Having spent some time with Firefox OS last week, I thought I would share some thoughts on the platform, and a few key aspects of the user experience.
FFOS is a very interesting experiment. The native vs web app (vs plain old web) debate has raged for years but has (in my opinion) been complicated by a lack of robust scenarios to evaluate. (As Jason Grigsby once said “We can’t predict future behaviour from a current experience that sucks”). Web apps don’t suck, but current ecosystems may have made it difficult to properly judge user preference. We need a properly orchestrated test case and Firefox OS might just be it.
FFOS is particularly interesting due to its high potential for growth. The platform has lots of backers, devices will be competitively priced, and will be released in parts of the world where many consumers have yet to buy their first smartphone. Each time a consumer considers one of these devices, they will (to varying, and very personal degrees) be evaluating several things: a new platform, a new and very personal piece of hardware, and a new set of mental models around the availability, discovery and use of apps.
Given the strength of its partners, i’m not too concerned the platform will suffer due to dodgy hardware or poor (and possibly performance sucking) software integration. The OS itself is fast, well designed, approachable and hackable. I think it will do very well. So that leaves the new bargain we’re making with users: here is a device that will enable you to discover fast, fun and useful web applications that you can download or used “just in time”.
Will users want this? Will they even understand the difference?
What will they miss? What will they hack?
The product’s acceptance will of course not happen in a vacuum. FFOS marketing and developer evangelism are bound to influence how people perceive these devices. The quality of the apps they come in contact with will also serve as a lens through which to judge the experience. But the web is a big place. There are currently millions more mobile optimised web sites than “mobile web apps”. I’m curious how people will feel about this mix and whether they will care (or even feel?) a lack of “appyness” in some (especially those that begin to experiment with the new Web APIs).
The fact that FFOS is open, and hackable also means that if people don’t like certain bits, they may well change them (there is good reference here from the many Android MODs, custom launchers and other variants that have sprung up around the world). If the platform proves popular we no doubt learn quite a bit from its evolution!
Now for a few thoughts about two key aspects of the FFOS user experience: Smart search and the experience browsing the web in “app” vs browser mode.
(A bit of a disclaimer. I’m well aware that i’m interacting with something embryonic. My time with the platform has taken place on a Geeksphone Peak. This is not an “official” FFOS device so there may be differences between what i’ve experienced and what others will see. I am also far too embroiled in the web and mobile industries to come at this without preconceptions. Still however, i’m struggling with certain aspects of the platform and this is my stab at explaining this disconnect).
“Smart adaptive app search”
Firefox OS devices include a smart search feature called everything.me. Marketing materials describe it as…
“…an adaptive app search that literally transforms the phone to meet your needs at any moment, without the need to download anything…start to type what you are interested in to get a whole new customized phone experience based on your needs instantly“.
The goals here seem clear enough: enable personalised discovery leading to “download” or one-time-use.
Smart search is easy to use. Swipe to the right on the device and you’re presented with a search bar and series of categories (such as Social, Shopping and Sports). You can either type keywords in the search box, or choose from the categories. Tapping a category displays a list of icons representing apps in this category. A keyword search does the same, but based on the search terms.
What’s a bit uncanny about the whole thing is that neither of these actions correspond to a search on the “real” web. Results are (so far) limited to sites that are (in some way) mobile optimised. While this kinda makes sense from a “give you the best possible first mobile web experience” point of view, you quickly end up browsing a very disjointed list. The Travel category for example shows me common stuff like Tripadvisor, then fairly quickly moves on to Michigan Travel and New Orleans Tourism (I live in Scotland).
Keyword searches seem even more constrained. The following scenario is often quoted: “enter the name of a band, and we will show you places to download their music”. This scenario works quite well. The results include apps such as Soundcloud and Last.fm and deep-links right to the artist in question (yeah indexable web!). But enter keywords that don’t immediately conjure up a transactional opportunity, or don’t map to activities you might complete using a well known service, and your options peter out quite quickly. These results may be suitable for the device but certainly don’t match what you’d find in a “normal” web search and (ironically) end up giving the impression that the web has far less to offer than it actually does.
Despite its good intentions, the whole feature ends up feeling like some sort of portal. As it stands, I fear it will confuse people and when they don’t quite find what they expect, they will switch to the browser (…more on this later).
Above and beyond delivering more relevant results, a thing that could improve smart search is an ability to customise. It’s currently easy to add categories (from a list), but you can’t remove the ones you don’t like. This quickly bloats the list and gives it that all too familiar “please enjoy apps from our select partners” feel. It’s also impossible to filter or reorder search results. This might be fine if this were a full blown app store (with screenshots, reviews and app descriptions) or search engine (with brief description and the odd photo or sub-link) but with nothing but an icon and label, the longer the list gets, the harder it is to decide what to click on.
On the one hand, I kind of understand these design decisions. Smart search helps you find apps you might like, and the platform itself delivers a low commitment opportunity to try them out. If you find something you really like, you can go a step further and save it to your home screen. So it works well in a “try before you buy” scenario but misses an opportunity to align with the full breadth of app interaction use cases.
Most studies reveal that while people install lots of apps, most only use a handful of these (currently native) apps a week. The rest tend to fall into one of two categories: something you get bored with or forget about (and therefore don’t use again), or something you do use (maybe even quite passionately), but just not all that often.
This is where the web really shines. The fact that you don’t need an app on a daily basis doesn’t make that app any less useful or meaningful to your life. It doesn’t also automatically mean you are a less valuable customer to that brand (you may interact with them in lots of other ways). But because native apps take time to download, and you can’t simply “Google them and use them” the native app model compels people to download and hold on to all sorts of apps just in case they need them again. I have many apps like this. The IKEA app (I shop there maybe twice a year), a Berlin subway guide (I visit 1-3 times a year) or Agoda which I use (about twice a year) to book hotels in Asia. I might occasionally “promote” one of these to my home screen to suit my activities (why hunt for the Berlin map while i’m actually in Berlin?) but I don’t want them there all the time.
These are however just the kind of apps that I would be happy to access, with a few keystrokes (the fewer the better) using smart search. In fact, if I use them often, I might expect these apps to just be there at the top of the list, or tucked away in a “Frequently used” category. Firefox OS is great for “one-time” and repetitive use, but i’d love to see it free us from having to save and manage stuff we value but only occasionally use.
Am I in app or browser mode? And should I care?
Congratulations: you’ve found an “app” that you like enough to save it to your home screen! App is just a convenience term here. It could be a plain old “site”, something designed to feel appy and load data asynchronously, or something in between. Any web thing will do.
Once you’ve made this decision to save it, opening it will forever launch it in app mode (my term, not theirs). In app mode, common browser controls such as back/next are tucked away in an easy to access drawer, overall features are reduced (no tabs, bookmarks etc.) and the URL bar has been permanently removed. This mode feels more “appy” as there is little visible browser chrome, and the content can be enjoyed (basically) full-screen.
The problem with this scenario may simply be one of perception, but here goes: why is there no URL bar? Sounds like a quibbling, so let me illustrate why this might matter.
So i’m in the Twitter “app” (which i’ve saved to my home screen) and click on an interesting link. I read something fascinating, decide I want to Google it but now i’m stuck. Not hugely stuck of course. I can close the app, find the Firefox browser, open it, Google whatever it was and go about my business, but why exactly do I have to do all that? Am I not already in a browser? In fact, is *the entire device* not a browser? Why do I need to jump from an (imaginary) app silo to a (equally imaginary) browser silo?
If you still think this is only a matter of semantics, let’s turn the scenario around.
I’m browsing a random site, but this time using the Firefox browser. I’m reading an article by someone unfamiliar and decide to click their profile photo. This brings me to their Facebook profile. Facebook is an “app” which I just moments ago (back on the “app” side) logged into, and saved to my home screen. Except now, i’m accessing it from the browser.
Does that browser know it’s one of my “apps”? If Firefox OS knows, surely the browser must know. Why should the experience be different? And will I have to login again (at the moment I do) or download assets again even though on the “app” side, I may have cached those same assets (…interesting question, anyone know the answer to that last one?)
So that’s my problem. Why the division? I get that hiding a URL bar is nice, but why can’t I have it back if I need it without traversing the rest of the OS to get to the browser?
And to take this line of inquiry one step further, why do I need to open the Firefox browser at all if the whole device is actually…the Firefox browser?
For the record, i’m not suggesting there is a completely easy answer to this. There are no doubt technical challenges and people may not be ready for such a switch in mental models…but wouldn’t it be neat to find out?
Maybe the conceptual app/web division is entirely based on our inability to conceive of an operating system as a browser. Many Firefox OS consumers will be buying their first smartphone. They may have some level familiarity with apps (Android’s popularity in BRIC countries, and ten years worth of J2ME-based operator and OEM portals will have seen to that) but they may ironically be relatively new to the web.
We’ve provided them with this amazing new platform that seeks to tear down the barriers between apps and the web, and the first thing we go and do is create barriers between apps and the web.
Seems like a strange thing to do…