The Folklore Audit

Every corner of the missed-call and business-texting industry quotes the same handful of statistics. They appear in ads, in vendor blogs, and in listicles that all quote each other, an endless loop of citation. In July 2026, while rebuilding our own missed-call statistics page, we traced each one to its source instead of repeating it. This page is the full record of where each chain ends, re-checked in August 2026 before publishing. Two of these figures used to appear on our own site; that is part of why we ran the trace.

The verdicts

We deliberately don’t name the companies currently quoting these figures. Most are repeating, in good faith, what everyone else repeats; the problem is the loop, not any one link in it. We name a source only where that source itself published the correction.

Four ways a number turns out to be nothing

1. There is no study at the end of the chain

The 85% claim is the canonical example, and the one we followed furthest. The attributions in circulation contradict each other: the same number is credited to a well-known marketing-research company that never published any such survey, to a “report” whose link resolves to a vendor’s homepage, and to a named business-communications study nobody can produce. The deepest chain ends at an unsourced sentence in a 2016 vendor blog post; every other chain goes nowhere at all. The closest thing to a primary source anywhere near this family is a 2014 Forbes contributor column, still live as of August 2026, offering the related claim that 80% of callers sent to voicemail don’t leave messages, prefaced only by “recent statistics show.” A contributor column quoting no source is not a study.

The 78% claim fails the same way with a twist: it is usually pinned to real research. The Lead Response Management study exists, and it is good; we read it in full, along with its full report and its 2014 follow-up. The 78% figure appears in none of them. Its actual attribution, a “Lead Connect survey,” has no published report, no sample, and no method. What the LRM study did measure is worth quoting instead: across more than fifteen thousand leads and a hundred thousand call attempts, the odds of contacting a lead fell roughly tenfold once the first hour had passed, and the odds of qualifying one fell about sixfold.

The after-hours family is the emptiest of all. We could not locate a primary source for any published share of trade calls arriving after hours. If you need that number, the honest answer is that nobody has published a measurement we could find.

2. The document is real, the number isn’t in it

This failure is more dangerous than pure invention, because the citation looks checkable and almost nobody checks it. BIA/Kelsey is a real research firm; the “10 to 15 times better” conversion figure is not in the work it is credited to. The 2012 Google/CEB study is real; the “35% to 50% of B2B sales” figure is not in it. Twilio publishes a real State of Customer Engagement report; the SMS response and preference rates attributed to it are, in the edition we read, simply not there.

3. The source itself took the number back

The “98% open rate” is the industry’s most-loved statistic, and it and its sibling, “90% read within three minutes,” are the only two on this page with a real origin: a single 2010 white paper by the messaging analyst firm Mobilesquared. In August 2022 Mobilesquared publicly withdrew both, writing that “neither datapoints represent an accurate reflection of today’s world of mobile messaging,” and put the current SMS read rate at 55%. Correcting your own most-quoted number, twelve years on, is the most honest act anywhere in this trace, and the industry’s response was to keep quoting 2010. The retraction is published on Mobilesquared’s own site, in a post dated August 2022 in its URL; we re-read it in August 2026 and it still stands.

4. The seller ran the survey, and the evidence vanished

The 69% claim, usually worded “69% of callers who reach voicemail don’t leave a message,” is the newest trace in this audit; we ran it fresh in August 2026. Unlike the 85%, it has an identifiable origin: an answering-service company’s own 2016 press release, announcing a self-commissioned survey it described as 300 micro businesses plus call data from 10,000 businesses. That press release is where every chain ends. The report itself verifiably existed, because trade press linked the PDF in November 2016; as of August 2026 it is no longer publicly retrievable, the Internet Archive holds no capture of it, the current copy sits behind a lead-capture form, and no published version of the method names the survey firm, the field dates, or even the country. A number whose method cannot be read cannot be checked, and here the seller of the remedy ran the survey.

One warning for anyone tracing this family: two other 69% figures circulate nearby, and they are different claims. A real 2025 Harris Poll, run for a restaurant-phone vendor with a published sample of 2,065 US adults, field dates, and a precision statement, found 69% of Americans say they would give up on a restaurant that doesn’t answer the phone: a stated intention, already being reworded into measured caller behavior barely a year after it ran. And a 2019 figure that 69% would prefer a text from an unfamiliar company lives only in a texting vendor’s own press release. Check which 69% you are being quoted.

Methodology, in full

• When: traces first run in July 2026, during the rebuild of our statistics page; the ninth claim, the 69%, was traced fresh in August 2026; every verdict re-checked in August 2026 before this page shipped.
• What we traced: the nine claims above, selected because they are the figures we most often saw quoted in missed-call and business-texting marketing, including, at one time, our own.
• How: for each claim, collect the attributions in circulation; fetch every cited document and read it in full, from the publisher’s own site where it survives and from the Internet Archive where it does not survive and an archived copy exists; follow each citation to its terminus. Image-only PDFs were rendered and read rather than trusted secondhand.
• The counting rule: a claim counts as sourced only when a primary document with a described method contains the number. A vendor blog citing another vendor blog does not count, however many times the loop repeats.
• A trap worth passing on: search spelled-out numbers as well as numerals. Pew Research writes “Eight-in-ten,” so a search for “81%” returns nothing and makes a real, well-sourced figure look fake. We nearly “corrected” an accurate number this way.
• Honest limitations: a trace can prove where a chain ends, not that no study exists anywhere. Any of these numbers could be redeemed tomorrow by someone producing a primary source, and we would like that: send it to hello@getbeepback.com and we will link it here. Documents also move and get deleted; two already have, which is why the identifying details are recorded inline. And our own incentive is disclosed: BeepBack sells missed-call text-back, so inflated folklore about missed calls would flatter us. That is exactly why we won’t use it.

What a real number looks like

Real measurements of this problem exist; they are just less quotable than the folklore. Pied Piper’s 2026 home-services study ran 3,211 mystery-shopper evaluations across 31 brands and found no follow-up at all 72% of the time. Pew Research surveyed 10,211 US adults on a probability panel in 2020 and found eight in ten say they don’t generally answer their cellphone when an unknown number calls. And when we wanted to know how often the trades answer the phone at all, we called 133 New Jersey plumbing shops ourselves and published the full methodology. Every sourced figure, with its sample and its weaknesses stated inline, lives on our missed-call statistics page.

Common questions

Is the “85% of callers who can’t reach a business never call back” statistic real? We traced it in July 2026 and found no study behind it. The attributions in circulation contradict each other, several point to reports that were never published, and the deepest chain ends at an unsourced sentence in a 2016 vendor blog post. Treat it as folklore until someone produces a primary source.

Is the “78% of jobs go to whoever responds first” statistic real? We traced it in July 2026 and re-checked it in August 2026: it is credited to a “Lead Connect survey” that has no published report, sample, or method, and it does not appear in the Lead Response Management study it is often pinned to. What that study actually measured: the odds of contacting a lead fall roughly tenfold once the first hour has passed, and the odds of qualifying one fall about sixfold.

Is the “98% of texts are opened” statistic real? It comes from a single 2010 white paper whose own publisher, Mobilesquared, publicly withdrew it in August 2022, writing that “neither datapoints represent an accurate reflection of today’s world of mobile messaging.” Mobilesquared’s current figure, still standing when we re-read the statement in August 2026, is that 55% of SMS messages are read.

Is the “69% of callers who reach voicemail don’t leave a message” statistic real? We traced it in August 2026. It originates in an answering-service vendor’s own 2016 press release announcing a self-commissioned survey; the report it announced is no longer publicly retrievable, the Internet Archive holds no capture, and the published method names no survey firm, no field dates, and no country. A separate, real 2025 Harris Poll found 69% of Americans say they would give up on a restaurant that doesn’t answer the phone; that is a different claim, about stated intention, and it is often confused with this one.

What share of calls to the trades come in after hours? Nobody has published a real number, as far as we could find in July 2026. The widely quoted 42% to 48% family traces to studies nobody can produce: a contractor phone-line study nobody has ever shown, a trade-magazine survey that magazine never published, and a report whose URL was already dead when we traced it.

Why we did this (and what BeepBack is)

We build BeepBack: it texts your missed callers back in seconds, asks the few questions you’d ask, and hands you a sorted, ready-to-call lead. Selling that product is easier with “85% never call back” on the page, which is precisely why that number needed tracing before we would print it. It didn’t survive, so we publish our own counts instead. It’s $79 a month flat, no contract, and the case for it rests on the sourced numbers, which are bad enough.