Feature Request: Webhook Output for Sonos Voice Control (SVC) Transcripts / Intents

Super Idea, how do we get this upvoted to get Sonos to implement it.  

Sonos tried hosting a “vote-on-features” type site a while ago.  It was an unmitigated disaster.  One and done voters from call-to-arms enthusiast sites skewed the “yes” vote to far beyond any real market share (something we still see in these threads today), and the half dozen superfan voters who stuck around made the threads extremely toxic, driving out anybody who even slightly disagreed with their orthodoxy.  

So Sonos did the smart thing, it plowed it under and salted the earth, and now use far more reliable private surveys to determine what actual users would like to see.

Hey jgatie, 

Thanx for a snap and elaborate answer, it makes sense, even though the feature would be a ‘killer’ to get  real voice control of your house - if you use HA as well. 

Rather late I just dicovered the Sonos SVC using local processing, which I prefer. So matching the feauture with HA would be so usefull, 

Anyway - do you have any experience with type of feature request - even though I know it hard to predict 

Don’t really know if it’s possible. My guess is that because there is local processing on the device, it is fairly limited in scope, intentionally. More ‘functionality’ would likely need extra requirements of both CPU and RAM to be properly supported, not to mention the extra code necessary to control other devices.

But then, I’m no programmer, and don’t have any significant experience to rely on. 

Seems more likely to maybe be available on later devices that have (currently) spare space, but at the end of the day, Sonos is a whole home music company, and not a voice assistant company, so their expertise is likely elsewhere. 

Sonos has implemented thousands of requests over the years.  But just like any other request, especially those which require an integration with third party hardware/software, Sonos will implement it when and if there is a market ROI that makes sense for Sonos.  It was like this for the first iPhone app, the first Android app (which lagged behind the iOS app until Android achieved ~35% of the market), Spotify Connect, Atmos audio, Alexa functionality, etc., etc.  So expect HA capabilities when HA has a market penetration that can no longer be ignored.

By the way, the #1 thread on the aforementioned “Vote for Sonos features” site, for both votes and level of toxicity, was for a Windows Phone app, a mobile OS which peaked at about 2.7% of the global market.  If you looked at the thread, you would’ve thought it had the same market penetration as Facebook or Instagram.  Lucky for Sonos, they weren’t (and aren’t) fooled by easily rigged voting sites.

I understand the ROI argument and the need to prioritize based on meaningful market penetration. However, in the case of Home Assistant, the “wait until it’s big enough” logic may no longer apply.

According to publicly shared data from Home Assistant, there are more than 75,000 active installations using the official Sonos integration. That is not a marginal niche; it represents a substantial and highly engaged segment of advanced smart home users.

Additionally, considering that Sonos Voice Control already enables device control through third-party ecosystems such as Philips Hue, it is clear that:

1. Sonos has already solved the voice intent → third-party action model.
2. The architectural foundation for external integrations already exists.

From that perspective, the most complex layer — voice intent parsing within SVC — is already implemented. The remaining step for Home Assistant would not require additional processing inside the speakers themselves. Instead, once the intent is recognized by SVC, the execution layer could be delegated to Home Assistant via a secure webhook or authenticated local endpoint.

In this model, the automation logic and device orchestration would live entirely within Home Assistant, not on the speakers. Sonos would simply expose a controlled integration surface after intent resolution.

Furthermore, this flow would most likely operate within the local network. That means execution and automation processing would remain under the user’s control, reinforcing privacy — an increasingly relevant factor in ecosystem decisions.

This is not simply about forum votes; it reflects a concrete installed base that already treats Sonos as a core component of its automation stack. Providing a formal webhook interface would be a limited technical addition with meaningful strategic and privacy upside.

Certainly a good argumentation ​@Mberlin  - totally agree 

In an addition to this, the webhook model does not restrict the use purely to the HA domain only, it may even be used together with HomeKit via HomeBridge , even though Apple has it own Voice agent. 

I’m tempted to promote your  feature request on the HA Sonos Blog, but it sounds like its worthless.         

75,000 active HA users who also have Sonos? That works out to about 0.46% of Sonos households*.

I’m sorry, but 0.46% is the very definition of “marginal niche”.  I’m actually not even sure if it qualifies as a “niche”.  More like a blip, or noise.  If you are claiming 0.46% is something that should move the “wait until it’s big enough” needle, then you have a bit more to learn about market analysis.  

*The latest figures from 2024 have the number of Sonos households at 16.3 million.  

A webhook for SVC to send commands to Home Assistant is feasible and could run locally for privacy. Implementation depends on Sonos prioritizing it based on user demand.