A possible reason for that is the tracking code is installed in wrong location. For example - if you put code fragment in <head>, the tracker won't work. The place to put the code is in <body> tag.
We use JavaScript files to transfer voice commands to the browser. On top of that, we cache them to make sure that our libraries will be delivered very fast all around the globe. When you add a new voice command in the dashboard, we perform a CDN invalidation on all nodes in the world and it could take a little amount of time the files to be refreshed. If your browser cache on the local layer the old version, you need to "refresh" the libraries. The easy way is to go in the "Incognito" tab and check that all is working fine. It could happen only for the first couple of minutes and most of the time to the user which set-up the commands. The new user, which just gets the fresh version of the JavaScript files - can use the commands immediately. Another way to refresh is to open the tracker Javascript file and press enter.
A possible reason for that is that the domain you specified when you made the tracker is different from this, where the searches are sent to our servers. In other words - you are using our service to a domain, which is not in your dashboard. Voice tracker will show and seems to work smoothly, but some functions will not work as expected. For example, the user searches saving functionality.
Our advice is to choose the settings carefully when you create a new tracker.
If you use free trackers, change can take some time. Usually, settings files are refreshed in 12 or 24 hours. That it is something called "CDN Invalidation". We cache all the files to be delivered as fast as possible to the end customer. If you are in a hurry, you can make a new tracker with the right theme, language or message. Our advice is to choose the settings carefully when you create a new tracker.
In PRO trackers, we make a refresh every time you change an option. Every location in the world will get a new copy of your settings and instruction in a few seconds. That is an expensive operation, that is why we make this only in the PRO version.
See the answer above.
A possible reason for that is the language specified for the tracker is not correct. It is vital to choose carefully the language which your customers speak, otherwise, you can get extraordinary results.
The speech recognition API works extensively on the secure protocol. So your website must be served over HTTPS connection. It is good practice to implement a redirection (301 or 302) to the version with an SSL certificate, so http://example.com -> 301 -> https://example.com. That will guarantee that the voice tracker will work all of the time.
Unfortunately, Chrome on iOS does not support the Speech Recognition APIs toolset. Following Apple's requirements for browsers released through the App Store, the iOS version of Chrome uses the iOS WebKit – which is Apple's own mobile rendering engine, developed for Safari. That means that Chrome on an iPhone, iPad, or iPod Touch is more Safari than Chrome.