Please Hold. Someone Will Be With You Shortly

Your call is important to us. Your time, apparently, is not.

You know the sound before you hear it. A pan flute, a synthesizer that peaked in 1994, and every forty seconds a calm voice cutting in to tell you that your call is important to us. If it were that important, somebody would pick it up.

“Shortly” is the biggest overstatement in the English language when you’re listening to pan flute music on loop. We all know it. The title of this article is a promise nobody believes, and yet we keep holding, because what’s the alternative?

This isn’t just a tech support problem. It’s the phone company, the hydro bill, the water department, the city tax office, the bank. Any organization big enough to have a queue has figured out how to make you sit in it. And if you run a business, you’re on both ends of the line: the customer stuck on hold, and the owner deciding how your own customers get answered.

A special kind of torture

I’ve spent a good chunk of my life on hold. Speakerphone helps, sort of. Your hands are free, but your ears are hostage to the same thirty seconds of music and a recorded voice that interrupts just often enough to keep you from concentrating on anything else.

My personal favourite: I once called an organization at 3:30 in the afternoon. I didn’t know they closed at 4. So I held. And held. At 4 o’clock, the line went dead. No “sorry, we’re closing.” No “please call back tomorrow.” Just click. Thirty minutes of my afternoon, deposited directly into the void.

It’s like they’ve optimized the process of making people wait.

The callback is a landmine

Then somebody had a bright idea: instead of making you wait on the line, we’ll call you back. Sounds civilized. The callback feature is the cruelest joke of all.

Like a lot of people, I block calls from numbers that aren’t in my contacts. That’s my fortress against spam. But the callback won’t come from a number I know. So you have to voluntarily disable your own security system. You spend all this time building a fortress against spam, and then the only way to get help is to leave the front gate wide open.

It turns your phone into a landmine for the next three hours where you’re terrified to miss a call from a random area code, only to find out it was a robocall. There’s nothing quite like seeing “Potential Spam” pop up on your screen and wondering if that’s actually the person who can finally fix your internet. And if the real callback comes while you’re on another call? Missed. Back of the line. Start again.

The irony is that every “improvement” they add seems designed to make you feel more grateful just to be ignored.

“Hello? Hello, sir?”

I work across multiple computers and multiple screens, and when something takes too long (a support queue, or even an AI that needs ten minutes to chew through a big job), I don’t sit and watch the paint dry. I go do other work. Sometimes I get so absorbed that I forget I’m on hold entirely, until a voice cuts through: “Hello? Hello, sir?” Who the… oh, right. I’m on hold.

The real problem isn’t the wait. It’s what’s waiting for you.

Here’s what nobody puts in the hold message: getting a human on the line isn’t the same as getting your problem solved.

Recently I needed help with a client’s email. I’m a customer and reseller with my hosting provider, and this client needed an unusual split-domain email setup, one that Microsoft 365 supports. I opened a chat with a human technician. The answer came back fast and confident: not possible.

That didn’t sit right. So I did something I suspect the technician never noticed. I had my own AI open beside the chat. Every time a reply came in, I pasted it into the AI. It translated the jargon I wasn’t sure about. It flagged when a reply sounded like an answer but wasn’t one. It suggested how to rephrase my question so it couldn’t be brushed off, and it told me, basically, don’t take no for an answer. Ask for someone more senior.

About half an hour later, the answer came from a senior tech: Microsoft supports the configuration. My provider’s server doesn’t.

So I still got a no. But it was the right no. Not “this is impossible,” which was false, but “we can’t do it on our setup,” which was true and which I could actually act on. That’s the difference between a dead end and a decision.

I don’t blame the front-line rep. Senior people are expensive, and you don’t put them on the front line answering calls from people whose computer isn’t plugged in. But that first layer is where the danger is. A confident wrong answer creates a false floor: you stop digging because it sounds like you’ve hit bottom. Plenty of people take that first answer as gospel and walk away.

And to be fair about the AI’s part: it isn’t an oracle either. It can be confidently wrong too. Its job in that chat wasn’t to out-guess the technician. It was to help me ask better questions and point me toward the actual documentation, so the answer rested on facts instead of two competing hunches.

Armed and ready (and a little dangerous)

It used to be a lopsided fight. The company had a terminal with your entire history on it. You had a crumpled paper statement and a vague memory of what happened in March.

AI levels that. Before you call, use it to get your story straight: what happened, what outcome you want, what you’ve already tried. Gather the account number, the reference numbers, the last invoice. Most of the questions they’ll ask are predictable. Call about your car and they’ll want the year, make, model, VIN and kilometres. Have it in front of you before the music starts.

If you’ve connected an assistant to your email or documents, it may even pull up that latest invoice while you’re still on the line. What’s possible depends on the tool and what you’ve given it access to, and no assistant can handle every company for you. But the posture is the point: you show up armed and ready and a little dangerous, with everything on your side.

One guardrail. Keep your passwords and one-time codes to yourself, and if you’re calling about someone else’s account, they still have to authorize it or name you as their representative. Being prepared isn’t the same as cutting corners.

Was it a little smug, sitting there with my secret co-pilot while the technician thought it was a fair fight? Absolutely. It’s a game, and I like playing games where the odds are stacked against me. Beating a large entity because you knew how to play is its own reward.

California puts a clock on “please hold”

Some governments are getting tired of the wait too. California’s Assembly Bill 1609 (nicknamed the Right to Human Customer Service Act) was signed in September and takes effect January 1, 2027. It covers businesses with more than $500 million in annual revenue that serve California customers. During business hours, they have to give you a simple way to ask for a human on every customer service channel they use. Then they have to make a good-faith effort to connect you within 15 minutes, or book a specific time within one business day. Once you’re with a person, they’re supposed to keep any single hold under 15 minutes and total holds and escalations under an hour. And a chatbot can’t pretend to be a person.

Let’s not oversell it. It’s a “good-faith effort,” not a guarantee. Enforcement is up to public prosecutors, with civil penalties of up to $5,000 for a first violation and $10,000 for each one after. You can’t sue over your own lost afternoon. And it’s California, not the rest of us. In Canada, there’s no law like it yet, just a House of Commons petition asking the CRTC to cap telecom holds at 10 minutes.

The bill was sponsored by a union, the Communications Workers of America, District 9. The committee analysis called it “chatbot purgatory,” and the union framed the bill as protecting customer service jobs from automation. It’s a rare win where the person fighting for their paycheck and the person screaming into a speakerphone are finally on the same side.

But notice what a clock measures: how fast someone picks up. Not whether your problem goes away. A quick handoff to someone who doesn’t know the answer is just a faster route to the wrong “no.”

If you run a business, read this part twice

If you’re putting an AI attendant on your phones or a chatbot on your website, good. Done right, it handles the easy stuff and frees your people for the hard stuff. Done wrong, it’s a very expensive wall between your customers and anyone who can help them.

So measure what matters. Not how many calls the bot “deflected,” but how many problems actually got solved. Make the path to a knowledgeable human obvious. Let your front line say “I’m not sure, let me get someone who knows” without penalty.

And remember there’s a human, or a team of humans, behind every hold maze. Somebody designed it. Somebody can redesign it.

Somebody else can stand in the rain

Here’s what I keep coming back to. When something bugs enough people and a fix shows up, people jump on it. Caller ID. Call screening. Unknown numbers straight to voicemail. Nobody needed convincing. The solution sold itself.

Waiting in line is the same. People already pay others to stand in line for them. Now picture that for customer service. You can go to Starbucks, and somebody else is waiting in line. And you’re inside in a nice cozy place. Somebody else is queued up in the rain.

That somebody is going to be your AI. Part of it is here already (I wrote about where this is heading in I’ll Get My AI to Call Your AI): an AI beside you during the call, translating, checking claims, coaching your next question. The rest, an assistant that holds your place in the queue, works the support chat, and taps you on the shoulder when a real person with a real answer shows up, is the obvious next step. Whoever packages all of that into one well-rounded service is going to do very well.

And we may not be far from making the analogy literal: a personal robot, embodied AI, standing in a ticket queue on your behalf while you get on with your day. Same idea as an assistant holding the phone line. Just add feet.

Until then, take your time back. Prepare before you call. Verify what you’re told. And insist on a competent answer, even if it isn’t the one you wanted.

Or, as Toto put it: hold the line. Love isn’t always on time.

Neither is customer service.

Book a discovery call