Shaun posted a survey about how people use the widgets screen in WordPress a few days back and over 150 wonderful people took time out of their day to answer the questions he posed. Below you’ll find a breakdown of the data that the survey yielded in a format which is (hopefully) slightly easier to digest. The aim here is to help us – and you – understand what people really think about the widgets screen.
tl;dr: people want change; drag and drop is great and awful; there’s a lot to love and even more to hate about the current setup; bacon is awesome.
It’s important to note that this is only a relatively small sample of people (yet it’s significantly better than having none at all) and it must also be pointed out that – although I can’t guarantee this – the majority of respondents are probably WordPress developers; i.e. not an average WordPress user. This will perhaps lead to a bias in the results – something we may need to explore further down the line. So, without further ado…
The first thing to note here is that in this dataset, a large majority ( 121 out of 178 or 68% ) of people use the widgets screen infrequently (once or twice a month or less). The obvious question to ask here is: why? Is it because the theme’s they use don’t have a focus on widgets? Do people set them up once and then forget about them? Does the end-user (should the survey respondent be a developer) not even know about widgets? Does the UI/UX of the admin screen make it very difficult for people to understand the widgets screen?
46 out of 178 (26%) use the widget screen at least once a week.
Could we perhaps set some goals based on this metric? Would it be great to increase weekly (or more) usage of widgets by 15%? 20%?
This set of data may already answer one of the questions I posed in the previous section; Do people set and forget? This data suggests that may well be the case with 119 out of 177 (67%) people not changing a widget any more often than once or twice a month. This could well be a useful data point in the future – are we able to increase repeat usage for widgets? Is that what people actually want to do? Do they struggle so much with their first experience with the widget screen that they don’t want to repeat the process?
This set of data is – in my opinion – the most interesting. There are two very clear areas of interest here.
The first is just how unmanageable the widgets screen becomes when there are several sidebars and multiple widgets. 78 out of 185 (42%) of people mentioned this as a problem. That’s huge. Dragging and dropping a widget from the left hand side to a widget area on the right is cumbersome. Especially when there are above 5 or so sidebars. And even more so on smaller screens. What’s even more interesting is that the many widgets = pain issue is only one of multiple user interface gripes that the vast majority of people mentioned (124 out of 185 = 67%). (It’s the one which stuck out due to how many times it was mentioned and deserved its own section on this chart).
Widgets feel like a different content paradigm than the content I can edit for a post or a page. When I train clients, it’s hard to show them the WYSIWYG inside the post editor, and then when they ask about some text inside a widget, I have to tell them “Well, you need to access this content through a different interface, and you need to know HTML to format your text in this part.”
Other items of UI that are causing problems are; Unfolding each sidebar to look for an active widget, all widget settings takes place in such a tiny proportion of the screen, lack of feedback when saving/moving widgets, No revision history/audit trail, lack of ability to duplicate widgets or blocks of widgets and no export of settings. Each and every one of these were given at least five mentions.
I wish there was more feedback around the fact that it auto-saves. I’ve had a few isolated incidents where it didn’t auto-save and so I’m always wary while working with it.
I think it’s quite clear that there’s a lot of pain points in and around the way the widgets screen works. This may go towards hinting at answering one of my previous questions – are people not using the widgets screen because of the harsh learning curve/points of friction. This data may well suggest that is the case.
They don’t go along during site export and require extra time to setup again. I do site moves for clients all the time and this is a huge waste of time.
It’s not all bad
So, then. Are we doing anything right on the widgets screen? Is there anything that people actually like using? Well, as it happens it would appear that there is. Somewhat confusingly and slightly contradictory to our previous findings, drag and drop is actually something that folk like using. 68 out of 151 respondents (45%) mentioned that it was something they considered to be positive on our screen. So, if drag and drop is both a positive and a negative then it looks like it’s something people want to use, but are frustrated with its implementation. Perhaps this is something we should consider when moving on to the next phase of development.
15 people (10%) really liked the inactive widgets feature. It also got mentioned on 4 other occasions when people were talking about the saving of widget settings. It’d be interesting to see if we can maybe use this point of pleasure and make it better – perhaps with an export of widget settings or a UI for revisions of widget settings (I’m thinking out loud, don’t quote me on this 😉 )
With all this said, 24 people – that’s 16% – took the time to explicitly say they like absolutely nothing about the widget screen.
Nothing. It’s like water – tasteless but required for survival
There was also 14 blank responses for this question too, which, if you were being pessimistic, pushes this to over 25% of people like – quite literally – nothing about the widgets admin. That’s fairly damming. (If you’re a glass half empty sorta person, that is – I like to think of it as an amazing opportunity)
OK, let’s talk improvements
People want to see change. That’s what I get from the chart above. And there are lots of ideas that people have in terms of what may work. I’ve pulled out the most popular from the changes – Live Preview, Contextual Widgets and a way to Filter or Search but in reality these are all UI ideas. So with that in mind 153 out of 179 answers revolved around making UI changes. That’s 85% of people think there’s something that can be improved upon in the user interface on the widgets screen. This is great news; it looks like we’ve definitely found something about which people feel strongly. There were lots of ideas (some of which I’ve detailed below) but the one that stands out for me is the Live Preview of widgets. If you dream-a-dream for a moment and imagine that widgets are part of the live preview/theme customizer, then it’s possible that you can tie that in with contextual widgets – people could manipulate widgets on a page-by-page (or template-by-template) basis and get a live representation of what it will look like when they press save. 72 people (40%) would like to see something like that happen. The idea certainly piques my interest. I wonder if it would pan out through the mockups, wireframes and user testing we’ll do. That’s if it makes it off the drawing board, of course.
Many people (About 8% of respondents) talked about how they’d like to see the widgets screen mimic the menu editor screen. This to me says a couple of things; 1) The current widgets screen is so far removed from other admin screens in WordPress that it causes a problem in and of itself and 2) the menu editor screen has a structure which appeals to people. I wonder if a sense of familiarity may end up actually causing confusion between the two (especially when folk see the menu widget). Certainly something to consider and perhaps see how far we get with wireframes in this direction.
The interface for widgets is too different/foreign when compared to the interface for anything else in WP. I wonder if we can make widget/sidebar management more like menu or post management.
Another item that came up in nearly 7% of answers was the ability to have user-level control to access to widgets. The idea here being that admins could decide what user level could access the settings for each widget – complete granularity in the capabilities to modify widgets. Certainly an interesting idea.
One thing that was mentioned several times, across all of the questions, was that people really want a WYSIWYG widget. There are a couple of good plugin options for this but, well, perhaps it’s something to consider for core? (My personal opinion is that it’s plugin territory, but then, I also think that many of the default widgets fall under that category too. A jetpack for widgets, perhaps?)
Not everyone wants to see change, however…
nothing. easier = less jobs for me. please do not force users to rewrite 3.8 widgets.
Here’s the raw data from the survey (with any user-specific data removed). Note that people could give multiple answers to each question (or none at all) which is why the numbers don’t quite match across each question. I’ve also tried to group the answers into something easy to scan – my categorizations may not be the same as yours.