What missed calls cost a law firm: the arithmetic behind the number

Search for what missed calls cost a law firm and you will be given a number. It will be large, it will be annual, and it will look like it came from research.

43 claim records·11 sections·every figure attributed to a dated capture·no first-party testing

It came from a multiplication. Each page we captured that shows how its figure was reached uses the same three terms: a count of missed calls, times a conversion rate, times an average case value. The formula is not the problem. The problem is that the inputs are not sourced, the publishers disagree about all of them, and the one input your firm could actually look up is the one the most detailed of those pages tells you to assume.

This page shows that arithmetic, follows the figure underneath it to the end of its trail, sets out what has genuinely been measured about law firms and the phone, and finishes with the four numbers a firm can count for itself.

The number is a multiplication, and here are the multiplications

Two of the pages competing for this search print their inputs. Their arithmetic is easy to follow, which is to their credit and is why it can be examined at all.

One, published by a company that sells law firms an answering service, takes 10 qualified missed calls a week, a 25% conversion rate and a $3,000 average case value, and arrives at $390,000 a year in lost revenue, alongside a table that runs the same formula across five input sets and reaches $1,300,000 at the top, describing the 25% rate as conservative.

The other, also published by a seller of attorney call answering, takes 5 missed calls a week, a 20% conversion rate and a $3,000 case value, calls these realistic numbers, and arrives at $12,000 a month or $144,000 a year.

The largest figure in this territory is built the same way. A legal AI company estimates 557 million calls a year to US law firms, applies about 35% unanswered, a 7% conversion rate and an $8,000 average client value, and publishes $109 billion. The same company runs a near-identical chain on the product site it sells, from the same 557 million estimate and the same 7% rate, sources none of its inputs, and presents the result as arithmetic rather than as an estimate.

Four pages, three publishers, one formula, and not one of the conversion rates or case values is sourced on any of them. Every one of those publishers sells the remedy for the problem its own figure describes.

What the assumptions do to the answer

Hold the missed-call count fixed at 5 a week, which is the count two of these publishers use. Then apply only the input pairs a single publisher actually put together, because mixing one publisher's conversion rate with another's case value would be inventing a method nobody published.

Same firm. Same five missed calls a week. A factor of about 4.5 between the lowest and the highest, and every bit of that spread comes from two inputs none of the publishers sources. The bottom two rows are that publisher's own printed figures, which is the check that the formula has been read correctly.

One page does not even agree with itself. Five calls a week at 20% and $3,000 is one signed client a week, and over a fifty-two-week year that is $156,000. The printed annual figure is $144,000, reached by taking the monthly figure and multiplying by twelve, which treats a month as four weeks. Nothing turns on which convention is right. It is worth seeing because it shows how soft the number is inside a single page that calls its own inputs realistic.

The one input you could actually know

Of the three terms, exactly one is obtainable. Your phone system knows how many calls came in and how many were answered. Nothing in the evidence we hold tells you what share of your missed calls were prospective clients, what share of those you would have accepted, or what the average matter arriving by phone is worth to your firm.

So look at what the most detailed of these pages tells the reader to do. Its four-step method opens by saying to pull call logs for the past 30 days, and then: if unavailable, assume 35% of business-hours volume and 60% after hours.

Follow that as written and all three inputs are assumptions. The number at the end describes no particular firm, including yours.

The figure they share, and where it goes

That 35% turns up everywhere. Three of the eight ranked pages we captured carry it, and they credit three different sources.

The first credits Clio, in a one-word parenthesis, with no link and no report named. Clio's own captured page carries the figure and sources it to a Law Technology Today URL described in Clio's link title as the ABA's benchmark study on a firm's intake process. That URL now resolves to americanbar.org and returns a not-found page saying the site was recently redesigned.

The second links a Digital Journal press release, and converts the percentage into 195 million calls across the US. That URL returns HTTP 410 Gone.

The third introduces the figure as industry data and links it to a page on the hostname legal.fullfunnel.co. We followed it. Two attempts, one over https and one over http, both returned a DNS resolution failure. The parent domain resolves and serves a live site describing the company as a go-to-market and revenue operations services firm for business to business organizations; the captured homepage contains no occurrence of law, legal or attorney.

Three attributions. Following all of them took four fetches and not one arrives at a method.

That is not the same as saying nobody has a method. One of the two 35% figures in circulation does, and it deserves credit the pages quoting it do not give it.

The audit that does have a method

The newer of the two 35% figures comes from a published call audit. Over four weeks, between 10 a.m. and 4 p.m. local time, the publisher placed 1,200 calls to small and mid-sized law firms across the US, using local caller ID numbers and the firms' publicly listed phone lines, and reports that 34.8% went completely unanswered. Sample size, window, hours, caller identification and the source of the numbers are all stated. Not one of the three ranked pages carrying a 35% figure reproduces any of that.

Two things sit alongside it. The publisher sells an intake product whose function is answering every call, so this is the seller of the remedy measuring the problem. And the audit records whether a call was answered, not whether the caller was a prospective client, because no matter was presented.

Then the same publisher extrapolates. The dollar figure built on top of that audit takes a national call volume, a conversion rate and an average case value, and sources none of them. Four further consumer-behaviour figures in the same article carry no source either, sitting beside the measured one with nothing to distinguish them. The audit and the extrapolation are two different objects and only one of them has a method.

The version of that study which reaches lawyers as trade news is a press release. It carries the method, and it is labelled sponsored content with a disclaimer that the magazine's editorial staff did not contribute and that the views are solely the source's. So far everything a reader needs in order to judge it is still attached.

Now follow the other chain, the $109 billion one, one hop further. A state bar practice-management article reproduces it in full as established fact: the 7% conversion rate, the $8,000 client value, 13.6 million lost clients and the total, presented as findings with no caveat attached. Read there, by a lawyer, in a bar publication, it is professional guidance rather than a vendor's estimate.

That is the shape of the problem. The parts that let a reader judge a figure, the method on one chain and the sponsorship label on the other, are attached at the source and detached by the time the figure reaches the audience it was aimed at.

One more, because it is the sort of thing a reader skimming would take for the study. The publisher of the $109 billion figure describes exactly one thing it did itself: it called 24 firms in one state about a truck accident case, and four answered. That is a different proportion from 35%, it is not offered as the basis of the figure, and it names no state, no date and no call protocol.

A second claim that behaves the same way

The idea that legal clients hire whoever answers first shows up constantly, and it shows up at different sizes. One ranked page states that 67% of legal clients choose the first attorney who answers their call. Another publisher states 78%. A third gives it as a 62 to 78% range. None of the three names a source.

We are not publishing any of them.

What has actually been measured

Set the multipliers aside. Real measurement of law firm phone behaviour exists, and it is usable as long as you keep hold of what each study counted.

Clio hired a third-party research firm to contact 1,000 law firms by email and phone with inquiries designed to assess responsiveness, published in its 2019 Legal Trends Report. The sample was equally represented across family, criminal, bankruptcy, business formation and employment law. No personal injury, no immigration, no estate work. If you run a PI firm, that study is about other people.

On the phone half, 56% of firms answered, 39% of calls went to voicemail of which 57% did not return the call within 72 hours, 5% were unanswered, and in total 73% either picked up or phoned back, meaning 27% of firms could not be reached by phone at all.

A second study, assessing 2024, contacted 500 law firms by email and phone. Its shoppers reached 52% of the firms they phoned and only 40% picked up when called, and 48% of firms were unreachable by phone even after being given the chance to respond to messages.

Those two studies are five years apart, at different sample sizes, and the 2024 report is behind a lead-capture form so its window and its definition of unreachable cannot be read. Do not subtract one from the other. They are not a trend line, and this page is not presenting them as one.

Both are published by a company that sells intake software. That does not make the measurements wrong. It does mean no independent replication exists in our evidence, and every figure above is attributed for that reason.

The callback is not the recovery

That 2019 report contains three scorecards, and the ninth-ranked result quotes two of them. It reproduces the email bands and the 7% of firms rated excellent on the phone. The third scorecard, the one about the returned call, appears in none of the pages we captured, and it is the one that matters most to the decision this page is about.

The report scores what happened once contact was made. Among the firms that returned the shoppers' voicemail: 0% and 0 firms rated Excellent, 0% and 0 rated Good, 4% and 3 firms rated Adequate, and 96% and 81 firms rated Unsatisfactory, where Unsatisfactory means the firm did not meet the criteria on more than half of the report's performance drivers.

Read that against what most of this category sells. The offer is coverage, so that the call gets returned. In the one study in our evidence that scored the returned call, three firms out of 84 got as far as Adequate and none did better.

Among firms that answered live, the picture is better and still not comfortable: 7% and 20 firms Excellent, 9% and 26 Good, 22% and 62 Adequate.

Two limits on that, both important.

What is being rated is information content, not outcome. The score averages how many questions were answered, whether experience with the case type was demonstrated, whether rates or cost were given, and whether next steps were given. It says nothing about whether the caller hired anyone. Nothing we hold links a scorecard band to a signed matter.

One row of that phone table does not add up, and we are saying so rather than quietly using it. The report states 56% of firms answered, which on 500 calls is about 280 firms. Three of the four bands carry counts consistent with that base. The fourth reads 61% and 61 law firms, where 61% of the same base would be about 171. Either the percentage or the count in that row is wrong. We do not know which, so we use the other three.

How a figure loses its denominator

The ninth result for this search is a legal marketing agency's account of that same Clio study, and it reports the two headline findings accurately: that 60% of firms never responded to emails, and that 27% did not answer phone calls or return voice messages. Both match the report.

Then it writes that Clio rated only half a percent of the firms as excellent at handling emails, while 71% were unsatisfactory.

The percentages are Clio's own. The base is not. Clio's email scorecard prints a firm count beside every band: 2 firms Excellent, 20 Good, 93 Adequate, 284 Unsatisfactory. Those counts sum to 399, which is what you get from 1,000 firms emailed with 60% never replying. The base is the firms that replied.

Taken against every firm emailed instead, those 284 firms are 28.4%. That is not a finding about law firms and we are not publishing it as one. It is the size of the gap between the two readings.

This is worth being fair about. The same article states the base correctly one sentence earlier, when it writes of law firms that responded to emails. The denominator survives one sentence and not the next. That is how it usually goes, and it is why the rule on this site is that a percentage and its denominator travel in the same sentence or neither ships.

What these pages get right

The competitive set on this query is mostly sold by people with a position in the answer, and it is still right about several things. Saying so is not a courtesy, it is accuracy.

The observation that missed calls are invisible in a firm's own reporting is correct and the ranked pages make it well. A prospective client who reached voicemail and moved on leaves no record anywhere the firm looks.

The mechanics they describe are concrete and real: overflow forwarding when a line is busy or rings past a set count, after-hours forwarding, a written intake script, delivery of call summaries into the firm's own system.

One page names the solo constraint more honestly than most consultants do. An attorney handling consultations, court appearances, client meetings and the phone cannot answer every time it rings, and no amount of discipline changes that.

And the most quantitatively elaborate page in the set is the one that shows its whole method, which is the only reason its arithmetic could be examined here at all.

One page in the set carries no statistics at all, no percentage and no dollar figure, and argues entirely from screenshots of consumer reviews. One capture in our set is partial, and its own header says so: only the first of its three named segments was returned, and the segment we did not get is the one titled how much a missed call costs. That is recorded in the capture's own header rather than in a claim, because a truncated fetch is a collection fact rather than a statement any source made. Nothing on this page is a statement about what that page does or does not cover.

What is actually failing

The commercial writing frames this as speed. The sources with no product at this step frame it as ownership, and they are more specific.

A state bar practice manual instructs firms to establish the order in which staff are expected to answer, gives a worked example in which the phone rings at a named second station when the receptionist is on a call and is picked up there after no more than three or four rings, and names the failure mode in its own words: avoid the situation where everyone is responsible for answering the phone and no one does.

That is the clearest non-vendor statement available that the problem is undefined ownership rather than latency, and it costs nothing to fix.

Two different state bar programmes place first contact with the least equipped person on the payroll. One states that law firm receptionists are the most poorly paid, least trained and least respected staff members. Another states that the duty is often given to the least experienced and lowest-paid staff member, and that in some smaller firms secretaries rotate it during the day. Two organizations, arriving at the same description independently.

Having named an owner is not the same as having named a successor. A bar practice-management director builds a firm meeting agenda around four questions to settle in writing: who receives the inquiry, what happens when that person is already on the phone, whether lawyers should handle initial inquiries at all, and how questions are answered. The second of those is the one the rest of this research found nothing written about anywhere.

There is a callback standard in the same manual: end of the business day where possible, and within 24 hours at the latest, with support staff returning the call and determining whether an emergency exists when the attorney cannot. Treat that as one bar programme's recommendation, which is what it is, not as an industry standard.

And speed is not a complete account of the failure anyway. The same practice-management director writes that most lawyers who represented individual consumer clients have had the experience of returning an inquiry call within the hour and finding the person had already talked to and scheduled with another lawyer, which he offers as a shared recollection rather than as a measurement. He also states there is no one size fits all answer, contrasting an antitrust defence practice, which he says is probably not going to lose a client if a message is taken and returned within an hour or so, with DUI defence and family law firms whose callers may be working down a list.

Who should take the call is genuinely contested and we are not resolving it. Two practice-management consultants state that lawyers should not be taking initial calls at all, and that a well-trained assistant, receptionist or answering service working from a script can do it as well or better at a fraction of the cost. A named plaintiff-side employment lawyer states the opposite about his own practice, preferring to do intake himself in order to judge the prospective client's demeanour, how they respond to questions and their credibility.

One more thing a buyer should know before signing anything, because it is the failure that no dashboard shows. A firm reports agents at its answering service consistently telling prospective clients that their matter was outside the firm's services when it was not, alongside routing to the wrong attorneys and failing to capture phone numbers. A prospect who was told the firm cannot help them generates no missed-call log entry, no voicemail and no message. The service's own reporting shows calls answered promptly. Coverage bought to fix an invisible leak can create a second one, and finding it requires recordings or transcripts checked against the firm's own criteria.

The four numbers to count instead

Nothing above gives you a benchmark, because no defensible one exists. What it gives you is a reason to stop estimating your own missed-call rate when you can measure it.

1. Calls received, and calls answered by a person. Both are in your phone system already. The ratio is the only input in that formula you can obtain, and once you have it the published rates stop mattering to you.

2. Voicemails returned within one business day. Count them against voicemails received, not against calls received. In the one study that measured it, 57% of voicemails were not returned within 72 hours, so this is where to look before buying anything.

3. Of the calls that were returned, how many reached a person who could discuss the matter. This is the one the scorecards suggest matters, and no source captured for this research reports tracking it. It is also the only one of the four that requires listening rather than counting.

4. Whether an answer order and a named successor exist in writing. Not a number, a yes or no. It is the step the bar manual puts before every other step, and a firm that has not taken it is buying coverage for a problem it has not defined.

A firm that has those four can tell whether it has a phone problem, a callback problem or a conversation problem. Those need different fixes and cost different amounts, and no published figure can tell you which one you have.

If you want the dollar figure after that, run the same three-term formula the vendors run, using your own call count. Just be clear with yourself that the other two terms are still guesses, and that the answer moves by a factor of four and a half depending on which published guesses you borrow.

What this page does not establish

How this page was built

Nine results were returned in a dated US search capture for this query, and eight of them were fetched. One of those eight came back partial, and its own capture header says so, which is why no judgment about that page's coverage appears here. The result that was not fetched at all is a Facebook post ranked third, because that platform returns HTTP 403 to our collection tools, so it was not read and nothing here rests on it.

Three of the figures on this page were followed to their cited sources rather than quoted from the pages carrying them. That took four fetches and ended at a 404, a 410 and a hostname that does not resolve.

Every number here traces to a claim in our internal evidence base, each carrying a verbatim excerpt and a dated snapshot of the page it came from. The two derived figures on this page, the input sensitivity range and the email scorecard denominator, show their formula and their inputs.

The standards this page is held to, including what counts as a source, what may never be published, and how figures are verified, are set out in full at methodology. The companion research on how fast firms reply, rather than whether they reply at all, is at law firm lead response time.

We publish no dollar cost of a missed call. If a conversion rate for unanswered calls and an average matter value for phone-originated consumer matters are ever published by someone without a position in the answer, this page will change.

Nothing on this page is legal advice, and nothing here is a benchmark your firm should be held to.