Key takeaways
- Define the clock for each channel: what starts it, what stops it and whether nights and weekends count.
- Report the median, the slow tail and the share reached within target, not just an average.
- Build the measure from system timestamps, never from times staff type in afterwards.
- Audit with test inquiries; in a 2024 secret-shopper study only 33% of law firms replied to emails at all[1].
Why most response-time numbers are wrong
Ask a firm how fast it responds to new leads and you will usually hear a confident estimate, often "within the hour". Ask for the data behind it and one of three problems tends to appear. The clock starts at the wrong moment, such as when someone noticed the voicemail rather than when it was left. The stop event is an automated message rather than a person. Or the figure is an average distorted by a handful of very slow cases, which hides both how fast the typical response is and how bad the worst ones are.
Accurate measurement starts with definitions, then moves to instrumentation, then to reporting that leads to action. None of it needs special software, but all of it needs consistency.
Define the clock for each channel
| Channel | Clock starts | Clock stops | Common mistake |
|---|---|---|---|
| Phone, answered | The call connects to your system | A person or approved intake answers | Measuring from when staff noticed the ring |
| Phone, missed | Time of the missed call or voicemail | The first two-way conversation | Stopping the clock at an unanswered callback |
| Web form | Submission time | The first personal reply by phone, text or email | Counting the autoresponder |
| Text | Inbound message time | The first reply that addresses the message | Counting an automatic acknowledgment |
| Time received at the firm's address | The first personal reply | Measuring from when someone opened it | |
| Chat | Chat started | A person joins or the approved assistant answers the question | Counting a greeting bubble |
The stop event matters most. A response is the first time the prospect gets something useful from the firm: a real answer, a booked time, or a conversation. Acknowledgments and attempted callbacks are worth tracking separately, but they are not responses.
Business hours or clock hours?
Decide whether evenings and weekends count, then report both versions. Business-hours measurement is fairer to staff; clock-hours measurement is what the prospect experiences. People contact firms outside the working day all the time. In 2024-25, 15% of US wage and salary workers usually worked an evening, night or other non-daytime schedule, and 57% had flexible schedules that let them vary when they started and finished work[2].
Example: a form submitted at seven on Friday evening and answered at nine on Monday morning took more than sixty clock hours but almost no business hours. Both numbers are true. The business-hours figure tells you whether staff are keeping up; the clock-hours figure tells you whether you need after-hours coverage.
Instrument it step by step
Use system timestamps
Pull times from the phone system, form tool, texting platform and email, not from what staff type into notes later.
Join records to one lead
Match by phone number and email, so a caller who also submitted a form is one lead with one first-response time.
Mark the stop event
Have staff log the first two-way contact, or derive it from call logs, such as an answered outbound call above a minimum length.
Store both measures
Record clock minutes and business minutes for every lead.
Exclude non-leads
Remove spam, vendors and existing clients before reporting.
Automate the report
Produce it weekly, by channel, source and handler.
Joining records is the step most firms skip, and it changes the result. A prospect who called, got voicemail, then filled in a form ten minutes later should have one clock starting at the call, not a fresh one starting at the form.
Report medians and targets, not averages
Averages are distorted by outliers. One lead answered after four days can make a team that usually responds in minutes look slow, and a few instant answers can make a team that usually takes hours look fine. Report three numbers instead: the median, the 90th percentile, and the share of leads reached within your target.
Example: a fictional firm's 50 form leads in one week have a median first response of 18 minutes, a 90th-percentile response of six hours, and an average of just over four hours. The average suggests a slow team. The median and percentile show a fast team with a weekend gap, which needs a different fix.
- Median: the typical prospect's experience.
- 90th percentile: how bad the slow cases get.
- Share within target: for example, the share reached within fifteen minutes during business hours.
Match the channel prospects use
Speed matters less if you respond on a channel the prospect ignores. In Pew's research, 97% of smartphone owners used text messaging during the study period, and 93% of owners aged 18 to 29 used voice or video calls[3]. Many people will read a text before they answer a call from an unknown number; 80% of Americans say they generally do not answer such calls[4]. Record which channel produced the first two-way contact, not only when it happened.
If most first contacts after a missed call happen by text, it may be worth sending a short text straight after the callback attempt, with consent, rather than relying on voicemail alone.
Audit with test inquiries
System data shows what was logged. Test inquiries show what a prospect actually experiences. Professional secret shoppers found in 2024 that only 40% of US law firms answered phone calls and 33% replied to emails[1]. Run your own checks every month.
- Place two test calls at different times, one of them outside office hours.
- Submit one web form and send one text from a number the firm does not know.
- Record the exact time of the first real contact and compare it with what the system logged.
- Note what the prospect heard or read: was it useful, or just an acknowledgment?
- Share the results with the team without blame, then fix the process.
Setting a target
Set targets by channel and by time of day, starting from your current median rather than an ideal. If the median web-form response is three hours, a first target of one hour is more useful than an aspirational five minutes nobody believes in. Phone calls answered live need a target in seconds or rings; missed calls and forms need a target in minutes during business hours and a separate one outside them. Publish the targets to the team, show progress weekly, and tighten each one only after it has been met for a month.
Data quality checks
- Spot-check ten leads a month by hand against the raw system logs.
- Look for leads with no stop time at all; they are often the slowest cases, hidden.
- Check that clocks on different systems agree, especially after daylight saving changes.
- Confirm that existing clients and spam were removed before the report was run.
- Watch for staff marking a first contact before it happened, which shows up as impossibly fast times.
Who handles the lead matters
Break response time down by handler as well as channel. Front-desk staff, intake specialists, an answering service and any automated system will each have their own pattern. A team median can look healthy while one handler, or one time slot, is consistently slow. Seeing the split lets you fix schedules or handoffs rather than asking everyone to try harder.
Look at response time by source too. Leads from a new campaign often arrive in bursts that overwhelm normal staffing, and the response time for those leads can be far worse than the firm's usual figure, just when the marketing spend is highest.
Use the number well
Keep the report short: one line per channel with median, 90th percentile, share within target and lead count, for business hours and clock hours, compared with last week. Anything longer tends to go unread.
Response time belongs next to booking rate in weekly intake reviews. If response times improve and booking rates do not, look at what is said in the first contact. If both improve, the fix is working. If response times get worse in particular hours, adjust staffing or coverage for those hours.
Connect speed to money with the cost of a missed call, compare channels with cost per lead versus cost per signed case, and set expectations for written replies with email response standards for attorneys. For the billing measures that sit alongside intake in a managing partner's review, see utilization, realization and collection rates.
Frequently asked questions
When does the response-time clock start for a missed call?
At the time the call was missed or the voicemail was left. It stops at the first two-way conversation, not at an unanswered callback attempt.
Do autoresponders count as a response?
No. Track them separately. The clock stops when a person or your approved intake actually engages with the inquiry.
Should we report average response time?
Report the median, the 90th percentile and the share within target. Averages are distorted by a few slow cases.
How often should we audit response times?
Monthly, with test calls, forms and texts at different times, including outside office hours.
References
Illinois Supreme Court Commission on Professionalism (2Civility) (2024). Phone Calls Ignored, Emails Unanswered: Fixing the First Impression Problem for Law Firms (opens in a new tab) โ
U.S. Bureau of Labor Statistics (2025). Job Flexibilities and Work Schedules (2024-25 data) (opens in a new tab) โ
Pew Research Center (2015). U.S. Smartphone Use in 2015 (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