Sentence never understood when speaking

I’ve followed the guide to add custom sentences to trigger an automation, in this case changing the station on my Sonos speakers. Everything works when the different sentences are typed, but when spoken, I always get a ‘Sorry, I didn’t understand’. Pipeline is the standard HA Cloud, speaker used the Atom Echo. Other custom sentences work as intended. Can anyone shed light on what is wrong?

The automation yaml:

alias: Zet Q op
description: ""
triggers:
  - trigger: conversation
    command:
      - Qmusic
      - Zet qmusic op
      - Speel qmusic
conditions: []
actions:
  - action: media_player.play_media
    target:
      entity_id: media_player.woonkamer
    data:
      media_content_id: FV:2/20
      media_content_type: favorite_item_id
    metadata:
      title: Q Music
      thumbnail: >-
        https://sali.sonos.radio/image?w=60&image=https%3A%2F%2Fcdn-radiotime-logos.tunein.com%2Fs2398g.png&partnerId=tunein
      media_class: genre
      children_media_class: null
      navigateIds:
        - {}
        - media_content_type: favorites
          media_content_id: ""
        - media_content_type: favorites_folder
          media_content_id: object.item.audioItem.audioBroadcast
mode: single

Try q music instead

This give you s bit more flexability

  • command:
    - (schakel|switch|speel) [over|over naar|naar] (qmusic|q music) [af]