Phone tag is a part-time job

Add it up sometime. Not the calls you miss — the calls you catch, but only on the second or third try, after voicemail, after a callback that went to their voicemail, after a text to confirm a time that got confirmed six hours later than you needed it. That whole loop has a name in every trade, and it's phone tag, and for a one-truck plumbing shop it is not a minor annoyance. It is a second job you never applied for, running in the gaps of your first one.

I want to actually tally it, the way you'd tally billable hours, because "phone tag is annoying" doesn't move anybody and "phone tag costs you six hours a week you could've billed at $125/hour" does.

The double-cost that hides in "at least they left a voicemail"

Here's the trap in how most plumbers think about this. A missed call feels like it has two outcomes: bad (no voicemail, they call someone else) or fine (they left a voicemail, you'll get to it). The math above — the missed-call math — already covers the bad outcome. This post is about why the "fine" outcome isn't actually fine.

A voicemail is not an answered call. It's a task. Here's what "fine" actually requires: you listen to it (2 minutes, usually at a red light or between jobs), you decide if it's real work or a wrong number, you call back (attempt 1 — they don't pick up, they're at their job too), you try again later (attempt 2), you finally connect, you get the address and the actual problem, you check your schedule on your phone with grease on your hands, you propose a time, they need to check with their spouse, they call back an hour later to confirm. That's not a two-minute voicemail. That's forty-five minutes to an hour, chopped into five separate interruptions spread across a day that started at 6:30am and doesn't have room for them.

Do that for even a third of your missed calls in a week — say 6 voicemails out of 60 calls, a mix of estimate requests and small jobs — and you've burned three to five hours of scattered admin time that never shows up on an invoice. It just shows up as you, standing in a driveway in Arlington at 8:15pm, texting a customer back about a Tuesday appointment.

Tally it for your week

Try this with your actual numbers, not mine. For one average week:

• Voicemails left: count them.
• Minutes to listen + triage each one: call it 3-5 minutes.
• Callback attempts per voicemail that connects: usually 2, sometimes 3.
• Time per callback attempt, including redialing and waiting: 5-10 minutes.
• Scheduling back-and-forth per booked job: another 10-15 minutes, easily, once you count the "let me check with my wife" round trip.

Even a conservative shop doing 6-8 real voicemails a week is looking at 3-5 hours of pure phone admin — attempted at all hours, because customers call back when it's convenient for them, not when you're free. That's not "wrench time." That's time you're not under a sink, not driving to the next job, not home. It's the hours squeezed in around the real work, and it's worth naming because it's real cost even though no invoice captures it.

Now put a number on what an hour of your time is actually worth. If your average billed rate is anywhere near $100-150/hour, three to five hours of phone tag a week is $300-750 of your own labor spent on logistics instead of pipe — on top of, not instead of, whatever jobs you lost outright to calls that never got answered at all.

The fixes, cheapest to most complete

Do nothing. Free, and the 3-5 hours a week keep happening, quietly, forever. This is where most one-truck shops sit — not because it's a good option, but because nobody put a number on it before now.

Batch your callbacks. Free, and it genuinely helps a little: instead of returning calls whenever you notice the red bubble, set two blocks a day (say, lunch and end-of-day) and knock out voicemails in a row. It cuts down on the scattered-interruption tax but does nothing for the calls that never left a voicemail in the first place, and it doesn't stop the phone-tag part — you're still playing tag, just on a schedule.

A human answering service. Real cost, usually $150-400/mo in this market, often metered per call or per minute past a plan cap. This removes the "did they even leave a voicemail" problem, since a person answers live — but it doesn't remove the phone-tag problem entirely, because the answering service still has to hand you a message, and you still have to call the customer back to actually schedule anything, unless you're paying for a tier that books directly into your calendar.

Part-time help to manage the phone and schedule. This can genuinely close the loop — a real person booking real appointments — but for a one-truck operation the math is usually upside down: you're paying for coverage across hours where the phone might ring twice, to get help during the two hours a week it rings constantly.

An AI receptionist that books the job, not just takes a message. This is the category Dialkeep is in, so read this one with the appropriate grain of salt. The specific thing this fixes, versus every option above it: the call ends booked. No message waiting for you, no callback required, no phone tag, because there's no gap between "customer describes the problem" and "it's on your calendar and texted to you." At $199/mo flat, it's priced below the human answering service and it removes the admin loop those services leave in place. It's not free, and it's not a person — but the thing it's actually competing against isn't "a human," it's the 3-5 hours of tag you're already paying in your own time.

The honest bottom line

Batching your callbacks is free and worth doing regardless of what else you choose — it's a genuine five-minute fix. Beyond that, every option is a real tradeoff between what it costs in dollars and what it costs in your own hours, and the second number is easy to undercount because it never shows up on a bill. If you want the demo instead of the pitch, the number's live at the top of the plumbing page — call it and see whether it actually closes the loop or just takes another message.