breme.

Research · Retention and loyalty

Is keeping customers really cheaper than finding new ones? What the evidence says

By BREME · · 6 minute read

In short

The claim that a 5% rise in retention lifts profits by 25% to 95% comes from Bain and Company work led by Frederick Reichheld. It began in 1990 as 25% to 85% for service firms and was restated for online businesses in 2000. It rests on the economics of customer tenure, not a broad study of firms. The 'five times cheaper' rule has no primary study we could trace. Retention matters, but only for the right customers and at an honest cost.

Who this is for: CX leads, operations managers, finance partners and executives in membership bodies, subscription brands, professional services firms and multi-site organisations who are building a business case for retention work.

Two numbers that end up in every business case

Almost every retention proposal we see leans on one of two lines. A 5% increase in customer retention lifts profits by 25% to 95%. And acquiring a new customer costs five times more than keeping an existing one.

Both appeared together in a 2014 Harvard Business Review article by Amy Gallo. It attributed the first to research by Frederick Reichheld of Bain and Company. For the second it gave a range of five to 25 times, 'depending on which study you believe', without naming one.

Those numbers carry a lot of weight in board papers and budget rounds, so we went looking for where they came from, what the original work measured, and what has been found since.

Where the 25% to 95% figure came from

The idea starts with a 1990 Harvard Business Review article, Zero Defections: Quality Comes to Services, by Frederick Reichheld and W. Earl Sasser. Bain's summary of it says companies can boost profits by almost 100% by retaining just 5% more of their customers. It argued that service firms should track the customers who leave the way manufacturers track defects, because profits rise the longer a customer stays.

The article's best known example was MBNA, a US credit card company. Bain's summary says that after its chief executive committed to keeping every customer in 1982, about 5% of MBNA's customers left each year, half the industry average, and its profits grew sixteenfold.

When Ipsos later quoted the 1990 claim, the range was 25% to 85%. The 95% upper figure appears a decade later, in a July 2000 Harvard Business Review article by Reichheld and Phil Schefter on online businesses. Bain's summary of that piece says that by retaining 5% more customers, e-companies can boost profits by 25% to 95%.

Bain's own later framing was narrower. In a 2003 Marketing Management article, Reichheld and Christine Detrick wrote that in financial services, a 5% increase in retention produces more than a 25% increase in profit, because returning customers buy more over time, cost less to serve, refer others and often pay a premium.

So the famous range is a blend of two articles ten years apart: one about US service firms in the 1980s, one about early online businesses. We could not read the full 1990 article, which sits behind HBR's paywall, so we have relied on Bain's summary and on how others quote it.

What the original work actually measured

The strongest critique comes from Byron Sharp of the Ehrenberg-Bass Institute. He argues the headline is misleading in three ways.

First, the '5%' is five percentage points, a fall in defection from 10% to 5%. That is a halving of the customers you lose, not a small nudge. Second, he describes the core logic as a thought experiment: suppose a credit card company loses 10% of its customers a year, then work out what happens if average customer life doubles. Third, he says the authors assumed the halving cost nothing to achieve. His conclusion is that the article does not empirically show that lower defection causes large increases in company profit.

Ipsos made similar points in 2005. A 25% to 85% profit gain implies a very low starting margin, retention gets more expensive as it rises, and profit is far more uneven across customers than revenue. It cited activity-based costing work in which the top fifth of customers produced 150% to 300% of profits while the bottom fifth erased 50% to 200%.

Our reading is that the original work modelled what happens to profit per customer when customers stay longer, in service businesses with high acquisition costs and recurring revenue. That is a useful idea. It was never a measured law that holds across industries.

The 'five times' rule has no traceable source

The second claim is weaker still. In their 2005 book Loyalty Myths, Keiningham, Vavra, Aksoy and Wallard of Ipsos tried to trace it. In an excerpt published by Ipsos, they say the earliest sources they could find credit research by a Washington DC customer service research group in the late 1980s, and that other firms soon claimed the same finding as their own.

They say the line gained credibility in a 1990 Harvard Business Review article, The Profitable Art of Service Recovery, by Christopher Hart, James Heskett and W. Earl Sasser, and in Tom Peters' book Thriving on Chaos. HBR's page for that article showed us only its opening, so we could not confirm the wording ourselves. The authors also admit they had repeated the figure in their own earlier work.

They name three flaws:

  • Costs are misallocated. Advertising and promotions reach existing customers too, so loading all of that spend onto acquisition inflates the ratio.
  • It ignores the life cycle. A growing business spends heavily to win customers; a declining one spends heavily to keep them. The ratio swings either way.
  • Customers differ. Often the most profitable customers are the most expensive to keep, because competitors want them too.

What newer evidence says

In 2002 Werner Reinartz and V. Kumar published The Mismanagement of Customer Loyalty in Harvard Business Review. They studied the customer databases of four companies: a US high-tech corporate service provider, a large US mail-order company, a French retail food business and a German direct brokerage. Together that covered more than 16,000 customers over four years.

The link between how long customers stayed and how profitable they were was weak to moderate, with correlations from 0.20 at the mail-order company to 0.45 at the grocery retailer. They found little or no evidence that steady customers were cheaper to serve, less price sensitive or better at bringing in new business. At the service provider, about half of the customers labelled loyal barely made a profit, and about half of the most profitable customers bought heavily for a short time and then left.

Other work points the other way on the size of the prize. Sunil Gupta, Donald Lehmann and Jennifer Ames Stuart of Columbia Business School valued the customers of five US firms, including Capital One, Amazon and eBay, from public data as at March 2002. In their model, a 1% improvement in retention lifted customer and firm value by 3% to 7%, a 1% improvement in margin by about 1%, and a 1% improvement in acquisition cost by only 0.02% to 0.3%.

Read together, the evidence says retention can be one of the most powerful levers a business has. It does not say every retained customer is profitable, or that keeping customers is free. The Gupta figures are a valuation model, not a test of any retention program.

What an organisation should actually take from it

None of this evidence is Australian, and most of it is more than twenty years old. That is a good reason to stop quoting the headline numbers and build the case from your own data instead. In practice, this is what we put on the wall:

  • A retention curve by joining cohort, so you can see when customers leave, not just how many.
  • Margin and cost to serve by segment, so you know which customers are worth more effort and which are not.
  • The reasons people leave, taken from exit conversations, complaints and frontline staff. This was the most useful idea in the 1990 article: defections tell you where you are falling short.
  • The full cost of keeping customers, including discounts, retention offers and extra service, set against what they are worth.
  • A clear split of marketing spend between winning and keeping customers, rather than assuming all of it is acquisition.

The honest version of the claim

If you need one line for a business case, we suggest something like this: in businesses with recurring revenue and real acquisition costs, keeping the right customers longer usually improves profit, and here is what our own numbers say it is worth.

That is less dramatic than 25% to 95%. It is also a claim you can defend when finance asks where it came from.

Sources

See your journey from the customer's side

We map the end-to-end journey with you, find the moments that matter most, and build a strategy your teams and systems can deliver.

Start a conversation More insights