen. The point as I see it is not to get your basic user testing to make all your user interface and graphic choices for you. The point is to make sure your interface does not end up an impossible maze. Run a few basic tests on your neighbour, grandmother, your drinking buddy, and a couple of others. It is likely to be helpful, if for no other reason than that you get the confirmation that you have not forgotten something essential. And you might have overlooked something.

A good thing is to take one site you like and one you think are awful, and run it through a simple heuristics test. It will make you more aware of why bad choices are bad.

Web heuristics

10 Usability Heuristics for User Interface Design, Nielsen Norman group

A Guide To Heuristic Website Reviews, Smashing Mag

Heuristic Evaluation – a Step By Step Guide Article, SItepoint

What You Really Get From a Heuristic Evaluation, Uxmag


The design of everyday things, Donald Norman

Designing Interfaces: Patterns for Effective Interaction Design

Practically anything by Edward Tufte

Interaction design, beyond human-computer interaction, Rogers

  • Sorry but I think the question wasn't clear enough, but it is more about non-interactive products rather than the usual web and mobile device interfaces. Commented May 18, 2014 at 22:19

If you have buckets of money (and time) and isolated yet related markets you can do A/B testing.

It's cheaper, quicker and ultimately more resonant and effective to pay someone millions that's brilliant at it to get it right the first time. Throw another million at them to stipulate why its right for those that need to additionally convey meaning through other contexts and endeavours.

That's how you wind up with those brand/image guideline docs... that last million.

The best at this are worth every ounce of blood they charge.

  • I'm just reading this and thinking "Just hire a few brilliant minds to lead the project, rather than spend millions on A/B testing." If the team that puts it together is truly brilliant, you can make some innovation and functional improvements with the interface, rather than spending it all (millions of dollars) on figuring out which of the colors and features and layouts appeal to people. What I mean is that, the right guys will make a great product regardless of huge testing expenses. Commented May 17, 2014 at 2:19
  • Actually, I think you're saying the same thing. The first time I read this, I interpreted the idea of paying some testing group a ton to research and test the branding, interface, design, etc. Wasn't till I saw your comment on the question that I noticed we're on the same page. Commented May 17, 2014 at 2:26
  • I probably could have written this better. Yes, we agree. I believe design is best done instinctively, by those that "get it". A world of false equalities and the prodigious propaganda of those preaching the scientific method as superior to all others has lead everyone to believe they can do everything or anything... that all they need is the right data and analysis to determine how and what is right for all, anywhere, anytime. Let me know when they've discovered how to "make" another Beethoven, Mozart, Chopin or Brahms. Or how to shorten my run on sentences.
    – Confused
    Commented May 17, 2014 at 8:07
  • I guess it still comes back to the issue of knowing who "gets it" and who doesn't. Often people who are in charge of hiring have different opinions to those that are doing the work. And even then there are people who like Apple products and those that loath it. I would like to think that there are good design principles in non-interactive design, just as there are good design principles for user interfaces so it doesn't come down to opinions and subjective views. Commented May 19, 2014 at 5:06
  • If you have to ask, you don't.
    – Confused
    Commented May 19, 2014 at 7:49

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.