The Support Metrics Worth Tracking When You Run a Small Shop

A dentist I worked with kept a binder. Every morning the office manager printed a report with eleven columns of numbers, slid it across the desk, and nobody looked at it past the first cup of coffee. Calls per hour, average hold, abandonment, a satisfaction score from a survey two people answered. It was tidy. It changed nothing.

The problem was not the binder. The problem was that none of those eleven columns told her what was actually leaking money. After eighteen years of this work, I can tell you most owners track too much and act on too little. You do not need a dashboard with twenty gauges. You need four numbers, and you need to know what to do when one of them goes the wrong direction.

Answer rate, the one that bleeds quietly

Answer rate is the percentage of calls a live person (or a competent system) actually picks up. Not how fast you answer. Whether you answer at all.

This is the metric I start with every time, because it is the one owners are most blind to. Your front desk hears the phone ring. They do not hear the ring that goes to voicemail while they are checking out a patient, or the one that comes in at 6:10 after the doors lock. Those calls do not show up in anybody's memory. They just leave.

Across the shops I have worked with, missing roughly a quarter of inbound calls is normal, and the owner usually swears it is closer to five percent. Lunch hours, the second line during a rush, the entire stretch between closing and opening. A home services company I helped was losing most of its after-hours calls and could not figure out why their ad spend felt expensive. The ads worked fine. The phone did not.

To improve it, look at when you miss, not just how often:

• Pull your call log and bucket misses by hour. You will see a pattern within ten minutes.
• Treat the after-hours block as its own problem. A voicemail is not coverage. Most people calling a plumber at 8 p.m. will not leave one, they will call the next plumber.
• Stop relying on one person to be the failover for a busy line. People take breaks.

This is the gap where an AI agent earns its keep, because the calls you miss are clustered exactly where humans cannot be: simultaneously, overnight, during the lunch rush. Something that answers every line on the first ring takes your answer rate from "we get to most of them" to all of them.

Response time, and the difference between channels

Response time is how long a customer waits for a first reply. Simple idea, but the target is wildly different depending on the channel, and owners get this wrong constantly.

On the phone, the bar is seconds. A caller who hits three rings is already deciding whether to hang up. On chat, people will tolerate maybe a minute before they assume the widget is decoration. Email and web forms buy you more patience, but less than you think. I have watched leads from a contact form go cold because the shop "got to email at the end of the day." By then they had three quotes from competitors who answered in twenty minutes.

Here is the framing I give people: the clock that matters is not your internal clock, it is the customer's expectation for that channel.

To improve response time, the honest fix is coverage, not speed drills. You cannot train your way to a one-minute chat reply when one person is also running the front desk. You either staff for it or you let something handle the first response instantly and hand off the ones that need a human. An instant "yes we do that, what day works" holds a lead long enough for a person to take over.

First-call resolution, the quality metric that pays for itself

First-call resolution (FCR) is the share of issues handled in a single contact, with no callback, no transfer chain, no "let me have someone get back to you." It is the closest thing to a true quality number, and it is the one I trust more than any satisfaction survey.

Why it matters financially: every unresolved contact becomes a second contact. The customer calls back, you handle it twice, and they are annoyed both times. Low FCR is a hidden tax on your whole operation. A dental group I worked with had a "leave a message and the office will call you back" habit for anything past simple scheduling. Half those callbacks took two or three rounds of phone tag. That is real labor spent re-handling work you already touched.

To raise FCR:

• Write down the ten questions you get most. Pricing, hours, "do you take my insurance," "can you come out today." Make sure whoever (or whatever) answers can resolve every one without escalating.
• Kill the reflexive "I'll have someone call you back." It is often a habit, not a necessity.
• Track what gets transferred and why. The same three reasons usually account for most of it, and two of them are fixable with better information at the front line.

The goal is not to never escalate. Some calls genuinely need the owner. The goal is that the simple eighty percent gets fully handled the first time, so your humans spend their day on the calls that actually require a human.

Missed calls, and why it deserves its own line

You might think missed calls is just the flip side of answer rate, and the math overlaps. I keep it separate on purpose, because answer rate is a percentage and missed calls is a count with a dollar figure attached. Percentages let owners feel fine. "We answer ninety percent" sounds great until you do the arithmetic: ninety percent of two hundred calls a week still means twenty people who tried to give you money and got nothing.

Put a number on each missed call. Take your average job or ticket value, multiply by your close rate on answered calls, and that is roughly what one missed call is worth. For a home services shop where a job runs a few hundred dollars, twenty missed calls a week is not a rounding error. It is a part-time salary walking out the door, silently, every month.

What to do with missed calls specifically: set up a fast recovery for any that slip through. A text back within a minute ("sorry we missed you, how can we help") recovers a surprising share of them. Better, do not miss them in the first place. This is the entire reason businesses move to coverage that answers every channel at once, in any language a customer speaks, without a queue.

Putting it together without drowning in data

Four numbers. Answer rate tells you if you are even in the game. Response time tells you if you are fast enough to keep the lead. First-call resolution tells you if you are good enough to keep the customer. Missed calls puts a dollar figure on the gap so you actually act.

Track those weekly, on one page, and ignore the other seventeen columns until these four are healthy. If you want to see how an always-on agent moves all four at once, the way we think about it for specific trades is laid out on our industry pages, and the per-conversation pricing is on the pricing page.

The binder dentist eventually threw out nine of her eleven columns. She watched answer rate and missed calls, fixed her after-hours coverage first, and stopped wondering why the ad spend felt wasted. The reports got shorter. The schedule got fuller. That is usually how it goes once you measure the things that actually move.