Where the claim that 5% more retention lifts profit by 25% to 95% comes from

What the 1990 article actually argued

The source is "Zero Defections: Quality Comes to Services," by Frederick Reichheld and W. Earl Sasser, Jr. Bain & Company dates it September 1, 1990. Harvard Business Review lists it in the September–October 1990 issue.

Reichheld is a Bain Fellow and an advisory partner at Bain & Company. Sasser is a Baker Foundation Professor at Harvard Business School.

Its central claim, as Bain states it, is that companies can boost profits by almost 100% by retaining just 5% more of their customers. The 25% to 95% range is not in that sentence. The headline figure is a single number close to a doubling.

The article's subject is services. Its opening argument is that service companies are starting to learn what manufacturers learned in the 1980s: quality does not improve unless it is measured.

Why the article borrowed the language of manufacturing

HBR's introduction to the article describes how manufacturers came to see scrap, rework and jammed machinery as costly. They concluded that quality was the most profitable way to run a business, and they adopted "zero defects" as a guiding principle.

The article carries that idea over to customers. The article says service companies have their own kind of scrap heap: customers who will not come back. The title's phrase "zero defections" is the quality-control idea applied to the customer base.

The article defines the goal with a limit. Zero defections means keeping every customer the company can profitably serve. It also notes that customers are not captive, so a firm has to keep outperforming its competitors to hold them.

The company at the centre of the evidence

The one company Bain's summary names is MBNA America, a credit card company based in Delaware. The story begins in 1982, when its president, Charles Cawley, gathered all 300 employees and committed the company to keeping every customer.

The company gathered feedback from customers who left and adjusted its products and processes regularly. By the time of the article, about 5% of its customers left each year, which the article gives as half the industry average.

Over the same period its industry ranking rose from 38 to 4, without any acquisitions. Its profits grew sixteenfold.

It is also one credit card company, covering roughly eight years. The summary names no other company.

The arithmetic behind a 5% gain

The article says that as a customer's relationship with a company lengthens, profits rise.

A defection rate converts into an expected customer lifetime. If 10% of customers leave each year, the average customer stays about ten years. If 5% leave each year, the average stays about twenty. These are illustrative calculations, not figures from the article, but they use the MBNA numbers: a 5% defection rate described as half the industry average implies an industry rate near 10%.

Halving defections therefore doubles the lifetime over which each acquired customer pays. The article says the 5% gap "translates into huge earnings."

Companies can boost profits by almost 100% by retaining just 5% more of their customers.

Defection as a leading indicator

The article describes defection rates as an accurate leading indicator of profit swings, which makes them something a firm can watch before the financial results move.

It also says that soliciting feedback from defecting customers lets companies ferret out the weaknesses that really matter. That matches the MBNA method of collecting feedback from departing customers and changing products and processes in response.

The article ties defections to quality: as quality improved, fewer customers had reason to leave.

How the claim became a range

The 25% to 95% wording appears in a Harvard Business Review article by Amy Gallo, a contributing editor, published October 29, 2014, on the value of keeping the right customers. It credits Frederick Reichheld of Bain & Company, whom it also identifies as the inventor of the net promoter score, with the finding that increasing retention rates by 5% increases profits by 25% to 95%.

The article does not name a study, describe a method or list the industries covered. It presents the figure as support for the value of retention and does not qualify it.

A reader tracing the number therefore finds two forms of it. The 1990 summary gives almost 100% for a 5% gain. The 2014 article gives a band from 25% to 95% for the same gain.

The same 2014 article hedges a neighbouring claim. It says acquiring a new customer costs anywhere from five to 25 times more than retaining one, then adds that this depends on which study is believed and on the industry.

How far the figure can be applied today

The 1990 article concerned service companies, and its named case is a credit card issuer.

The 2014 article attaches a caveat to its acquisition-cost figure: it depends on which study is believed and on the industry. The profit figure carries no such caveat in that article.

The 1990 article's own limit points the same way. Keeping every customer a company can profitably serve is a narrower target than keeping every customer.