Sonos CEO admits to insufficient app testing : “We released it too soon”

Sonos extends hardware warranties by a year as act of contrition.

See full article...

 

[Iphtashu Fitz]

While I appreciate the Mea culpas from the CEO, what I'd really like to see are some timeframes as to when we can actually start to see improvements rolling out. We have a couple of the Ikea branded sonos speakers and they used to work relatively nicely with Home Assistant, but all that's gone haywire since this fiasco. I'd like some idea as to when I might be able to try getting them to play nicely together once again. The incremental improvements to the app don't seem to have changed the HA issues at all.

 

[E.Murphy]

My understanding was that the issues were not insufficient testing, but rather the execs not listening to the developers telling them that the app was not ready.

 

[ThatEffer]

The CEO reportedly admitted to the botched rollout stemming from a lack of proper testing and a desire to push out a lot of features simultaneously

Oh, they pushed 'em out alright.

 

[healthcamp]

Compared to the Sonos fiasco, Ars’ redesign has been a huge success.

 

[Fatesrider]

Sounds to me like they have a lot of trust to build back up.

Once bitten, twice shy, is a apt metaphor here. Sonos marketed (and priced) itself as "the premier" thing to have for the best audio quality. Then they fumble the ball in their own end-zone, and now have to punt and hope they don't fuck up at all, ever again, going forward.

Cutting exec bonuses is a clever move. But my trust and faith in companies these days tells me I'd have to see that on the financials in writing before I'd even consider that to be true. I don't know how involved investors are in Sonos, but since the head guy is STILL THERE after this, I'd say there's more smoke and mirrors here to appease pissed customers than to placate potentially petulant investors.

After all, these days, the investors come first.

 

[Mental Gear Reduction]

The biggest reason I avoided Sonos, when I was originally considering them, is because they were proprietary. I opted, instead, to assemble my own system the traditional way, with speakers, a receiver, and (at the time) a Squeezebox 2. That was a fantastic setup, and even when Logitech discontinued the whole Squeezebox line, the server software stayed available. (It's still being worked on today, although with its lack of hardware players, it's not very attractive anymore.)

I'm still using the same speakers, with a different receiver and subwoofer, and I now drive them with my PC instead of a dedicated player. Buying standard components has let me replace each piece when I was ready. I can still sub out any piece, twenty years later. Everything is interoperable, nobody can lock me into anything. Even all my archived music is in FLAC format, which works almost anywhere, and can be easily converted to any lossy format as needed.

I can substitute anything, and nobody can decide to wreck my system for their benefit. It's not turnkey like a Sonos, but is superior in pretty much every other way. It just took some effort. I had to plan the system out before buying, I couldn't just show up with a credit card.

I don't remember the detailed cost analysis anymore, but I think it was also a lot cheaper than a Sonos.

 

[OrangeCream]

While I appreciate the Mea culpas from the CEO, what I'd really like to see are some timeframes as to when we can actually start to see improvements rolling out.

You realize they don't know either right?

We have a couple of the Ikea branded sonos speakers and they used to work relatively nicely with Home Assistant, but all that's gone haywire since this fiasco. I'd like some idea as to when I might be able to try getting them to play nicely together once again.

They do too. If they knew in the first place they would have also known about this entire fiasco up front.

 

[adespoton]

I do wish this perspective was more common in publicly traded companies: when the employees are all doing what they can to get the product shipped, including voicing concerns to management, when management makes the wrong call, the response to investors should ALWAYS first be "this was a business decision, and it cost us X amount, and it will be coming out of these management line items."

Once exec, and then management, bonuses are eaten up completely, then come the employee bonuses. Once those are gone, we start trimming other executive perks. Once those are gone, we trim management perks. Once those are gone, we trim employee perks.

Only after THIS point is it OK to move on to things like "which contracts do we cut?" And by contracts, I first mean external, and then employee.

 

[citizencoyote]

Forgoing their bonuses, the horror! Those poor executives, surely they will work extra hard next time!

Meanwhile the rank and file already laid off as a result of this go back to job hunting, certainly mollified by the knowledge that the people who fired them won't get extra money for doing so.

 

[h2500]

A company is candid about having screwed up and makes customers whole?

Let’s for a moment recognize the rarity of such an occurrence.

If it was google, they would simple have discontinued the speakers and spun it into a success story.

 

[50me12]

Remember folks:

https://www.joelonsoftware.com/2000/04/06/things-you-should-never-do-part-i/

 

[50me12]

A company is candid about having screwed up and makes customers whole?

Let’s for a moment recognize the rarity of such an occurrence.

If it was google, they would simple have discontinued the speakers and spun it into a success story.

It's not like they could do anything BUT be candid about it.

It was that total a failure.

I know folks who couldn't reliably change the volume on their devices since MAY...

And in May they weren't the least bit candid about it.

 

[OrangeCream]

A company is candid about having screwed up and makes customers whole?

They haven’t made them whole. They’ve just promised to extend the warranty and are still trying to make them whole.

Let’s for a moment recognize the rarity of such an occurrence.

Sure, let’s recognize that Sono has admitted to screwing up and are trying to fix the problem they created. That’s just how business works. If you don’t fix your problem you lose customers.

If it was google, they would simple have discontinued the speakers and spun it into a success story.

That’s only because Google’s ongoing revenue stream is independent of their product customers.

 

[LuDux]

Technically there was insufficient testing.

If they measured everything broken then they would have known it was too broken to release. By not measuring certain aspects of the new architecture they could naively release it without realizing it was broken.

In other words, QA's job here is to stop a release. If you don't want to stop a release, you don't QA the release.

QA doesn't have that power.

If QA tells the execs 'stop the release, it's broken' - and the execs say 'fuck you loser I don't care release it anyway' - who do you think wins?

 

[s73v3r]

Sounds to me like they have a lot of trust to build back up.

Once bitten, twice shy, is a apt metaphor here. Sonos marketed (and priced) itself as "the premier" thing to have for the best audio quality. Then they fumble the ball in their own end-zone, and now have to punt and hope they don't fuck up at all, ever again, going forward.

Cutting exec bonuses is a clever move. But my trust and faith in companies these days tells me I'd have to see that on the financials in writing before I'd even consider that to be true. I don't know how involved investors are in Sonos, but since the head guy is STILL THERE after this, I'd say there's more smoke and mirrors here to appease pissed customers than to placate potentially petulant investors.

After all, these days, the investors come first.

Why do these execs still have jobs? If you or I fucked up a tenth of how badly these people fucked up, we'd have been fired long ago.

 

[50me12]

QA doesn't have that power.

If QA tells the execs 'stop the release, it's broken' - and the execs say 'fuck you loser I don't care release it anyway' - who do you think wins?

Yeah reports had the developers screaming and yelling to not do the release, they didn't listen to them.

QA (if they have any) wasn't going to be listened to either.

 

[OrangeCream]

QA doesn't have that power.

If QA tells the execs 'stop the release, it's broken' - and the execs say 'fuck you loser I don't care release it anyway' - who do you think wins?

You’re right. It’s not QA’s fault the product got shipped, they’re the ones finding how broken the product is.

I guess it sounds like I’m faulting QA here, but as you say they don’t have the veto power. QA did as much as they could but it was someone else with the power to push the release. It was that person who decided enough QA had been done to release the product.

 

[metamatt]

I'm one of those longterm Sonos customers who is very passionate about this. Sonos was a great (if expensive) system for years, and even through the S1/S2 migration where they botched the messaging and alienated a lot of people, it was still possible to get good performance out of either an S1 or S2 system. Up until this year -- it's not been at all possible to get it to work right since this "version 80" app release in May.

Before S2... and even to a large degree even after S2... Sonos was my go-to example of a company that knows how to do software updates right; products would continue getting updates that actually made them work better for years and years (much more so than other companies). With v80... not so much.

It's been interesting to me watching the corporate messaging on this botched v80 release shift from "this is great" to "it has problems but this is fine" to "this is what COURAGE looks like" to "ok, we have some egg on our face but we'll fix it soon" to "we are starting to admit/realize how bad this really is".

Thanks to Scharon and Ars for continuing to keep focus on this... a bunch of us longtime customers have gone through the stages of grief on this, maybe sticking with anger or maybe moving past that, but it's a pretty important story and lesson for the tech industry.

 

[healthcamp]

The biggest reason I avoided Sonos, when I was originally considering them, is because they were proprietary.

Same here. Sonos was early to the market with a set of very cool concepts, but the lock-in killed it for me. It’s cool that you found an affordable way to implement your favorite parts using commodity components.

For me the most interesting part was taking a compact speaker with high-excursion drivers, sticking in an active DSP, and letting it tune itself to its acoustic environment. Sure, it will never sound as good as good hifi equipment, but the flexibility is priceless. I’m perennially disappointed that you can’t buy a system like that without some kind of proprietary lock-in.

So, I recently built something similar using a Raspberry Pi, HifiBerry, and cheap IC amps. Even with so many premade components it was still a pretty involved process, but the results and worth it. Of course, back when Sonos launched, a project like that was completely out of the reach of an ordinary home hobbyist.

 

[OrangeCream]

Before S2... and even to a large degree even after S2... Sonos was my go-to example of a company that knows how to do software updates right; products would continue getting updates that actually made them work better for years and years (much more so than other companies). With v80... not so much.

Huh. I would have thought the whole S2 debacle would have been proof that they don’t get SW updates right:
https://www.digitaltrends.com/home-theater/before-upgrading-to-sonos-s2-read-this-separate-systems/At the time, Sonoswas encouraging its customers to trade up to newer devices by bricking the old products with a special recycling mode, in return for a 30% discount on a new Sonos product.

It sounds like they never intended to support S1, after S2, but customer backlash demanded they back down. Their initial plan was essentially the same as their current debacle:
The easiest thing to do, it reasoned, was to get its customers onto newer devices so that all Sonos products in a single home could run on the same system.

Instead today they had the problem that a key feature was integrating legacy products with Ace Headphones, which means they had to make both systems use the new architecture. The S1/S2 split was not in the cards this time:
In response, Sonos came up with a new option: Keep your legacy products if you want, but you will need to stick with its older S1 software. And if you own newer devices that are S2-compatible, these will also need to stay on S1 if you want them to be part of the same, easy-to-use single system.

Maybe they could have kept both systems operational but then the user wouldn’t be able to forward sound from a soundbar to the headphones:
https://www.digitaltrends.com/home-...-your-sonos-ace-headphones-to-sonos-soundbar/About as literal as it sounds, TV Audio Swap lets you listen to whatever audio is being routed through a Sonos soundbar, be it the Sonos Arc, Sonos Beam (Gen 1 and 2), or Sonos Ray, with your Sonos Ace headphones.

In hindsight they should have made TV Audio Swap an optional feature and also kept their old system concurrent.

 

[starglider]

Given how awful this was for customers locked in to Sonos, I can't imagine ever trusting the company again. Maybe if they axed the entire executive team and brought in new management. Even then, I'd want to see something genuinely different. Maybe a commitment to open-sourcing a portion of the system, especially for EOL hardware? Speakers last for decades and are pretty much on the flat part of the innovation asymptote; there should never be a reason to have to buy a new one except a desire to upgrade or some electrical fault.

That said, I hate this app-based lockin bullshit anyway; I'm the guy with a Raspberry Pi Zero 2 connected to a DAC plugged into a receiver from circa 2000, so granted I probably wasn't the target customer in the first place.

 

[Dreadalus]

Forgoing their bonuses, the horror! Those poor executives, surely they will work extra hard next time!

Hey, that $72,000 isn't nothing! But I'm sure the $4.5 million in stock compensation from last year makes up for it some (in fact the stock price is pretty even with one year ago despite rising and falling in the interim period).

 

[dlux]

Reuters reported today that Spence and seven other execs "would forgo their bonus in the most recent fiscal year"

They will be reduced to a diet of avocado toast and Juicero salvage lots for a full week. I don't know if they'll survive!

 

[uncaughtexception]

$72k bonus for a CEO in 2023? That's about L7 level bonus in a FAANG. It's a nice virtue signal, but I doubt it affected exec TC significantly. Just means their package doesn't have a large bonus component.

 

[50me12]

I worked at a company where the executives where denied bonuses for a given amount of time.

As soon as they were allowed again ... they just paid the past bonuses that quarter...

 

[BlandMushroom]

Technically there was insufficient testing.

If they measured everything broken then they would have known it was too broken to release. By not measuring certain aspects of the new architecture they could naively release it without realizing it was broken.

In other words, QA's job here is to stop a release. If you don't want to stop a release, you don't QA the release.

It wasn't just insufficient testing. There is a whole re-arhitecture and I'm sure there are features that are marked as "don't worry about it, we'll do it in phase 2" on the testing plan. Classic "deprecate the old thing but replacement is not ready yet" scenario.

 

[Readercathead]

I do wish this perspective was more common in publicly traded companies: when the employees are all doing what they can to get the product shipped, including voicing concerns to management, when management makes the wrong call, the response to investors should ALWAYS first be "this was a business decision, and it cost us X amount, and it will be coming out of these management line items."

Once exec, and then management, bonuses are eaten up completely, then come the employee bonuses. Once those are gone, we start trimming other executive perks. Once those are gone, we trim management perks. Once those are gone, we trim employee perks.

Only after THIS point is it OK to move on to things like "which contracts do we cut?" And by contracts, I first mean external, and then employee.

The executives in this case are only giving up a tiny pittance of their total compensation. It’s a drop in the bucket compared to all the salaries and health care for the many people they fired, or the total stock value lost or the lost revenue… certainly not enough to make up for future lost revenue from all the customers who now hate them. The whole executive team needs to be fired. If you think that’s harsh think about the many blameless employees they threw away almost immediately after this app rolled out. Punish the “deciders”.

One guy commented on his from-scratch stereo solution, that seems like a lot of work but not compared to spending hours per item trouble-shooting Sonos speakers that used to work or schlepping them to the router every week to get them back on the internet (why my Sonos speakers all got permanently unplugged years ago.)

 

[agt499]

The biggest reason I avoided Sonos, when I was originally considering them, is because they were proprietary. I opted, instead, to assemble my own system the traditional way, with speakers, a receiver, and (at the time) a Squeezebox 2. That was a fantastic setup, and even when Logitech discontinued the whole Squeezebox line, the server software stayed available. (It's still being worked on today, although with its lack of hardware players, it's not very attractive anymore.)

I'm still using the same speakers, with a different receiver and subwoofer, and I now drive them with my PC instead of a dedicated player. Buying standard components has let me replace each piece when I was ready. I can still sub out any piece, twenty years later. Everything is interoperable, nobody can lock me into anything. Even all my archived music is in FLAC format, which works almost anywhere, and can be easily converted to any lossy format as needed.

I can substitute anything, and nobody can decide to wreck my system for their benefit. It's not turnkey like a Sonos, but is superior in pretty much every other way. It just took some effort. I had to plan the system out before buying, I couldn't just show up with a credit card.

I don't remember the detailed cost analysis anymore, but I think it was also a lot cheaper than a Sonos.

Everything you said.

I'm a Squeezebox user for all the same reasons - the community which is now Lyrion - is amazing.

A note on hardware players that the WIIM products seem to be an excellent option at present, supporting Squeezebox along with a pile of other protocols (like Spotify connect and Tidal Connect along with Chromecasting audio, Airplay and bluetooth) at very good value prices.

A point to note is that the Server software has been open source for a very long time, which along with some talented and committed contributors has kept it stable and adding features for years.
I'll call out Michael Herger specifically, who seems to be a total superhero - seemingly from (past/present?) Logitech, who is omniscient.

 

[bzz-erk]

My understanding was that the issues were not insufficient testing, but rather the execs not listening to the developers telling them that the app was not ready.

Reminds me of a failry typical scenario. Having lived this time and time again, leadership above the dev team, having zero interest in the dev process pressures dev for a wildly optimistic delivery date then obstinately refuses to heed missed milestones. The dev factory abuse continues until they are silenced through implied threats until everyone has sufficiently buried their head in the sand up until the deployment date... It's a common organizational dysfunction, I think.

 

[Lonyo]

Hey, that $72,000 isn't nothing! But I'm sure the $4.5 million in stock compensation from last year makes up for it some (in fact the stock price is pretty even with one year ago despite rising and falling in the interim period).

I was wondering how an exec of a global company only had a 72k bonus.

The massive stock comp makes it make a lot more sense and adds context

 

[Drizzt321]

The biggest reason I avoided Sonos, when I was originally considering them, is because they were proprietary. I opted, instead, to assemble my own system the traditional way, with speakers, a receiver, and (at the time) a Squeezebox 2. That was a fantastic setup, and even when Logitech discontinued the whole Squeezebox line, the server software stayed available. (It's still being worked on today, although with its lack of hardware players, it's not very attractive anymore.)

I'm still using the same speakers, with a different receiver and subwoofer, and I now drive them with my PC instead of a dedicated player. Buying standard components has let me replace each piece when I was ready. I can still sub out any piece, twenty years later. Everything is interoperable, nobody can lock me into anything. Even all my archived music is in FLAC format, which works almost anywhere, and can be easily converted to any lossy format as needed.

I can substitute anything, and nobody can decide to wreck my system for their benefit. It's not turnkey like a Sonos, but is superior in pretty much every other way. It just took some effort. I had to plan the system out before buying, I couldn't just show up with a credit card.

I don't remember the detailed cost analysis anymore, but I think it was also a lot cheaper than a Sonos.

You should check out Snapcast. Has both server and client, and does some timing sync for the clients. I run server on my NAS, multiple RPi clients with Amp hats for speakers in different rooms

 

[darwinosx]

While I appreciate the Mea culpas from the CEO, what I'd really like to see are some timeframes as to when we can actually start to see improvements rolling out. We have a couple of the Ikea branded sonos speakers and they used to work relatively nicely with Home Assistant, but all that's gone haywire since this fiasco. I'd like some idea as to when I might be able to try getting them to play nicely together once again. The incremental improvements to the app don't seem to have changed the HA issues at all.

Why would you appreciate anything from Spence. He is the person responsible for this fiasco. It's amazing that he still has a job. It must be a pretty captive board that he hasn't been fired yet. This is far from the only debacle he has overseen and all resulted in layoffs of Sonos employees but never him.

 

[darwinosx]

You should check out Snapcast. Has both server and client, and does some timing sync for the clients. I run server on my NAS, multiple RPi clients with Amp hats for speakers in different rooms

Get back to us when it does home theater.

 

[woo90]

100% accurate. They have a very robust public beta testing program. I have been a member for several years.

The problem is an arrogant CEO who thinks he knows better than clients and forced a load of crap upon them, then tried to tell them (effectively) to stop whining, he knows better.

Sonos has never been for the Apple HomePod or Amazon Alexa crowd, or folks looking for a cheap home theatre set up. Talk about a CEO who clearly didn’t (and still doesn’t) understand his clients.

Despite their timeline, I am currently beta testing the iOS app and it’s still a flaming dumpster fire. My older Play 1 speakers simply won’t play content. At all. If I want to change SiriusXM channels, I have to use my bedroom Amazon Echo device and have Alexa change channels. The only way to get the app working is to (constantly) power cycle the speaker. And that doesn’t consistently make it work.

 

[darwinosx]

Everything you said.

I'm a Squeezebox user for all the same reasons - the community which is now Lyrion - is amazing.

A note on hardware players that the WIIM products seem to be an excellent option at present, supporting Squeezebox along with a pile of other protocols (like Spotify connect and Tidal Connect along with Chromecasting audio, Airplay and bluetooth) at very good value prices.

A point to note is that the Server software has been open source for a very long time, which along with some talented and committed contributors has kept it stable and adding features for years.
I'll call out Michael Herger specifically, who seems to be a total superhero - seemingly from (past/present?) Logitech, who is omniscient.

That's nice. What about home theater?

 

[darwinosx]

It wasn't just insufficient testing. There is a whole re-arhitecture and I'm sure there are features that are marked as "don't worry about it, we'll do it in phase 2" on the testing plan. Classic "deprecate the old thing but replacement is not ready yet" scenario.

A lot of stuff running through the cloud now that doesn't need to. Creating lag and other issues. I would bet it's so they can harvest users data and sell it. Apparently the price for Sonos speakers isn't enough for them. Ever read their terms of service? Oh boy...

 

[1349]

My understanding was that the issues were not insufficient testing, but rather the execs not listening to the developers telling them that the app was not ready.

this is what needs to be said out loud

"i didn't listen and i fucked everything up because of this"

 

[agt499]

That's nice. What about home theater?

You do multiroom home theater?

I do theater in one room at a timebonly, and Squeezebox outputs music to the AV receiver as a DLNA renderer.
Otherwise the theater side is Blu-ray or streaming via CCWGTV.