How can I do voiceover testing at scale?

Voiceover as a screen reader for both Mac and iOS, requires different approaches. 

Here are some basic guidelines: 

• At scale means different things to different people, thus, there may not be a one-size-fits-all model. Generally, think about the QA process and where accessibility fits in the QA process. 
• Evaluate the QA process, come up with an accessibility testing procedure, and make voiceover a part of that testing.  
• You can’t test everything, there isn't really a QA process which includes testing every single component on every single page or every app to the most rigorous degree. Test for blockers, barriers, anything that's going to cause the user to potentially fail or not be able to complete a process. Test across different assistive technologies on different devices. 
• If thinking about an iOS app uniquely or specifically, incorporate testing from the voiceover perspective. Also think about different iPhone devices, as well as iPads. Test for any types of issues that might arise across different versions. 
• To test accessibility on a mobile app versus a web page on a Mac OS device, expand beyond testing simply with voiceover and incorporate testing with PC based screen readers as well. Usually, it's not sufficient just to identify a single assistive technology and say, “we'll do testing on voiceover, but we won't really do anything else beyond that”. Testing should include screen reader testing, ideally by a person or by people who are end users of that assistive technology.