For more than a decade, consumers have steadily disengaged from the phone call.

Robocalls, spoofed numbers, scams, and aggressive dialing have conditioned people to treat a ringing phone as an intrusion rather than an invitation. That shift isn’t merely anecdotal — it’s now one of the most well-documented behavioral trends in modern communication.

Pew Research found that 80% of Americans rarely or never answer calls from unknown numbers:

Similarly, TNS reports that 75% of consumers decline unidentified calls by default:

This is the communication environment that “agentic voice AI” tools are entering. These platforms promise next-generation outreach: automated conversations, synthetic voices, and scalable real-time calling.

But what we’ve seen across hundreds of deployments at Meera is consistent:

Voice isn’t broken. Uninvited, interruptive, synchronous-first voice is broken.

Voice still plays a critical role in sales and service — but only after trust is established and the consumer is ready.

When voice leads the engagement instead of supporting it, it falls into the same trap as robocalls: forced timing, low trust, high friction, and low answer rates.

Voice AI used improperly doesn’t solve this problem — it accelerates it.

The Problem: Voice Is Synchronous & Today’s Consumers Are Not

A live phone call demands immediate attention. Consumers must either stop what they’re doing and answer or decline entirely. There is no “respond when convenient” option.

But modern communication has shifted decisively toward asynchronous channels:

• Messaging
• Texting
• Email
• Chat
• Social DMs

Gartner reports that digital-first service tools (chat, messaging, async channels) are rapidly overtaking phone calls as the preferred customer experience channels:

Meanwhile, SMS remains one of the most effective channels precisely because it respects timing. CTIA reports 90%+ open rates, with most messages read within minutes:

And when consumers are asked directly, 69% say they prefer that an unfamiliar business text them rather than call them:

This shift aligns exactly with what we’ve observed at Meera: once communication becomes async-first, consumers engage far more willingly — because they remain in control.

Synchronous-first voice AI ignores that shift entirely.

A Human-Sounding Synthetic Voice Doesn’t Restore Trust

Voice AI companies focus on making synthetic speech more natural. But trust was never a sound problem — it was a channel problem.

Truecaller’s U.S. spam study found that Americans receive dozens of unwanted calls per month, many of them robocalls or scam attempts:

Hiya’s State of the Call report shows that one in every ten calls globally is spam or fraud, and carriers are blocking more calls than ever:

When a channel is this polluted, realism doesn’t repair credibility. A synthetic voice that sounds “almost human” still arrives at the wrong moment, from an unrecognized number, in a context where trust is already eroded.

What we’ve seen at Meera is straightforward: when a channel loses trust, adding more automation amplifies the distrust. Voice AI used as cold outbound is simply a more sophisticated version of the problem consumers have been rejecting for years.

Efficiency in Voice AI Doesn’t Matter If Nobody Answers

Voice AI platforms tout efficiency: infinite scalability, automated qualification, rapid dialing, 24/7 throughput. But these efficiency gains only matter if someone answers.

And they increasingly don’t.

Meera’s own research found that: 70% of B2C sales teams require three or more call attempts to reach just one lead.

That pattern worsens every year. Businesses are placing more calls than ever; consumers are answering fewer calls than ever.

Scaling the least answered channel doesn’t produce better engagement — it magnifies the failure mode.

When Consumers Choose the Channel, Engagement Soars

Across industries like insurance, lending, auto warranty, education, and home services, we’ve consistently observed the same dynamic:

Engagement increases dramatically when the consumer controls the pace of the conversation.

That is the fundamental advantage of SMS.

It is:

• Persistent
• Non-intrusive
• Familiar
• High-deliverability
• Context-rich
• Easy to respond to later

Consumers can read a message without committing to a real-time conversation. They can reply when they’re ready. They can re-engage days later without awkwardness or pressure.

This is why SMS-first engagement consistently outperforms voice-first models in almost every B2C vertical we serve.

Voice isn’t the issue — the timing is.

Voice works beautifully when it happens after the consumer expresses interest, not before.

Compliance Risk Increases With Unsolicited or Automated Voice

Voice — especially automated outbound voice — sits in one of the most heavily scrutinized areas of consumer communication law (TCPA).

Unsolicited automated calls raise questions around:

• Opt-in status
• Prerecorded or synthesized voice definition
• Caller identification
• Frequency limits
• Carrier blocking
• Legal exposure

Companies adopting synchronous-first voice AI face all the compliance challenges of traditional robocalling — plus the added uncertainty around how AI-generated voice content will be regulated in the coming years.

Async-first SMS, by contrast, provides clearer consent visibility, easier audit trails, and significantly lower perceived intrusiveness.

A Better Way: Async First Then to Sync ONLY When Earned

Voice still matters.

What doesn’t work is using voice as the starting point.

Our data across thousands of campaigns shows that the highest-performing approach is:

Step 1 — Begin Asynchronously (SMS)

AI engages the consumer via text:

• Gives context
• Answers questions
• Builds trust
• Qualifies intent
• Respects timing
  
  

This removes the pressure and skepticism associated with unsolicited calls.

Step 2 — Move to Voice Only When the Consumer Opts In

Once a consumer expresses readiness, Meera instantly transitions into:

• Warm call transfers to real humans
• Scheduled appointments
• High-intent conversations
• Dramatically higher close rates

Voice becomes the synchronous layer on top of an asynchronous, consumer-led conversation — not the opening gambit.

This is the proper role of voice in modern outreach.

It’s about more than just voice AI 

The voice channel is not dead.

But cold outbound voice — human or AI — is fundamentally misaligned with modern consumer behavior.

Consumers trust messages.

They distrust unexpected calls.

They respond to async communication.

They ignore synchronous interruptions.

And when companies use voice at the right moment — after context, after trust, after consent — engagement and conversion increase significantly.

Voice is powerful when it’s earned, not imposed.

Meera’s async-to-sync model isn’t an attack on voice. It’s the blueprint for using voice the way consumers actually want:

Start async.

Earn trust.

Then — and only then — move to voice.

It’s not just more effective.

It’s more respectful.

And that’s why it wins.