What Actually Happens When Someone Asks Their Smart Speaker to Find a Plumber

Say a pipe bursts on a Sunday morning and a homeowner shouts at a smart speaker on the kitchen counter: 'find me a plumber open now near me.' Within a second or two, the assistant reads back a single business name, maybe a phone number, and offers to call it. There was no scrolling, no list of ten results to compare, no map to squint at. One answer got chosen and read aloud. Understanding what makes an assistant pick that one answer, out of every plumber in the area, is the entire game for a business that wants to be the one it picks.

Voice answers are a single-result format, not a search results page

This is the central difference between a voice query and a typed search. A typed local search returns a page of results and a map with several pins, and the person does the comparing themselves. A spoken query to an assistant almost always returns one answer, occasionally two or three read aloud in sequence. There's no browsing involved — the assistant has already done the comparing on the user's behalf. That means the assistant's underlying ranking and eligibility criteria matter enormously more for voice than for typed search, because there's no second-place finish that still gets seen.

Where the answer actually comes from

Voice assistants don't independently research businesses. They pull from structured local business data sources — most centrally, business listing profiles like Google Business Profile, along with the same signals that inform local map rankings: proximity to the person asking, category match to what was asked, hours of operation relative to when the question was asked, and review signals. A business with an unclaimed, sparse, or inaccurate listing simply isn't eligible to be the one answer read aloud, regardless of how good the business actually is, because the assistant has nothing reliable to pull from.

• An accurate, complete, and claimed local business profile is close to a hard prerequisite for being surfaced at all
• Hours of operation need to be genuinely current — a listing showing incorrect hours can make a business appear closed at exactly the moment someone is asking
• Category selection matters more than it seems — a business mis-categorized in a listing platform may simply never surface for the query that matches what it actually does
• Proximity is weighted heavily for 'near me'-style spoken queries, more so than for some typed searches where a person is willing to travel farther

Why review volume and recency matter more for voice

Because the assistant is picking one answer rather than presenting a list, review signals function less like general reputation evidence and more like an eligibility filter. A business with very few reviews, or with reviews that stopped coming in months or years ago, reads to these systems as a business with less confidence behind it — even if the service quality is excellent. Consistently gathering fresh reviews isn't just about persuading a human reader who compares options; for voice, it can be the difference between being selectable at all and not.

In a results page, being third place still gets seen. In a spoken answer, third place is invisible.

What businesses can actually control here

1. Claim and fully complete the business's Google Business Profile — every field, not just the minimum required ones
2. Verify hours of operation are accurate, including holiday and seasonal hours, and update them proactively rather than after a customer complaint
3. Confirm the primary and secondary categories genuinely match the services offered, checking against how a customer would actually phrase a request
4. Keep the phone number and address perfectly consistent across every listing platform the business appears on — inconsistency undermines the confidence these systems place in the data
5. Build a steady, ongoing stream of reviews rather than a one-time push, since recency appears to matter as much as total volume

The phone call is often the actual conversion point

Many voice assistant local queries end with an offer to place a call directly, skipping a website visit entirely. This means the business's phone system — who answers, how fast, and whether the call gets picked up outside normal staffing hours — becomes part of the same conversion chain that starts with a spoken search. A business that wins the voice answer but doesn't answer the resulting call has effectively won a coin flip and then thrown it away.

This isn't a separate discipline from local SEO — it's the same data, held to a higher bar

There's no special 'voice optimization' toolkit distinct from doing local business data correctly. The businesses that show up reliably in voice answers are, almost without exception, the ones with clean, complete, actively maintained local listings and a genuine, ongoing review pipeline — the same fundamentals that support local map rankings generally, just with much less margin for error, because voice hands out one answer instead of ten. A free audit checks exactly this kind of listing and data consistency across a business's online presence.

← Back to all articles