How to test your own after-hours phone line

Published 2026-09-24 · EagleCallPro original research

Every restoration owner believes their after-hours line is covered. Most have never tested it. We placed real after-hours calls to 201 U.S. restoration companies, one attempt each, and 57% reached nobody. Those companies all believed they were covered too. The method below is the same one we used, stripped down so you can run it on your own business without any tools you do not already have.

What is an after-hours line test?

It is a secret shop of your own phone number. You call the number a homeowner would call, at the hour a homeowner would call it, from a number nobody on your team recognizes, and you write down what happened. That is the whole thing. It takes about four minutes per attempt and it is the cheapest diagnostic in the business.

The reason it works is that it tests the system rather than the intention. Your on-call rotation may be perfect on paper. The test tells you what a stranger actually experiences at 1:40am on a Saturday, which is the only version of the truth that produces revenue.

When should you place the test calls?

Place them in the windows where restoration emergencies actually land and where coverage actually breaks. Pick times across at least three separate nights, and make one of them a weekend. A single Tuesday-at-9pm call proves almost nothing, because that is the easiest hour to pass.

Use a phone your staff has never seen. A personal mobile belonging to a friend or family member works. If your on-call tech sees the office number or your name on the screen, you are testing their relationship with you, not your intake.

What script should the test caller use?

Use a real, plausible emergency and let the person on the other end drive. Do not volunteer information. The point is to find out which questions they ask, not to hand them a clean summary.

Then wait. Answer only what you are asked, and note whether they get to the five things that actually determine whether a truck can roll: the address, whether the water is still running, what kind of water it is, how long it has been going, and a callback number. If they ask for insurance information, note that too. If they say "someone will call you back in the morning," that is a fail even though a human answered.

Always end by identifying yourself. Tell them it was a test of the line, thank them, and do not use the call as a disciplinary event the first time. You are measuring a system. If you turn the first test into a punishment, you will never get an honest second reading. The word-for-word version of what a good intake sounds like is in the water damage intake script.

How do you score each call?

Score one of three outcomes, then a quality grade for the calls that were answered. Keep it in a spreadsheet with one row per attempt: date, local time, day of week, number dialed, outcome, seconds to answer, and whether the five intake items were captured.

Two rules keep the data honest. Let it ring for a full 30 seconds before scoring it, because a rotation that answers on ring seven still answers. And count "answered but told to call back in the morning" as a separate half-fail in your notes, because it looks like a pass in the raw count and behaves like a loss in the revenue.

How many attempts does it take to trust the result?

For a single business, six to eight attempts spread across three or four nights is enough to tell you whether you have a coverage problem. You are not running a study. You are looking for a failure, and if your coverage is genuinely broken, a failure will show up inside the first three calls.

Where you need more attempts is when you want to compare months. If you test six calls in March and six in June and get five passes then four, that difference is noise, not a trend. To claim improvement you need a consistent schedule, the same windows each time, and a larger count. That is why our own test is an ongoing sample rather than a one-off audit.

What should you test besides the main number?

The main line is the biggest hole but it is not the only one. Four more take about twenty minutes total.

• Your published emergency number, if you have one. Only 12% of the 6,241 restoration websites we have scanned publish a separate emergency number. If you are one of them, it is the number in your ads and your truck wraps, and it needs its own test.
• Your web form. Submit it as a stranger, at night, and start a timer. 4,586 of the sites we scanned have a quote or contact form. A form that emails an inbox nobody reads until Monday is a lead you paid for and threw away.
• Your Google Business Profile number. It is often an old tracking number or a line that was retired. Call it directly rather than assuming it routes to the main line.
• Your website itself. Load it on mobile data, not office wifi. 17% of the sites we scanned were down, broken or timed out entirely. Owners almost never know, because their browser has the site cached.

What do the results usually show?

Here is the full picture from our own ongoing test, so you have something to measure your result against.

Alongside it, 21% (1,332) of the 6,241 restoration websites we scanned publish an explicit speed promise on their own homepage: "24/7 live answering," "30 minute response." The gap between that number and the 43% who answered is the whole problem in one line.

What do you fix first?

Fix in the order of what the test found. If calls rang out, the problem is routing: your forwarding is broken or nobody is genuinely on call, and that is a same-week fix. If calls hit voicemail, the problem is coverage, and the honest options are a real rotation, an answering service, or an AI agent that picks up every time. We compared the three, including the salary math, in night dispatcher vs answering service, and the structural reasons rotations fail are in the on-call rotation guide.

If calls were answered but the intake was thin, you do not have a coverage problem, you have a training problem, and it is the cheapest of the three to fix.

Before you decide how much to spend on any of it, run your own numbers through the missed-call calculator. Put in your average job value, your rough monthly call volume, and the failure rate your own test just produced. The output is the annual cost of the gap, and it is usually the number that settles the argument.

Questions owners ask

How do I test my after-hours phone line without my staff knowing?

Call from a number nobody on your team has seen, such as a friend's mobile, and do not warn anyone that a test is coming. Place the calls on at least three separate nights including a weekend, and always identify yourself at the end of the call.

Is it legal to record a test call to my own business?

Recording laws vary by state, and roughly a dozen U.S. states require all parties to consent. You do not need a recording for this test: written notes are enough. If you do want to record, state that the call is being recorded at the start, or check your own state's rule first.

What counts as a failed after-hours call?

Anything other than a live person who engages with the emergency. Voicemail is a failure, a ring-out is a failure, and a human who says to call back in the morning is a failure in revenue terms even though someone picked up.

How often should a restoration company test its own line?

Monthly, with a short burst of calls across the same windows each time, plus an immediate test after any change to your rotation, your phone system, your answering service, or your call forwarding. Changes to routing are where silent failures get introduced.

What percentage of restoration companies fail an after-hours call test?

In EagleCallPro's ongoing test of 201 U.S. restoration companies, 57% of after-hours calls reached nobody: 32% went to voicemail and 25% rang out with no answer at all. Only 43% reached a live human.