The failure of Google Voice

I have been a Google Voice customer since its launch. Kudos to Google to bring the VoIP to everyday users. In theory, this service can mask our real phone numbers, leverage Google’s sophisticated machine learning technology to block spammer, and screening unfamiliar callers with prompts.

In practice, however, these security features can often be easily circumvented at the best, if not outright problematic. For instance, the callerID screening works well for direct calls but fails when the caller uses VoIP — common practice among scammers. It may display my Google Voice number instead! This also makes it impossible to return missed calls.

The screening prompt for unknown numbers was a nice feature. Until a recruiter half-jokingly complained about it, I disabled it to reduce friction for legitimate contacts. Ironically, while it might not deter some persistent scammers, it proved more of a hurdle for professionals to reach me.

Using Google Voice to mask my real number comes with another trade-off: outgoing calls are routed via mobile data rather than the traditional voice network. Generally, the coverage of the mobile voice is much broader than the mobile data network. In most case, this compromise is OK, — until something like today happens.

A massive power outage hit on the Seattle’s east side, triggered by bomb cyclones. Despite having four bars of signal, I couldn’t make a call because of the unreliable data network. Worse still, disabling the feature in Google Voice’s settings also required network connectivity. Trading availability for security or convenience can backfire as the system is more fragile.

I think I should stop using my Google Voice number ubiquitously, maybe only use it for untrusted parties, such as sign up for merchants’ membership. Google Voice, despite its ambition, just failed to materialize its vision it once promised.