Profile cover photo
Profile photo
John Shanks
John's posts

Post has shared content
I was writing a blog post about three recent changes to Google's UX that made me feel that Google has changed how they think about design, specifically toward simplifying the UX for current users at the expense of the new user learning curve, when I decided maybe I was being arrogant, seeing as how I didn't do the user research Google did, so instead I'm going to talk about them here in hopes that in this Google-rich population I might learn rather than preach.

First up is the 'new tab' chicklet in the Chrome tab bar. Somewhere around Chrome 16 the '+' disappeared from it, leaving a little ghost of a button that, to my mind, wouldn't be recognized for what it was by a user who hadn't already formed their mental model on earlier versions of Chrome.

IE 8 and 9 do it this way, with a small ghosted (but full-connected) tab, while Firefox explicitly retains the '+'. Do users who are coming from a less-than-modern browser understand that the chicklet is a baby tab waiting to be given form?

Second is the Gmail conversation view. Gmail isn't my baby anymore, and I don't pipe up about design decisions that are different than how I would have designed it, because Gmail's design team's goal isn't to do what I would do. However, the new conversation view is unfortunate. One of the main goals of the 'card stack' design was to give a visual metaphor of a 'new-stuff-first' list, even when it's actually in chronological order. The 'stacking' of read cards tested extremely well and people understood not only where one reply stopped and the next began, but how to expand cards to re-read earlier parts of the conversation.

The new design flattens conversations completely, turning them into just a list of boxes, some of which are grey (which means closed) and some of which are white. The reply box at the end is no longer tightly coupled to the actual email you're replying to, which is a problem if the most recent reply was to a subset of the original recipients.

For the experienced user this doesn't represent much of a problem. Their mental model of conversations and collapsion was formed in the more explicit UI, and they understand the underlying meaning even when the cues are removed. My guess is that the user who never sees anything but the new UI gets a very different picture. All the problems we had when testing a flat 'expand contract' UI should crop back up if the new UI is tested on folks who have never used Gmail before.

I understand the design goal of a cleaner, sparser UI where any gradient or visual complexity should be cut but I can't help but feel that, as with the Chrome 'new tab' chicklet, a little bit of baby got thrown out with the bathwater.

Lastly, the new Gmail and Google+ 'clicking on the logo does nothing' behavior seems just absurd. Nothing this significant could have actually gotten pushed out without a huge internal conversation about it, with one side saying it's stupid and the other side thinking about how they just didn't understand the bigger picture and would get used to it.

Speaking as a practitioner who's been on both sides of that field and been right and wrong on both, this design decision is definitely one of those ones where the best of design intentions is flat out wrong. As long as there is a property logo on your page, clicking on that logo should take you to the top level of that property, and if you're already on the top level and it's a dynamic site, clicking on it again should perform the same action as clicking a refresh button on the same page.

This isn't a Google convention that will be acclimated to if changed. It's an Internet convention that predates Google's existence by a good many years. It's like if Audi started shipping all 2012 vehicles with gearshifts on the driver's left, no matter which side of the road folks drive on in your country, because it creates a more consistant experience across Audi cars or supports a future Audi strategy.

Chrome's new-tab button losing the plus? Doesn't feel right to me, but data could easily prove me wrong.

Gmail's new flat conversation view? I'm pretty sure it's not as friendly for new users struggling with threaded conversations for the first time, but maybe I'm an old fogey.

Clicks on logos no longer taking you to the top page of that site and/or refreshing content? That's just batshit crazy.

And don't even get me started on the eldritch logic that dictates that clicking on the 'News' link from a search result page doesn't carry over the search term, while clicking on 'Images', 'Maps', or 'YouTube' does.

(Update: So much for trying to make the post lower-key by writing it here instead of my blog)

I worry this place won't be as cool when its not just me and Robert Scoble.

Post has shared content
For the noobs. Feel free to share at will.

Created by +Simon Laustsen. Props.

Post has shared content
Best overview I've seen.
Additive Thoughts on Google Plus

I do love me some new social software.

Google +1 -- Rolled out before Google Plus, initially within search results and then as a share button that helps both personalize and socialize (your friend's breadcrumbs factor in), +1 now becomes a sharing action people can understand. But Google needs to do a better job in sharing case studies for how implementing +1 benefits publishers who adopt it. Especially because the glut of sharing buttons is now comparable to the days of the RSS coffee cup. Right now you +1 in public and they appear as a minor post (buried) on your profile, but this could evolve as they way to share into circles more prominently.

Profiles -- Nice and clean. Has a purpose now with Plus. My favorite feature is being able to view your profile as others would see it (a specific person, anyone on the web, etc.) which helps people understand what is actually a complex privacy model.

Circles -- this is the core of Google Plus.

Discovering people and adding them to your circles (lists of people you share to) is a great experience. You might recall during the early days of Orkut there was a fast and visceral UI to collect faces. Since the number of faces you collected was a prominent number, many played it as a competitive game (by now we just let +Robert Scoble go there for us) that went too far. They've actually made list forming fast and fun. And consistently let you add someone to a circle anywhere you see their face across Google Plus. But there are some issues:
* This isn't Ridiculously Easy Group Forming. You aren't forming groups, you are forming lists you share to. A group and its membership are a shared entity. I'm sure they made this conscious tradeoff to reduce noise and keep things simpler. But it is at the cost of not leveraging Reed's Law. You can't understand people and things through the groups they are a part of. You can't discover people and things through groups. You can't form a group and take action with people. Instead you make lists of people you share to in different ways, but those lists are only yours (and google's for mining).
* It doesn't leverage existing groups to help form lists. For example, I want to create Circles for SlideShare and Socialtext people. The people search doesn't leverage profile data yet, or email domains. Every group entity across Google's properties should be explored as a source for a person to create a Circle.
* One usability snafu. I used select all an then held down the shift key to remove people before dragging it to my Friends circle. I didn't realize I had selected all past the first screen so added 500 people by accident with no way to undo, so I had to delete the Circle (including the <100 that I had more carefully chosen) and start over.

Stream and Sparks -- pretty solid and simple.
* People over algorithms. If I was to give the Google one piece of advice its to beware of creating algorithms that people don't understand. I think they have largely done this, but please remember to save the magic for things I couldn't possibly understand.
* Its not clear if the activity stream is the firehose, or if it is filtered for me in some way. And if its filtered, how do I remove that filter? What about simple filters like by type (link, video, location), time, activity? This is important because Circles provides filter-out, but there will be a strong demand for filter-in.
* The stream functions almost exactly like Socialtext Signals (plus location sharing, minus shared group constructs), right down to wikitext, so you know I love it.
* More permalinks, for sub-items.
* I tried editing a post that was initially Limited to some Circles and couldn't figure out how to make it public. I could however share it with somebody named public.
* Sparks are promising, a good way to pull in the core of Google search. While I wouldn't want an unread count, I'd like a hint if there is something new while I'm engaged in Plus. And a blended feed.

Hangouts -- this is exactly what my teenager wants to hang out with her closer friends, beyond chatting with friends on Facebook. The shared experience of watching a Youtube video (a la +Om Malik Alive Web,, etc) is just right for her. However, Facebook will indeed launch a direct competitor to this in partnership with Skype.

Photos -- Nice UI, but I'm not a Picassa user and most of the pictures at the moment are people's profiles. Patents probably prevent them from doing people tagging, but I don't see the value add here yet.

Huddle -- the group messaging app isn't available within the web UI, or the mobile web UI for the iPhone. They have a rich mobile app for Android and its on teh way for iPhone.

Mobile -- Otherwise the mobile web app is pretty good. Its pretty clear that Plus will be the driver for overhauling their mobile app efforts.

Some closing thoughts:
* While the above is my reaction to a feature playground, I can say that Google has made a major step forward into social. Even though the more constrained privacy model doesn't hold back rollout for sake of noise, the public noise about Plus will drive them to accelerate rollout soon (also, there is an empty Family circle).
* Profiles, activity streams and social objects are the trifecta for social sharing. But the really interesting stuff starts to happen when they integrate their more collaborative properties. Let alone bring Plus into their one social property, YouTube, which is one hell of a property.
* I haven't heard anything yet of if this platform will be open for third party developers
* Its not clear to me that Plus will be able to hang on to its early adopters, and is mostly getting feedback from them right now. Some of us want to move over to something new, especially from Facebook. Right now there are good conversations limited to some of the same people that were the first on services like Twitter, so the conversations just feel good. But I don't see how a persistent group identity can form here, or how the utility is sustainably differentiated. My guess is that the Circles model will retain certain conversation leaders in the same way LinkedIn does for its professional demographic. But further integration will be required for true differentiated utility.
* That said, I believe that Google has a full commitment behind this, is iterating quickly upon a good start. Google will be a player in social and thats largely a good thing.

* Maybe they won't solve the filter-in problem by exposing human-understandable facets. But Google sure could have a big win by having an activity stream of updates and Sparks that actually has memory. And Search.
* Having trouble in the Circles UI, it won't let me drag a person that is already in one Circle into a second one.
* I'm just now realizing how this is a Reverse Asymmetric Follow model, or Asymmetric Sharing model for ties.
Wait while more posts are being loaded