Marketing's role in client retention: what works

What is client retention?

Client retention is the share of your customers who keep buying from you over a set period. It is the opposite of churn, which is the share who stop. Retention covers repeat purchases in retail and D2C, renewals in subscriptions and contracts, and continued use in apps and services.

Retention is not the same as loyalty. A customer may stay because switching is hard; loyalty is choosing you even when a rival is cheaper. Marketing influences both, so know which one you are measuring.

For most businesses, the first sale only covers the cost of winning the customer. Later sales make the margin.

How is customer retention measured?

Retention is measured with four numbers: a retention or churn rate, cohort curves, customer lifetime value (CLV), and, for B2B and subscription businesses, net revenue retention (NRR). Each answers a different question.

Measure	Formula	The question it answers
Retention rate	(Customers at end of period − new customers won in the period) ÷ customers at start	What share of the people we had did we keep?
Churn rate	Customers lost in the period ÷ customers at start (or 100% − retention rate)	What share did we lose?
Cohort curve	Share of each starting group (say, everyone who first bought in January) still active in month 1, 2, 3 and so on	When do customers drop off, and is it getting better?
Customer lifetime value	Average margin per customer per period × expected number of periods they stay (discounted for time)	What is a customer worth, and so what can we spend to win and keep one?
Gross revenue retention	Revenue this year from last year's customers, excluding upsells and price rises ÷ last year's revenue from them	How much revenue did we keep, ignoring growth?
Net revenue retention	The same, but including upsells, cross-sells and price rises	Did our existing customers grow or shrink as a group?

Example: a skincare brand starts the year with 1,000 customers, wins 600 and ends with 1,250. Retention is (1,250 − 600) ÷ 1,000, or 65%. Churn is 35%. The 600 new customers do not hide the 350 who left.

Cohort curves matter more than a single rate, because a blended number mixes five-year customers with last week's first-timers. Most curves fall steeply after the first purchase and then flatten. The height of that flat part is the number to improve.

CLV turns retention into money. It is the expected sum of future margin from a customer, discounted for time. That is the definition used by Columbia's Sunil Gupta, Donald Lehmann and Jennifer Ames Stuart in "Valuing Customers", published in the Journal of Marketing Research in 2004. Compare CLV with what you spend to win a customer: if CLV is not well above acquisition cost, growth will burn cash. I covered the acquisition side in why D2C acquisition costs rise in India.

NRR is the headline number for B2B. SaaS Capital defines it as the revenue in a given month from customers who were also customers a year earlier, divided by their revenue a year earlier, including upsells and price rises. Its 2025 retention benchmarks for private B2B companies put median NRR at 102% for companies with annual contract values of $25,000 to $50,000. Above 100% means the existing base grows even before you win a single new client.

Why does client retention matter?

Retention matters because it drives the value of the whole business more than almost any other marketing lever. The best evidence for that is not the statistic most people quote, so it is worth separating the two.

The strong evidence: retention moves customer value

Gupta, Lehmann and Stuart modelled the customer base of five companies (Amazon, Ameritrade, Capital One, eBay and E*Trade) from public data. In their paper, a 1% improvement in retention raised customer value, and so firm value, by about 3% to 7%. A 1% improvement in margin raised it by about 1%. A 1% cut in acquisition cost raised it by only 0.02% to 0.3%.

In those firms, keeping customers a little longer was worth far more than squeezing the cost of winning them. It is a model built on five companies, not a law, but it is peer reviewed and the method is open to check.

Winning a customer earns you a chance. Keeping them is where the money is made.

The famous statistics, and what they actually said

Two numbers appear in nearly every retention deck. Both trace back, loosely, to Frederick Reichheld of Bain & Company, the consultant who later created the Net Promoter Score. Neither says quite what people think.

"A 5% increase in retention increases profits by 25% to 95%." The trail runs like this:

1. In 1990, Reichheld and Harvard professor W. Earl Sasser published "Zero Defections: Quality Comes to Services" in Harvard Business Review. As Ipsos summarises it, the article said that reducing customer defections by 5% could boost profits by 25% to 85%. The range was 85%, not 95%, and it came from service businesses.
2. In 2001, Reichheld wrote a Bain brief, "Prescription for cutting costs". It says that "in financial services, for example, a 5% increase in customer retention produces more than a 25% increase in profit." That is one industry and one floor figure.
3. In 2014, a widely shared HBR article, "The Value of Keeping the Right Customers", said Reichheld's research "shows increasing customer retention rates by 5% increases profits by 25% to 95%," and linked to that same Bain brief. I read the brief and could not find 95% anywhere in it.

The 1990 claim has also been challenged on its method. Professor Byron Sharp of the Ehrenberg-Bass Institute calls it a "thought experiment" with no empirical research behind it. He notes it treated a fall in defections from 10% to 5% as the "5%", which halves churn, and assumed that cutting defections cost nothing. Ipsos makes a similar point: the profit gain depends heavily on a firm's current retention rate and margin, and a 25% to 85% jump would need a very thin starting margin.

The fair reading: Reichheld's work showed that in some service businesses, small gains in retention could produce large gains in profit. It did not show that every company gets 25% to 95% more profit from 5% more retention.

"Acquiring a customer costs five times more than keeping one." This one has no clear origin I could trace. The 2014 HBR article gives a range of "anywhere from five to 25 times," and adds that it depends "on which study you believe, and what industry you're in." It does not cite a study. Reichheld's Bain brief argues that serving a long-standing customer costs less than courting a new one, but gives no multiple. Treat any fixed ratio as a rhetorical device and calculate your own.

The counter-evidence: not every loyal customer is profitable

Werner Reinartz and V. Kumar tested the loyalty-profit link directly in "The Mismanagement of Customer Loyalty" (HBR, July 2002). They tracked more than 16,000 customers at four companies: a high-tech service provider, a US mail-order firm, a French grocery retailer and a German brokerage. The correlation between how long customers stayed and how profitable they were was weak to moderate, from 0.20 to 0.45. They found little evidence that long-standing customers were cheaper to serve, less price sensitive or better at bringing in new business.

At the service provider, about half of the customers flagged as "loyal" barely made a profit. That does not mean retention is unimportant. It means retaining the right customers matters more than retaining everyone.

How does marketing drive client retention?

Marketing drives retention by shaping what customers expect, how quickly they get value, and how often they are reminded of it. Six levers do most of the work. I would run them in this order, because each one depends on the one before.

1. Set honest expectations before the sale

Retention starts in the advert. Promise results in seven days for a product that takes six weeks, and churn is built in. Reichheld's Bain brief makes the same point from the other side: reward sales teams and channels for customers who stay, and rank acquisition campaigns by the share of loyal customers they bring in, not just by cost per lead.

2. Design onboarding around the first moment of value

Onboarding is the period between the first purchase and the moment the customer gets what they paid for. Map the steps a new customer must take, find the slowest one, and use welcome messages, how-to content and a check-in to get them through it: the first session in an app, the first appointment in a service, the first live use in B2B.

3. Run lifecycle CRM, not campaigns

Lifecycle CRM means sending the right message at each stage: welcome, first use, repeat purchase, lapse risk and win-back. Trigger messages on behaviour (a missed refill date, a drop in log-ins) rather than on the calendar. In India, this work has rules: consent for personal data under the DPDP Act, and registered templates and headers for SMS under TRAI's DLT system.

4. Build loyalty that rewards the right behaviour

A loyalty programme should reward what is profitable for you and useful for the customer: frequency, a wider basket, referrals, a longer plan. Points that only discount what people would have bought anyway cost margin and teach customers to wait for offers. Reinartz and Kumar's findings are a warning here: spend more on the customers who are both loyal and profitable, and less on the rest.

5. Create community and content worth returning for

Community gives customers a reason to stay that is not a discount: other customers. That can be a WhatsApp group, a user forum or events for top clients. Content helps customers use what they bought, and every successful use is a reason to renew.

6. Treat service recovery as marketing

Every complaint is a moment when a customer decides whether to stay. Respond fast, fix it and follow up. Do not count on the "service recovery paradox", the idea that a well-handled failure leaves customers happier than no failure at all. A 2007 meta-analysis by de Matos, Henrique and Rossi in the Journal of Service Research found the effect held for satisfaction but was not significant for repurchase intentions or word of mouth. Good recovery limits the damage; it rarely improves on getting it right first time.

When should each retention lever start?

Retention work runs in a sequence over the first 90 days of a programme.

• Week one: calculate retention, churn and a simple CLV for the last 12 months, and build one cohort chart.
• Weeks two to four: find the biggest drop-off in the cohort curve and fix onboarding around it.
• Month two: set up three behaviour-triggered lifecycle messages: welcome, lapse risk and win-back.
• Month three: review which acquisition channels bring customers who stay, and move budget towards them.

Loyalty and community come after that, once you know which customers are worth rewarding.

Where does retention marketing differ between B2B and B2C?

B2B and B2C retention use the same levers, but the unit, the metric and the owner differ.

	B2C and D2C	B2B
What you retain	Individual customers, often without a contract	Accounts with several users and decision-makers
Main metric	Repeat purchase rate, cohort retention, CLV	Logo retention, gross and net revenue retention
Main moment	The second purchase	The renewal, and the expansion before it
Main channels	CRM messages, app notifications, loyalty, community	Account management, business reviews, user training, client events
Who leaves	One person, often silently	A committee, often after a champion leaves the client

In B2C, the danger is silent churn: customers stop buying and never tell you. In B2B, it is a renewal decided by people who never use the product, so marketing's job is to arm the client's champion with usage reports, case studies and proof of return.

In India, both sides have one extra feature: price sensitivity is high, but so is relationship loyalty. KPMG's 2019 customer loyalty survey of 1,721 Indian consumers found 47% unlikely to switch to a competitor even after a bad experience, and 74% said personal connection drove their loyalty. The data is now several years old, so treat it as a direction, not a benchmark.

Where does retention marketing go wrong?

Retention marketing goes wrong in three common ways.

• Retaining the wrong customers. Discounts that keep unprofitable customers flatter the rate and hurt the business.
• Hiding the exit. Hard cancellation keeps the number up for a month and damages trust for years. In India, the dark patterns guidelines name subscription traps as a banned practice.
• Messaging instead of fixing. No email sequence saves a product that does not deliver.

Who owns client retention: marketing, product or service?

One senior leader should own the retention number, and three teams should each own a part of it. Splitting the number across teams with no single owner is the most common reason retention plans stall.

• Product owns whether the offer delivers value: features, quality, pricing, the core experience.
• Service and customer success own support, complaint handling and, in B2B, renewals.
• Marketing owns expectations, onboarding communication, lifecycle CRM, loyalty, community and content, and the analysis that shows which customers and channels are worth keeping.
• Sales owns the fit of the customers it brings in, which is why Reichheld suggested tying commission to customers who stay.

My view

My view is that marketing teams spend most of their energy on the first sale and treat everything after it as someone else's job. That is backwards. At VLCC I supervised AI-driven projects that automated CRM journeys and lead scoring, and led the launch of a D2C app and PWA with subscription models. Subscriptions make retention impossible to ignore, because every month the customer decides again.

I would stop quoting the 5% and 25% to 95% figures. They are not quite what Reichheld said, and an expert in the room will know it. Gupta, Lehmann and Stuart's 3% to 7% is smaller, better supported and more persuasive to a finance team.

The trade-off is that retention work is slow and shared. A campaign shows results in a week; better onboarding takes a quarter to show in the cohort curve, and the credit is split with product and service. I would accept it, because it is the work that keeps growth from depending on ever-rising acquisition budgets.

Start with one cohort chart

This week, take every customer who first bought from you twelve months ago and plot how many bought again in each month since. Find the month where most of them stopped. That single drop-off is your first retention project, and it tells you whether the fix belongs to marketing, product or service.