The follow-up statistic its own author quietly deleted

Every sales blog on the internet repeats it: 80% of sales need five follow-ups, and 44% of salespeople give up after one. We traced both. The page everyone cites is a 404, the author revised the number down and deleted the claim, and the 44% comes from a paperback published in 1987.

You have seen this sentence, probably this month: 80% of sales require five follow-ups, yet 44% of salespeople give up after one. It appears in sales training decks, CRM marketing, LinkedIn posts and roughly every article ever written about persistence. It is usually credited to something called Marketing Donut, sometimes to Scripted, occasionally to the National Sales Executive Association.

We went looking for the study. There is no study. There is a guest column, and its author has since changed his mind.

The page everybody cites returned a 404 while we were writing this

The Marketing Donut article is the canonical citation for this statistic. It is gone. The URL returns HTTP 404, and the last archived capture showing a working page is dated 4 February 2026. It died this year, quietly, and every citation still points at it.

We recovered the full text from the Internet Archive and read it. Two things are immediately clear. It was never Marketing Donut's research, it is a bylined guest column by Robert Clay of Marketing Wizdom, a marketing consultancy. And its entire evidential basis is one sentence: the figures come from "different studies ... by different market research companies over a number of years". No study is named. No year is given. There is no sample size anywhere on the page.

The author revised it, and cut the famous line

Clay still publishes the article on his own site, revised on 4 August 2025. Comparing it against the archived Marketing Donut version that the industry still quotes:

The version everyone citesThe author's own current version, August 2025
8% of salespeople get 80% of the sales8% of salespeople are responsible for 60% of all sales
80% of prospects say no four times before they say yes60% of customers say no four times before they say yes
80% of non-routine sales occur only after at least five follow-upsAbsent. The sentence is gone.
Only 20% of sales leads are ever followed upAbsent.
Only 2% of sales occur when two parties meet for the first timeAbsent.
A five “nos” strategyA four “nos” strategy
Left: the Marketing Donut version, archived 4 February 2026, now returning 404. Right: Robert Clay's own revision on marketingwizdom.com, dated 4 August 2025, read in full. The marked row is the claim the entire industry repeats.

The man who originated the most-cited statistic in sales has walked it back from 80% to 60% and deleted the five-follow-ups claim entirely, and nobody citing him has noticed. Every article published since August 2025 that credits "80% of sales require five follow-ups" to Marketing Donut is quoting a version its own author no longer stands behind, on a page that no longer exists.

The 44% is real, and it is from 1987

Clay's current version does something the Marketing Donut one never did. It names a source for the second half of the pairing: Michael LeBoeuf's How to Win Customers and Keep Them for Life, G.P. Putnam's Sons, 1987. The cascade appears verbatim in the book. 44% give up after one no, 22% after two, 14% after three, 12% after four.

The other numbers in this family, traced

The touch-count statistics are a related genre. Some are real, and the difference between them is worth knowing before you quote one.

  • 5.7Contact attempts to a win. VanillaSoft with the Telfer School of Management, University of Ottawa: 130 million interactions across roughly 550 companies. Preliminary results, 2018.
  • 8Touches to secure an initial meeting, not to close. RAIN Group, 488 buyers and 489 sellers, self-reported. Fieldwork 2018 or earlier, restamped 2026, gated behind the course it sells.
  • 7 to 13Not a measurement. A hand-written illustrative sequence. The authors' own two case studies came in at 4.28 and 7.51 touches.
  • 12The fifth-to-twelfth contact cascade. Traced by the attributed body itself to a 1942 survey of the Long Island chapter, sample under 40.

No sample, or an illustrative listStated sample, measured

Touches or contact attempts, by what is actually behind each figure. Note that these do not measure the same thing: 8 is touches to secure a first meeting, 5.7 is contact attempts to a win, and the 7 to 13 range is a hand-written example sequence rather than a measurement.

That last one deserves its own sentence. The famous cascade, that 2% of sales are made on the first contact and 80% between the fifth and twelfth, is attributed to the National Sales Executive Association. In 2014 VentureBeat went looking for that organisation and reported it did not exist. In 2021 the body that does exist, Sales and Marketing Executives International, answered from its own archive: NSEA was its former name, so the attribution was not fabricated. The survey was conducted in 1942, among members of the Long Island chapter, and the sample size was under 40.

What the research actually supports

Two figures here have a stated sample and are worth more than the rest of the genre combined.

5.7 contact attempts to a win. VanillaSoft, working with the Telfer School of Management at the University of Ottawa, analysed 130 million call interactions from around 45 million leads across roughly 550 companies. The same study puts the optimal first-response window at 10 to 60 minutes, which contradicts the five-minute rule the same industry repeats. VanillaSoft sells auto-dialling software, so it is selling the persistence its finding recommends, and the study was published as preliminary in 2018 with no completed version we could find.

70% of marketing-generated leads are never pursued by sales. Sabnis, Chatterjee, Grewal and Lilien, Journal of Marketing, 2013, peer-reviewed. We could not read the full text, which is paywalled, so we cannot tell you its sample and we are not going to pretend otherwise.

And one finding that explains why this whole category is the way it is. A 2022 paper in the Journal of Business Research states plainly that rigorous academic research on prospecting is virtually nonexistent. That is the actual state of the evidence. Into that vacuum went a 1942 survey of forty people, a 1987 paperback, and a guest column with no sample, and they have been circulating ever since.

What this means if you are trying to fix your follow-up

We build follow-up systems for service businesses, so we are in the same category as most of the publishers above and you should weigh that. Which is exactly why there is no touch-count anywhere else on this site.

The practical version:

  1. Stop quoting the number, including at yourself. Any target derived from it is a target derived from nothing. If your sequence is five steps because of that statistic, it is five steps for no reason.
  2. Count your own. How many attempts did your last twenty won deals take? That is a real number, it is specific to your trade, and you can have it this afternoon. Almost nobody has it.
  3. Separate touches to a meeting from attempts to a close. They are different measurements and the published figures are not interchangeable. Most of the confusion in this genre comes from merging them.
  4. The failure is almost never the wording. Nobody's follow-up collapses because the second email was phrased badly. It collapses because the second email never went out. That is a systems problem and no template fixes it.
  5. If you take one figure from this article, take the one about leads never pursued. 70% is peer-reviewed and it describes the actual failure: not insufficient persistence, but no contact at all.

The honest summary is that nobody knows how many follow-ups it takes, and the people who told you they did were quoting each other. The number you should be working from is your own, and the reason to build a system is not that five is the magic figure. It is that whatever your figure turns out to be, it will not happen reliably while it depends on somebody remembering.

Sources

Publisher, year, sample, and what the publisher sells. The same four things every cited figure on this site carries.

Why just 8% of salespeople win 60% of the business

Publisher, year and sample
Robert Clay, Marketing Wizdom, current version dated 4 August 2025, read in full. This is the author's own revision of the article that Marketing Donut published and that the sales industry cites second-hand. The 80% is now 60%, the five “nos” strategy is now four, and the sentence “80% of non-routine sales occur only after at least five follow-ups” has been removed, along with the 20% and 2% claims. The Marketing Donut version, archived 4 February 2026, now returns HTTP 404. Neither version states a sample size for any figure; the original attributes them only to “different studies by different market research companies over a number of years”.
What the publisher sells
Marketing Wizdom is a marketing consultancy. The same page promotes a growth programme with an earnings disclaimer in the footer. Marketing Donut is a UK small-business advice site funded by sponsorship and advertising; the article was a guest contribution rather than its own research.

How to Win Customers and Keep Them for Life

Publisher, year and sample
Michael LeBoeuf, G.P. Putnam's Sons, 1987; revised edition Berkley, 2000, ISBN 0425175014. Source of the 44% / 22% / 14% / 12% cascade, which appears verbatim in the text. The book states the percentages as received wisdom: no survey, no sample size and no citation. This is the true origin of a figure commonly attributed to Marketing Donut, to Scripted and to the National Sales Executive Association, all of which are incorrect.
What the publisher sells
A popular business paperback by a management professor and author, not a research publication. Naming it identifies the origin without supplying a denominator.

Sales statistics: the 1942 survey behind the fifth-to-twelfth contact claim

Publisher, year and sample
Sales and Marketing Executives International, 4 August 2021, by Willis Turner. SMEI is the current name of the National Sales Executives Association, the body to which the 2%/3%/5%/10%/80% cascade is attributed. Searching its own archive, SMEI reports that the survey was conducted in 1942 among members of the Long Island, New York chapter, and that “the sample size was less than 40”. This answers VentureBeat's 2014 investigation, which reported it could not locate the organisation at all: it existed under a former name, and the underlying statistic is worse than the accusation.
What the publisher sells
SMEI sells professional sales and marketing certifications and courses. It published this against its own interest, which is unusual enough to be worth saying.

Top performance in sales prospecting

Publisher, year and sample
RAIN Group Center for Sales Research. 488 buyers representing $4.2 billion in purchases across 25 industries, plus 489 sellers, self-reported. Reports an average of 8 touchpoints to secure an initial meeting, and 5 for top performers. The page carries “Last Updated June 11, 2026” but archive evidence places the research live by March 2019, so the fieldwork is 2018 or earlier and has not been repeated. Measures touches to get a meeting, not follow-ups to close a sale, and the two are routinely conflated.
What the publisher sells
RAIN Group sells sales training, coaching and a prospecting course. The full report is gated behind a lead-capture form on the page selling that course.

Lead conversion study with the Telfer School of Management

Publisher, year and sample
VanillaSoft with the Telfer School of Management, University of Ottawa. Collected 220 million call interactions across 70 million leads from roughly 550 companies; the analysed sample is 130 million interactions from about 45 million leads, spanning 2005 to 2017. Reports an average of 5.7 contact attempts to a positive outcome in business-to-business and 5.9 in business-to-consumer, and an optimal first-response window of 10 to 60 minutes rather than five. Published as preliminary results in 2018 with a completed version promised for that autumn, which we could not find.
What the publisher sells
VanillaSoft sells sales-engagement and auto-dialling software, so it is selling the persistence its own finding recommends. The sample is nonetheless the largest with a stated denominator that we located on this subject.

Getting the most out of prospecting (Journal of Business Research)

Publisher, year and sample
Gopalakrishna, Crecelius and Patil, Journal of Business Research, 2022, volume 149, pages 916 to 926, DOI 10.1016/j.jbusres.2022.05.008. Peer-reviewed. States that rigorous academic research on prospecting is virtually nonexistent, which is the context every figure in this article sits inside.
What the publisher sells
A peer-reviewed journal. Nothing is being sold.

Sales lead qualification and the leads never pursued

Publisher, year and sample
Sabnis, Chatterjee, Grewal and Lilien, Journal of Marketing, 2013, volume 77 issue 1, pages 52 to 67, DOI 10.1509/jm.10.0047. Peer-reviewed. Source of the finding that roughly 70% of marketing-generated leads are never pursued by sales. The full text is paywalled and we did not read it, so we cannot state its sample and have not implied one.
What the publisher sells
A peer-reviewed journal. Nothing is being sold. Listed with its limitation stated because that is the standard the rest of this article is applying to everybody else.

We would ratheruse your numbers.

Every article here ends the same way, because the honest answer is always the same: the benchmark is somebody else’s and your own figures are the ones that decide anything. On the call we work them out with you.

Fourteen questions, about seven minutes. No price, no purchase, and nobody calls you unless you ask them to.

Application1 / 14

Next question: where we send what we prepare.