I Went Looking for the Sources Behind Sales Coaching's Favourite Statistics

The short version: Four statistics repeated constantly in sales enablement — 90% of training unapplied within 90 days, an 88% productivity lift from coaching, managers reviewing under 1% of calls, and the CEB "middle 60%" finding — do not trace to locatable primary research. The underlying observations are often directionally sound. The numbers are not evidence.

I have used some of these numbers. That is why I went looking.

You know them. They appear in every enablement deck, every vendor one-pager, every LinkedIn post about why training doesn't stick. They have the shape of research. They carry decimal points and percentage signs and the quiet authority of something somebody once measured.

I tried to find who measured them.

---

"90% of training is not applied within 90 days"

This is the load-bearing statistic of the entire sales enablement industry. It justifies reinforcement platforms, coaching programmes, follow-up curricula, and a great deal of consulting revenue.

Follow it back and it goes nowhere.

Every article citing it cites another article. Those articles cite blog posts. The blog posts cite nothing, or cite each other. The trail runs for pages and terminates in an empty room.

There is real, serious literature on training transfer. Baldwin and Ford's 1988 framework. Blume, Ford, Baldwin and Huang's 2010 meta-analysis of 89 studies. Arthur and colleagues on skill decay. That work is careful, and it finds real problems — transfer depends heavily on the work environment, on whether the skill gets used quickly, on whether anyone locally reinforces it.

None of it produces "90% within 90 days."

The finding is directionally consistent with the literature. The number is invented.

"Coaching improves productivity 88% versus 23% for training alone"

Two decimal-free numbers, side by side, delivering a clean four-to-one story. It appears everywhere. It has appeared in my own material.

I could not find the study.

The citation chain leads to vendor blogs, which lead to other vendor blogs, occasionally to a statistics-roundup page that cites a source that cites a source. At no point does anyone produce a sample size, a methodology, a population, or a year.

A number that specific should have a paper behind it. This one has a lineage of people repeating it.

"Managers review less than 1% of sales calls"

This one is instructive because the industry cannot keep its story straight.

Some sources say less than 1%. Others say under 5%. Others say 5–10%. The same claim, cited to "industry benchmarks," varying by an order of magnitude depending on which vendor is telling you.

When a statistic ranges 1% to 10% across sources that all claim to be reporting the same measurement, nobody measured it. Somebody estimated it, someone else rounded it, and a third party put it in a headline.

The underlying observation is almost certainly true. Managers review a small fraction of calls — you can derive that from arithmetic alone. Eight reps, two hundred conversations a month, one manager with a calendar. You do not need a study.

But "obviously true" and "measured" are different claims, and the moment you attach a number you have made the second one.

"The middle 60% is where coaching pays off"

This is the most widely repeated finding in sales management, attributed to CEB, now part of Gartner. Coach the core performers. The stars will succeed anyway. The strugglers have a fit problem coaching can't fix.

It is repeated with total confidence by consultancies, training firms, and enablement platforms, all of whom cite CEB.

I could not locate the underlying publication. Not the methodology, not the sample, not the analysis. Only the conclusion, quoted onward, for fifteen years.

It may well be right. It has the ring of something an experienced sales leader would recognise. But "sounds right and everyone repeats it" is the definition of folklore, not evidence, and the distinction matters most precisely when the folklore is plausible.

Why this is worth caring about

The easy objection is that this is pedantry. The claims are roughly right. The direction is correct. Sales leaders aren't reading footnotes.

Three reasons it matters anyway.

Unsourced numbers cannot be corrected. A real finding has boundaries. It applies to this population, under these conditions, with this effect size. You can check whether your situation resembles the study's. A folk statistic has no boundaries, so it is applied everywhere, including where it is false, and there is no mechanism by which anyone ever discovers this.

They crowd out the real research, which is better. Kluger and DeNisi's finding — that over a third of feedback interventions decreased performance across 607 effect sizes — is more surprising, more useful, and more actionable than any of the numbers above. It is also almost entirely absent from sales enablement discourse. The fake statistics have taken the shelf space.

And they are corrosive. If you build your case on numbers that don't survive a search, you have handed a sceptical buyer a reason to dismiss everything else you said. He won't tell you why. He'll just stop returning calls.

The standard

I now hold myself to a rule, imperfectly and with some past sins to answer for.

If I cannot name the author, the year, and the publication, I don't use the number. I make the argument in words instead — which is usually stronger anyway, because an argument invites thought and a statistic invites nodding.

The real literature is available. It is less convenient than the folklore. Nothing in it fits on a slide.

That's the trade.

Common questions

Where does "90% of training is not applied within 90 days" come from?

It does not trace to a locatable study. Articles citing it cite other articles, which cite blog posts, which cite each other or nothing. The genuine training-transfer literature — Baldwin and Ford, Blume and colleagues, Arthur and colleagues — finds real transfer problems but produces no such figure.

Is the "88% productivity increase from coaching" statistic real?

No primary source is locatable. The citation chain leads through vendor blogs and statistics-roundup pages without ever producing a sample size, methodology, population, or year.

What percentage of sales calls do managers actually review?

Nobody has measured it. Published claims range from under 1% to 10%, all attributed to unnamed industry benchmarks. The underlying observation is almost certainly true and can be derived from arithmetic — but "obviously true" and "measured" are different claims.

What sales coaching research is actually trustworthy?

Kluger and DeNisi (1996) on feedback interventions, Blume and colleagues (2010) on training transfer, Baldwin and Ford (1988), Arthur and colleagues (1998) on skill decay, and Leonard and Masatu (2010) on the know-do gap. All peer-reviewed, all locatable, all more interesting than the folklore.

A statistic you cannot source is not evidence. It is a rumour wearing a lab coat.

Real sources, for anyone who wants them

• Kluger, A. N., & DeNisi, A. (1996). Psychological Bulletin, 119(2), 254–284.
• Blume, B. D., Ford, J. K., Baldwin, T. T., & Huang, J. L. (2010). Transfer of Training: A Meta-Analytic Review. Journal of Management, 36(4), 1065–1105.
• Baldwin, T. T., & Ford, J. K. (1988). Transfer of training: A review and directions for future research. Personnel Psychology, 41(1), 63–105.
• Arthur, W., Bennett, W., Stanush, P. L., & McNelly, T. L. (1998). Factors that influence skill decay and retention. Human Performance, 11(1), 57–101.
• Leonard, K. L., & Masatu, M. C. (2010). Journal of Development Economics, 93(2), 226–234.