Yes, your online store should have a phone number

The claim

Publish a phone number. Not because customers will call — most of them never will — but because the number does work for you on the pages where people are deciding whether you are a real business.

Typed into a search box the question reads: should my online store have a phone number? My answer is yes, with one condition that carries the entire argument. Decide, before you publish it, what happens to every call you cannot personally take.

An unanswerable number is not a neutral mistake. It is worse than nothing. That is the honest counter-argument and I will make it properly further down, because the position only survives if it survives that.

Why a number works even when nobody dials it

A phone number is read, not used. It sits in the footer and on the contact page doing three jobs.

It signals that someone is there. A store with no address, no number and a generic contact form is indistinguishable from a store that took the money and shut down last month. Shoppers cannot audit your company. They look for the cheap signals, and a number next to a real location is the cheapest pair of them available.

It de-risks the first purchase. The second order needs no reassurance. The first one, from a brand nobody has heard of, does — especially above the price point where a customer starts imagining what happens if this goes wrong. The number is not there to be dialed. It is there to answer the question "and if there is a problem?"

It unlocks the buyers who do not email. trade, wholesale and bulk inquiries still largely arrive by phone, as do most genuinely urgent problems — a delivery at the wrong address, a gift that has to arrive before Saturday. Anyone over about sixty is far more likely to call than to open a chat widget. If any of those are your customers, a missing number is a missing channel, not a tidy design decision.

None of this requires high call volume to pay off. It requires the number to exist and to be believable.

The strongest argument against it

Here it is, stated as well as I can make it, because this is the argument that should decide the question for some readers.

A phone number is a promise with a response time attached. Email implies a day. Chat implies a few minutes. A phone number implies now. When a customer rings at two in the afternoon on a weekday and it rings out, they have not learned nothing. They have learned something specific and bad: the number on the website is decoration.

Voicemail that nobody returns is worse than ringing out, because it added hope to the disappointment. A menu with no route to a person is worse again — now you have spent their time deliberately. And an AI voice that cheerfully cannot help, with no way through to anyone, manages to be the worst of all four, because it reads as a wall you paid to build.

Against all of that, no number at all is honest. A contact form that says "we reply within one business day", kept reliably, generates less distrust than a line nobody answers. The critic is right about every part of this.

Why the position still holds

Because "answered" stopped meaning "a person is sitting next to the phone."

A number is answered if the caller always, without exception, gets one of three things: a person, or a clear and kept promise of a person, or a confident answer to the thing they rang about. All three are now reachable by a one-person store, in three different ways.

Publish narrow hours and honor them. "Phone: 9–5, Monday to Thursday" is not a weaker promise than 24/7, it is a stronger one, because you will keep it. A scoped promise kept beats an unscoped promise broken, every time.

Or pay a human answering service. They take a message and send it on. Unexciting, inexpensive, and it has worked for decades.

Or let an AI answer and escalate. Order status, shipping, stock and policy questions get resolved on the spot. Everything else becomes an accurate message with a callback window — which is exactly what a human answering service does, minus the per-call fee. What actually happens on one of those calls is worth understanding before you buy one, because the escalation path is the part that makes it work or fail.

The critic's objection is not an argument against publishing a number. It is an argument against publishing an unanswered one. Those are different claims, and only the second one is true.

What a number you can actually answer looks like

• Opening hours stated immediately next to it, in the same font size, not buried on an FAQ page.
• One hop to a person. Not a menu with four levels.
• An out-of-hours message that names a specific window: "we return calls before 11am the next working day." Then you do that.
• The number on the contact page, in the footer, and on order confirmation emails — nowhere that implies someone is listening at 3am.
• A real callback log, so a promise that is made is a promise that gets kept.

If you cannot commit to the third item, do not publish the number yet. Everything else is cosmetic by comparison.

When I would skip it

Three honest cases where the right answer is no number.

You sell digital products delivered instantly, you have no delivery anxiety to absorb, and you know you will not return a call within a day. The number buys you nothing and costs you trust the first time it rings out.

You are one person, your customers are eight time zones away, and the only honest hours you could publish are hours they are asleep. Publish a response-time promise instead.

The number would be your personal cell, and you know from experience that you will start resenting it by week three. A channel you quietly abandon is worse than one you never opened.

In all three cases, publish what you can honor: real hours, a stated reply time on email, a WhatsApp number you actually read. The underlying goal was never the phone. It was removing the suspicion that nobody is home.

Where mAutomate fits

On Pro, $149/month, the AI Call Center answers on your website. On Scale, $349/month, it answers on a real phone number — Scale is the tier with an actual line, and it is worth being clear about that rather than letting the two blur together. Both include multi-channel messaging from the plans below them, every plan has a 14-day free trial with no card required, and the full table is on pricing.

One caveat I would rather say than leave out: an AI answering your phone is only an improvement on silence if the escalation path is real. Decide which calls it is allowed to handle, decide who takes the rest, and write the never-automate list before you publish the number anywhere.

The test that settles it

Call your own store, from a phone your system has never seen, on a Sunday afternoon. Then call again on a Tuesday at 4pm and ask for a refund you are not entitled to.

Whatever happens on those two calls is what your customers get. If you are comfortable with both, publish the number. If you are not, fix the second one first — and until it is fixed, a well-kept contact form is the more honest choice.