Sources on website design stability vs. tinkering?
June 27, 2020 7:49 PM   Subscribe

Can you point me to any sources that address the value of website design stability vs. tinkering?

By "tinkering," I mean changes such as changing the menu from horizontal to vertical, or the background from red to blue ... frequent changes, for reasons that are apparent to only the person making the changes. The person making these changes does not consult with anyone else about the changes.

I do not mean a website redesign that is deliberated upon, spurred because it hasn't been updated in years, or there is a new content focus, etc.

This concerns a site for an activist group of a few hundred people.

If you know of sources that generally address stability vs. tinkering that aren't focused on websites, that could be helpful, also.
posted by NotLost to Computers & Internet (9 answers total) 1 user marked this as a favorite
 
It sounds like you're looking for some concrete resources to back you up so that you can say,
“This person's constant tinkering with our website is hurting the site's effectiveness. Here is why.”
Is that a fair interpretation of your question?
posted by D.Billy at 8:49 PM on June 27


Yes, you are right. :)

I want the issue be not just two personal opinions (although I outrank the other person within the group).
posted by NotLost at 8:53 PM on June 27


“Websites that are meant to communicate information or solicit particular actions from their visitors — such as sites for activist groups of a few hundred people — should be clear, thoughtful, and consistent in the way that they set visitors' expectations and then meet them. NotLost is right, stop f***ing around.”
– Me, a person who manages UX design projects and builds small websites

But if they won't take my word for it, here are a few (more intentionally written) things from other people who work in the UX design field:

Consistency — A Key Design Principle by Saadia Minhas
Design principle: Consistency by Anton Nikolov
Consistency: MORE than what you think by Mads Soegaard
Achieving and Balancing Consistency in User Interface Design by Michael Zuschlag

Or if a metaphor is likely to be persuasive:
Say you just checked in to a hotel. You finished talking with the desk attendant, who explained how your keycard works and where you can find the elevator and the bar and the ice machine. You go out to you car and get your bags. When you come back in, the attendant waves you over to tell you that the elevator and bar and ice machine are all in different places now. You're kind of confused about it, but okay. You forgot your phone charger in the car, so you go back out to get it. When you come back in again, the attendant waves you over again. They say that all of that stuff has moved again. And also your keycard is now a keychain fob. But it might be a regular old key when you actually get up to your room. A room which, by the way, they're thinking of moving from the third to the fifth floor at some point during your stay. But if you have any questions, just come down to the lobby again by pressing the elevator button which will be labelled either “L,” “1,” “G,” or “★” ...probably.
Would the average guest find that experience satisfying, or stressful? Would they find it easy to achieve their objectives of getting to and from their room, getting ice, etc.?
posted by D.Billy at 9:48 PM on June 27 [4 favorites]


https://metatalk.metafilter.com/tags/pony is a pretty solid data source.

sources that generally address stability vs. tinkering that aren't focused on websites

The Ubuntu window controls debacle of 2010 is the main reason I switched from Ubuntu to Debian and have been happily using that ever since.
posted by flabdablet at 5:24 AM on June 28


Tinkering vs not tinkering is a moot point. Any resources that argue in favour of one or the other is just another opinion. The real question should be “is the tinkering providing a solution to a known problem”. And the answer is highly dependent on every individual website.

The objective way to approach this would be to A/B test.

Unfortunately it doesn’t sound like your organisation warrants the resources to support such a thing. Additionally, I would argue that you don’t even have enough users (“a few hundred people”) to even get good data even if you were to A/B test.

In fact, your user group is small enough that I would say neither side could argue they are having an impact or not with the tinkering. Your data set is just too insufficient.

Now, one caveat. Are your site visitors *regular* visitors? You might be able to get some data out of that. If they are an active, engaged audience that you saw participating in regular behaviour (such as signing up for events, responding to information, interacting with key actions on the website) that you no longer see happening from the same group of people, then you should be able to tie that historical data from before and after certain tinkering from your analytics. Are sign ups down? Are event sign ups down? Are fewer people viewing important articles? Are people spending less time on your site? Are “bounce rates” up? Have “exit rates” from specific pages gone up dramatically? Etc.

Because you also work with a relatively small set of users, you could also try just asking them directly for some qualitative feedback. Send out a free survey asking if it is easier or harder to access specific types of features and pages on your website. If you have regular visitors they will have noticed and have opinions. Now, human nature being what it is, their opinions may be split 50/50. But that will still tell you something.

Background: 20 years in UX design and product management.
posted by like_neon at 6:24 AM on June 28 [3 favorites]


With all due respect, the articles that D. Billy cites are not exactly what I would describe is the consistency problem that you are facing. These articles are more focused on consistency within a system. Ie All forms work the same way. Radio buttons work like radio buttons and check boxes work like check boxes. The same words used across different pages mean the same thing. Etc.

This issue you are describing is more about how often to make changes to a design of a site. Like, I don’t think any of these articles are advocating “never change your website design.”

If you had a hot mess of a website there would be nothing wrong about tinkering with a site to achieve consistency of experience.

In fact, one could argue that “tinkering” is one way to describe the very popular and valid Agile approach to software development.

The real problem is tinkering with no purpose before and no evidence after.

And how important that is really depends on what is changing.

What problem your designer trying to solve by changing the menu pattern from horizontal to vertical? Something that is so important to the effectiveness of navigating your users should have solid justification and a plan for judging success. I mean, if you had a horizontal menu on your mobile view with small text and all cramped across a small width of scree, changing to vertical with larger menu items probably makes a lot of sense and now you are more “consistent” with other mobile websites. So all this totally depends and outsiders can’t judge if either of you are right or wrong.
posted by like_neon at 6:41 AM on June 28 [5 favorites]


Thanks for all your comments.

For a little context, this person has handled our web design from the beginning. I have seen at least four versions (probably more) of the menu in about two years, and at least two of those versions have been in the past two weeks, maybe three in the past month. And so on.

He makes small and large changes without consulting with anyone, although I have told him not to make changes without checking. I am not the only person who has raised this issue with him.
posted by NotLost at 10:46 AM on June 28


I think any designer worth their salt should be able to articulate why they are making any changes and what they are aiming to achieve with it. I think it’s a worthwhile conversation for everyone involved in the website. Then you can take the conversation to a place where you are having a shared vision of the site, and not just seeing what feels like random changes that are disconnected from the organisation’s goals.
posted by like_neon at 11:10 AM on June 28 [2 favorites]


If your web designer can't justify his decisions and/or provide a plan to get the data that backs them up, then he's just tinkering because he wants to. Honestly, the idea that one person can push out website changes without anyone else's consultation sounds like a gigantic process red flag in general for me, leaving aside the issues of consistent design.
posted by Aleyn at 3:11 PM on June 28 [2 favorites]


« Older Name this pop philosophy!   |   Easy assemble computer desk Newer »

You are not logged in, either login or create an account to post comments