Key takeaways
- The voicemail count is the floor of after-hours demand: hang-ups at the closed greeting, calls forwarded to attorneys' cells and carrier-blocked calls never reach it.
- Pull the phone-system report, voicemail, on-call cells, call tracking and the carrier's blocked-call list for the same weeks; a carrier that blocks on analytics must list the past 28 days of blocked calls on request[2].
- Strip spam, clients and repeat dials with written rules, then count unique new callers by time band per week.
- Convert the count to calls and minutes a month, at normal volume and in a surge month, before asking anyone for a quote.
- Count again in the first month of coverage, when every call is recorded rather than reconstructed.
Why the after-hours number is usually a guess
Ask a managing partner how many calls come in after hours and the answer is usually anchored on the voicemail box: five or six a night, more on Mondays. The voicemail count is the floor, not the number. Callers who hang up at the closed greeting leave no message. Calls forwarded to an attorney's cell end up in that phone's own call list. Calls the carrier blocked as suspected spam never reach the firm's system at all.
Buying coverage on the guess goes wrong in both directions. Too low, and the firm signs for a message-only plan that cannot handle a Saturday intake load. Too high, and it pays for full overnight intake when nearly all the demand sits between closing and 9 p.m. on weekdays. Four to eight weeks of careful counting settles it. The after-hours coverage plan covers the coverage decision itself; this guide is the measurement that should come first.
Who is calling while you are closed
National schedule data explain why the evening tail is real. In 2024โ25, 85.2% of US workers usually worked a regular daytime schedule; the rest worked evenings (5.3%), nights (3.4%) or irregular, rotating, split and other shifts[1]. In 2025, 81% of employed Americans worked on an average weekday, against 30% on an average weekend day[3]. A daytime worker calls a lawyer before 9, at lunch or after 5. A night-shift worker calls when they wake up.
None of this gives you your own numbers. Practice area matters far more than national averages, and which practice areas get the most after-hours calls explains why. The national data only say that an after-hours tail exists for nearly every firm. Your logs say how long it is, and which hours it falls in.
Collect every source, not just the phone system
A firm's after-hours calls are spread across more places than most administrators expect. Pull all of them for the same weeks, and note for each source exactly how it records a call that nobody answered.
| Source | What it shows | What it misses |
|---|---|---|
| Phone system call report | Every call that reached your system: time, caller number, duration and how it ended | May log a call sent to a night greeting or voicemail as answered; check how each outcome is coded |
| Voicemail box | Messages left, with timestamps | Everyone who hung up at the greeting |
| Attorney and on-call cells | Forwarded calls in the phone's recent-calls list | Calls that rolled to the cell's own voicemail, unless someone saves the list each week |
| Carrier blocked-call list | Calls the carrier blocked before they reached you | Owed only by carriers that block on analytics, and only for the past 28 days |
| Call-tracking numbers | Calls by ad, listing or campaign, with times | Calls to the main number that bypass tracking |
| Web forms and chat | After-hours demand that did not phone | Phone volume itself; keep it as a separate column |
The blocked-call list is the one firms rarely ask for. A carrier that blocks calls through an analytics program must, on the subscriber's request and at no charge, provide within three business days a list of calls to that number it blocked in the previous 28 days[2]. Ask at the start and again at the end of the window, so the two lists cover most of it. A run of blocked calls from local mobile numbers deserves a closer look; some of those callers may have been trying to hire you.
Set the window and the time bands
Measure for at least four weeks, and eight if the practice is seasonal. Demand for legal services moves through the year: unadjusted legal services revenue peaked at $132.1 billion in the fourth quarter of 2025 and fell 18.9% to $107.2 billion in the first quarter of 2026[4]. Calls follow their own seasons by practice area, so write down what kind of month you measured and repeat the count in a busy one before signing anything long-term.
Include a holiday weekend if you can, and log anything unusual in the window: a new ad campaign, a local news story, a day the office closed early. Then fix the bands before you count, so every week is sliced the same way.
- Weekday early: 7 a.m. to opening. People calling on the way to work.
- Weekday evening: closing to 9 p.m. Usually the largest band for consumer practices.
- Overnight: 9 p.m. to 7 a.m. Often small, but this is where arrests and safety calls sit.
- Saturday and Sunday, counted separately, because they rarely look alike.
- Office holidays, tagged so they do not distort the weekly average.
- Lunch hour, if nobody covers it. Not after hours, but those callers meet the same voicemail.
Clean the list: spam, clients and repeat dials
Raw after-hours logs are noisy. In 2025, 68% of US adults said they get scam calls at least weekly and 31% at least daily[5], and an industry estimate cited by the FCC puts the average consumer at 13 spam or fraud calls a month[6]. Business lines get their share, and after closing time spam makes up a bigger slice of the log, because ordinary business calls stop and robocalls do not.
Remove internal and outbound calls
Keep only inbound calls from outside numbers to the lines you are thinking of covering.
Strip spam with a written rule
Example rule: under 10 seconds with no voicemail from a number never seen before, or labelled spam by the carrier. Count what the rule removes.
Collapse repeat dials
Treat calls from one number within two hours as one attempt, but note the repeats: three calls in ten minutes is an urgency signal.
Match clients and contacts
Existing clients, opposing counsel, courts, vendors and staff go in their own column. They still need handling, but they are not new demand.
Sample the unknown callers
Listen to the voicemails left by unknown numbers, estimate the share that were genuine new matters, and apply it to the unknown callers who left none.
Count by band per week
Unique new callers, clients and other contacts for each band, averaged across the window.
Write the rules down before you start and keep them for the whole window. If the rules change halfway, so do the numbers, and nobody will be able to say whether demand moved or the method did. The same discipline applies to the daytime metrics in measuring answer rate and abandoned calls. And if your own callbacks show up on clients' phones as suspected spam, why law firm calls show Spam Likely covers the fix.
Worked example: four weeks of a five-lawyer firm's nights
Example: a fictional five-lawyer firm with family law and criminal defense practices closes at 5:30 on weekdays and all weekend. Its voicemail box held 74 messages over four weeks, about 18 a week, which is the figure the partners had been quoting. The full count looked different.
| Time band | Raw inbound calls | Spam removed | Clients and contacts | Repeat dials | Unique new callers |
|---|---|---|---|---|---|
| Weekday 5:30 to 9 p.m. | 118 | 22 | 34 | 10 | 52 |
| Overnight 9 p.m. to 7 a.m. | 61 | 29 | 8 | 5 | 19 |
| Saturday | 47 | 9 | 7 | 4 | 27 |
| Sunday | 39 | 8 | 5 | 3 | 23 |
| Four weeks in total | 265 | 68 | 54 | 22 | 121 |
inbound after-hours calls in the raw log
left after removing 68 spam calls
left after setting aside 54 calls from clients and contacts
unique new callers once 22 repeat dials are collapsed
likely new matters, at the 60% rate found in sampled voicemails
Example figures for a fictional firm. Use your own cleaning rules and voicemail sample.
Example: the firm had about 30 unique new callers a week after hours, and roughly 18 of them were likely new matters, compared with 18 voicemails a week from every kind of caller, clients included. The overnight band, which the partners had worried about most, carried fewer than five new callers a week. The weekday evening carried slightly more than both weekend days combined.
That shape points to a specific purchase: live answering from closing until 9 p.m. and through the weekend, with an urgent path to the on-call attorney overnight rather than full overnight intake. The criminal defense night-calls example starts the same way, with a month of the on-call log sorted into calls that could not wait and calls that could.
Turn the count into a buying spec
Providers price after-hours coverage on calls or minutes, so convert the count before you ask for quotes. Time a sample of calls first: new-matter intake runs much longer than a client leaving a message.
Example: 30 new callers a week at 7 minutes each and 14 client or contact calls at 2 minutes is 238 minutes a week, or about 1,030 minutes a month (238 ร 4.33). Ask each provider to quote that volume, the same volume plus 50% for a surge month, and the split by band.
Then match the coverage to the shape of the count, not to a package name. Demand concentrated on weekday evenings and weekends points to a closed-hours schedule; nights and weekends answering for law firms covers how that switch is set. Heavy lunch and late-afternoon misses point to daytime overflow as well, and overflow answering vs full-time coverage compares the two.
- Which bands carry most of the new callers, and which carry almost none?
- How many after-hours calls a week were urgent under the firm's own rules, and who should get them?
- Which languages did after-hours callers use?
- How many voicemails were returned next day and actually reached the caller? In Pew's 2020 survey, only 19% of adults said they generally answer calls from unknown numbers[7].
- How did the holiday week compare with ordinary weeks?
Measure again once coverage starts
The first month of coverage gives the best measurement you will ever get, because every call is answered and recorded instead of being pieced together from logs. An SI Employee answering the forwarded line leaves a recording, transcript and summary for each call in the firm's account, with urgent calls marked, so the second count needs no voicemail sampling. Compare it with the pre-purchase count band by band. If the weekday evening turns out busier than the logs suggested, the difference was hang-ups at the old greeting. You can hear a test call before scoping anything.
Keep counting each quarter after that. A call log that records the band, the caller type and the outcome turns the next renewal into a short review instead of another four-week project. If after-hours numbers keep growing, look at the daytime pattern too; law firm call volume and peak hours shows how to build the hour-by-day picture.
Frequently asked questions
How long should we measure after-hours calls?
At least four weeks, and eight for a seasonal practice, including a holiday weekend if possible. Unadjusted legal services revenue fell 18.9% from the fourth quarter of 2025 to the first quarter of 2026[4], so note what kind of month you measured.
Why not just count voicemails?
Because callers who hang up at the greeting leave nothing behind, and calls forwarded to cells or blocked by the carrier never reach the box. Count calls from the phone system and carrier records, then use the voicemails to estimate what share of unknown callers were new matters.
How do we get the list of calls our carrier blocked?
Ask the carrier in writing. A carrier that blocks calls using analytics must provide, free and within three business days, a list of calls to your number it blocked in the previous 28 days[2].
What counts as spam in the count?
Whatever your written rule says, applied the same way every week. A workable example: calls under 10 seconds with no voicemail from numbers never seen before, plus anything the carrier labels as spam. Record how many calls the rule removes so the cleaning can be checked.
Do existing clients' after-hours calls matter when buying coverage?
Yes, but count them in their own column. They usually need a message and a next-morning callback rather than intake, and they take fewer minutes, which changes the volume a provider quotes on.
References
U.S. Bureau of Labor Statistics (2026). Job Flexibilities and Work Schedules 2024-25, Table 7: Workers by shift usually worked and selected characteristics (opens in a new tab) โ
Electronic Code of Federal Regulations (Office of the Federal Register) (2026). 47 CFR 64.1200 Delivery restrictions (opens in a new tab) โ
U.S. Bureau of Labor Statistics (2025). American Time Use Survey - 2025 Results (opens in a new tab) โ
U.S. Census Bureau (2026). Quarterly Selected Services Estimates, Second Quarter 2026 (CB26-146) (opens in a new tab) โ
Pew Research Center (2025). Online Scams and Attacks in America Today (opens in a new tab) โ
Federal Communications Commission (2025). Advanced Methods to Target and Eliminate Unlawful Robocalls, Eighth Report and Order (FCC 25-15) (opens in a new tab) โ
Pew Research Center (2020). Most Americans don't answer cellphone calls from unknown numbers (opens in a new tab) โ