Android upgrades have been a source of frustration for years now — but it's rare to get a glimpse at just how extensively the various manufacturers' performance has been degrading over the past few years.
I've been tracking Android upgrade progress closely since the platform's earliest days. Each year, I publish the results in an Android Upgrade Report Card that measures how long different device-makers take to get the most recent major Android OS update onto their current and previous-gen flagship phones (using the first availability of the software in the U.S. as a metric).
With Oreo, to put it mildly, it wasn't a pretty picture. It was by far the worst overall result I've seen in all the years I've been measuring this stuff, with just a single passing grade followed by one D-level score and then a bunch of F's. Two of those F's were actually zero percent, too, due to a complete failure by the manufacturers to start any level of rollouts within the first six months of Oreo's release.
Un-frickin'-believable.
Android upgrades: A zoomed-out perspective
Given all of that, I thought it'd be interesting to zoom out to a bigger picture view — to pull together data from the past four years of Android OS updates and see how the major manufacturers' performance has evolved from the time of Lollipop, in 2015, to Oreo, now.
Lollipop, you may remember, marked the first time Google provided an early preview of its software — months ahead of the public release. The idea was that with more advance notice, manufacturers would be able to get a head start on prepping their devices and then be ready to roll out upgrades faster than they'd managed in the past. Google pushed the preview up earlier and earlier with each subsequent year, giving companies more and more time to work with the software.
And yet — well, let's turn to some telling visuals:
JR Raphael
(Click image to enlarge)
This chart represents each major manufacturer's Android Upgrade Report Card score from Lollipop through Oreo. You can see the exact formula for the score's calculation here, but in short, 60% of it is based on how long it took for an update to reach a company's then-current flagship, 30% is based on its time to the company's previous-gen flagship, and 10% is based on the company's overall communication with customers throughout the process.
Google, as you can see, has remained relatively consistent. Its scores stay steadily in A territory, with two 95's, a 93, and a 94.
HTC was on a good streak from Lollipop to Marshmallow, meanwhile — with respectable scores of 85% and 86%, respectively. Then, with Nougat, things started heading south, with a drop down to 77%. That trajectory picked up pace considerably with Oreo, which saw HTC earning a mere 49% grade.
LG's baseline has always been lower, but it actually made a teensy bit of positive progress from Lollipop to Marshmallow — moving from a 68 to a 71. After that, however, the downward slope hit hard. The company earned an underwhelming 47% for its Nougat performance and a big fat zero for its embarrassing failure to do anything within Oreo's first six months of existence.
Motorola's been on a steady decline from Lollipop on — not entirely surprising, perhaps, considering that Lollipop was the start of the company's Lenovo-owned era and the end of its Google-owned glory days.
And as for Samsung — well, it's always done pretty poorly with upgrades, but it's slumped further down the path from "bad" to "worse" with every passing year. (And yes, that almost-missing final bar in the green spot represents a zero, just like it did with LG.)
Upgrade delivery times for flagship phones
Those are composite scores, so let's break things down even further and look at the actual number of days it took each company to get an Android release onto its U.S. flagship — and remember, this is measuring only the first appearance of the software on a U.S.-available device, so it doesn't even take into account the variance and additional delays we frequently see across multiple carriers and models.
JR Raphael
(Click image to enlarge)
This should be pretty obvious, but it seems worth emphasizing: Here — unlike in our first chart — the lower the number (and the shorter the bar), the better.
You can see that the same general progression more or less holds true in this domain: Google holds steady with very fast and reliable updates to its flagship devices, while HTC shows delivery times slowly but steadily growing worse.
LG had that one year of mild improvement, then worsened a bit the following year and completely dropped the ball this latest go-round. Motorola actually did okay with Lollipop (on its current-gen flagship, at least — you'll see the other part of the story in a moment) but then shot up considerably with its rollout time the following year and continued to get slower with its rollouts on each subsequent year.
And Samsung started out underwhelming and got meaningfully worse, without exception, every single year.
But again: My scores deliberately factor in both current-gen and previous-gen flagships — because manufacturers should be providing timely and reliable support to their top-tier devices for a minimum of two years. And the data for those previous-gen phones is perhaps the most telling part of all:
JR Raphael
(Click image to enlarge)
Basically, if you buy a phone from LG, Motorola, or Samsung, you really shouldn't hold your breath for an upgrade in your device's second year. LG completely failed its customers with its previous-gen U.S. flagship update for Nougat and has yet to deliver Oreo to its previous-gen device as of today, more than six months after the software's release.
Motorola took 433 days to get Nougat onto its previous-gen flagship, meanwhile — a device that was sold only unlocked, at that! — and is also still currently a "TBD" for Oreo, half a year into that update's life.
And Samsung's steady tick upward stays true in this realm, too, with another "TBD" in place for its previous-gen Galaxy S7 flagship as of this writing.
One thing you can say for the major manufacturers is that, with a couple of glaring exceptions — LG, as noted above, and also Motorola, which abandoned some of its U.S. flagship customers a few years back — most of them do provide updates to their recent flagship devices eventually. But waiting in the dark for months upon months, not knowing if or when increasingly stale-growing software will reach you and receiving little to no communication along the way, isn't exactly an ideal experience.
So where do we go from here?
In discussing Android upgrades over the past several months, one thing I hear come up a lot is Project Treble — Google's plan, as announced last May, to create a "modular base" for Android that should make it easier (at least in theory) for manufacturers to update their devices in a timely manner.
Treble is something that'd apply mostly to devices released after that announcement — so not most of the ones included in the Oreo analysis, in other words, but quite possibly some of the ones we'll look at with the upcoming Android P release. Early reports indicate the newly announced Galaxy S9 will support Treble, and other 2018 flagships seem likely to do so as well.
We have to keep in mind, though, that Treble itself is not an end-all solution. Sure, it's another tool that'll help make the process of updating devices easier for Android manufacturers — but we've seen such efforts before, and yet device-makers' performance has generally only continued to get worse. (See the discussion about the "early preview" program, above.)
The real question in my mind revolves around motivation — and whether, even with more tools at their disposal, manufacturers will have the incentive to make OS upgrades a priority. Even if Treble makes them easier, after all, it doesn't make them automatic. What it basically does is eliminate the need to update the "lower-level" parts of the code, namely the areas related to the silicon inside a device, with every new release. In real-world terms, that means a company like Samsung won't have to wait on Qualcomm to do its share of the updating every time a new Android version comes out.
But Samsung will still have to do its own share of the work — and that includes updating Android to incorporate all of its user-facing interface changes and feature additions, which are factors that Treble does not, by all counts, address. The level of effort and resources involved is certainly less substantial than it was before, but it's still not negligible.
And the cold, hard reality is that despite the fact that they're investing time and money in managing software upgrades, most Android manufacturers don't make any additional revenue directly off of those efforts. In fact, if you think about it, providing fast and frequent OS updates to existing devices actively works against most device-makers' financial interests — as if anything, it makes you less likely to feel the need to upgrade your hardware and plunk down cash for a new phone. Google is the sole exception to this rule, and it's probably no coincidence that it's also the sole company that takes Android upgrades seriously.
There is actually an answer already in place for all of this, of course, but it's not the answer many people are hoping to see (nor the answer that'll cause these grades to go up). So for now, we'll have to just wait and see if this latest effort to give Android manufacturers a kick in the caboose actually does any good.
I'd like to end on a positive note and say at least things can only go uphill from here — but, well, I've thought that before, and look where we are now.
Sign up for JR's new weekly newsletter to get this column along with bonus tips, personal recommendations, and other exclusive extras delivered to your inbox.
[Android Intelligence videos at Computerworld]
