Business Strategy / Economics / Accounting & Finance
The Fortune in the Second Payment
Every business celebrates its first sale. The great ones
build a reason for customers to pay again. Inside the
economics of recurring revenue—and the costly mistakes
hiding behind the subscription boom.
Monthly recurring revenue: a run-rate,
not cash in the bank
Net revenue retention: how an existing
customer cohort changes
Customer acquisition cost: the price
of winning a relationship
Imagine two businesses that each earn $10,000 this month.
One sells an expensive machine and must find another buyer
next month. The other maintains 100 machines for $100 each,
every month, under service agreements. On the income
statement, the headline is the same. In the owner’s
imagination, the future looks entirely different.
The first business wakes up to a blank sales pipeline.
The second starts with a base of customers who already
know the company, have a continuing need and expect
another invoice. It still has to deliver excellent service.
Customers can leave; machines can break; costs can rise.
But it has escaped one of commerce’s most expensive habits:
having to persuade a stranger to make every purchase
from scratch.
This is the attraction of recurring revenue:
income generated repeatedly by continuing customer
relationships, whether through subscriptions, service
contracts, memberships, repeat purchases or
consumption-based arrangements. It is the commercial
equivalent of entering a new month with part of the
next month’s business already spoken for.
Recurring revenue is not new. Landlords, insurers,
utilities, newspapers and maintenance contractors
understood it long before software companies gave it
the labels MRR and ARR. What changed was the ease with
which digital payments, cloud technology and customer
data allowed businesses to package more kinds of value
as ongoing services. Almost anything, it seemed, could
become a subscription. Not everything should.
The secret is not getting someone to pay every month.
It is giving someone a reason to be glad they do.
The difference between recurring revenue and
recurring resentment
Why investors love revenue that comes back
Ordinary sales are episodic. A restaurant sells dinner
tonight, a consultant sells a project, a retailer sells
a chair. The business knows what it has already earned,
but next month’s revenue depends on another round of
persuasion, foot traffic or luck. Recurring income
replaces some of that uncertainty with a customer base,
contracts and observable renewal behaviour.
That improvement matters in five ways.
Forecasting becomes more credible:
managers can plan staffing and inventory around existing
relationships.
Customer-acquisition spending can be spread
over time: the cost of winning an account may
be recovered across many payments rather than one
transaction.
Cash flow can become steadier,
especially when billing is reliable.
Customer knowledge compounds:
the provider learns which features, repairs or deliveries
clients actually need. And
business value may rise when profits
are durable and less dependent on the founder constantly
hunting for work.
But the last point has an important qualification.
Investors do not pay a premium simply because an invoice
repeats. They care whether repeat revenue is profitable,
collectible, diversified and likely to continue.
A portfolio of unhappy subscribers locked into contracts
can be less valuable than a loyal customer base that
returns voluntarily. Predictability is earned through
retention; it cannot be manufactured by changing the
billing frequency.
Contractual: customers agree to ongoing
payments, such as software access or a maintenance
agreement.
Repeat purchasing: customers reliably buy
again without being contractually obliged, such as
a café’s regulars or monthly consumables.
Usage-based: payment recurs whenever the
customer uses a service, such as cloud computing
or transaction processing.
All three may be attractive; only some qualify cleanly
as contracted recurring revenue or conventional ARR.
Four companies, four ways to make the second sale
Adobe: turn a product into continuing access
Adobe once depended heavily on selling software licences
and upgrades. Under its subscription-led model, customers
pay for continued access to creative tools, cloud services
and updates. Its fiscal 2025 annual filing reports
$25.20 billion in total annualized recurring
revenue at year-end.
Adobe’s ARR is its defined operating measure, not revenue
already earned or guaranteed cash. The strategic
achievement is not merely annual billing. It is the
ongoing bundle of useful software, workflow compatibility,
cloud features and updates that makes renewal
commercially plausible.
Costco: charge for belonging, then earn the shopping trip
Costco’s membership model shows why recurring revenue
is not exclusively a technology story. In fiscal 2026,
the retailer reported $5.907 billion in
membership-fee revenue and an
89.8% worldwide member-renewal rate.
The annual fee is only part of the economic equation:
customers continue renewing because they believe the
combination of price, assortment and shopping experience
justifies the privilege of entering the warehouse.
Membership strengthens the relationship; the merchandise
still has to deliver value.
Rolls-Royce: sell performance, not merely machinery
Aircraft engines are expensive, but their operators
care just as much about avoiding unscheduled maintenance
and keeping aircraft flying. Rolls-Royce’s TotalCare
service contracts price engine support through a payment
mechanism based on engine flying hours.
The manufacturer gains an ongoing service relationship;
the airline obtains a more predictable maintenance-cost
arrangement. Yet the provider also assumes complex
engineering and operating risks. Recurring revenue becomes
a promise to deliver an outcome—not just an entitlement
to collect a fee.
Microsoft: a long backlog is not today’s income
Microsoft’s fiscal 2026 annual report disclosed
$678 billion in commercial remaining performance
obligations at June 30, 2026.
That is revenue associated with remaining contracted
performance obligations under accounting rules, expected
to be recognised in future periods subject to the contract
terms. It is not the same as money already received, nor
does it automatically equal ARR or future profit.
The figure demonstrates how valuable visibility can be
while reminding investors to distinguish a commercial
commitment from a deposit in the bank.
Company figures come from the respective annual filings
and official product disclosures listed in the source notes.
The nine ways to build revenue that returns
The best recurring-revenue model usually begins with a
recurring customer problem. A company that solves a
once-a-decade problem should not invent monthly charges
to satisfy a business-plan spreadsheet. A company that
solves a weekly inconvenience may be leaving substantial
value—and stability—on the table by selling only once.
Charge monthly or annually for software, research,
education, content or a digital service that
continues to be useful.
Provide continuing privileges: preferential pricing,
expert access, community, facilities, service
priority or bundled benefits.
Offer inspections, repairs, updates, compliance
checks, equipment servicing or preventive
maintenance at agreed intervals.
Give clients defined ongoing support such as monthly
bookkeeping, IT monitoring, financial reporting,
design or advisory hours.
Deliver products people consume repeatedly:
filters, office supplies, hygiene items, food
staples or industrial materials.
Bill for transactions, storage, processing, machine
hours or other measurable consumption; useful, but
less predictable than fixed fees.
Provide continued use of equipment, vehicles,
premises or specialist tools in exchange for
periodic payments and clear obligations.
Earn periodic payments for intellectual property,
brands, patented technology, data or licensed systems
where agreements permit.
Sell a core product, then provide optional
installation, consumables, training, warranties
or managed support over its useful life.
These models have different accounting, margin and
regulatory profiles. Rental income is not automatically
treated as customer-contract revenue; regulated services
and extended warranties may have special rules; royalties
may fluctuate; replenishment orders are not necessarily
committed. The business model should follow customer need
and sound contract design, not a desire to label everything
as ARR.
The easiest place to start is the sale you have already made
Founders often search for a brilliant new subscription
idea while overlooking the existing customer who already
has a recurring need. A web designer can offer security
monitoring, updates and accessibility checks. An accounting
practice can shift from a once-a-year compliance engagement
to monthly reconciliations, management accounts and
cash-flow reporting. A cleaning company can sell a fixed
maintenance schedule. An equipment distributor can provide
replacement parts, calibration and on-site inspections.
These are not tricks for collecting payments after the
real work is finished. They are extensions of the service
customers need after the original purchase. The web
designer’s client’s website is not “complete” forever.
Books do not reconcile themselves each month. Equipment
does not stop wearing out after installation. A useful
recurring offer turns that ongoing responsibility into an
explicit service standard.
Win trust and solve the immediate problem
Identify what must happen again
Deliver measurable continuing value
Make staying a deliberate, worthwhile choice
Every cycle either strengthens the relationship
through real value or weakens it through disappointment.
Consider a hypothetical independent bookkeeper serving
small businesses. One-off clean-up jobs bring revenue in
bursts. A monthly package that includes transaction
reconciliation, payroll coordination, management reports
and quarterly reviews could create a steadier business.
But the firm must calculate staff hours, peak workloads,
software costs and support boundaries. Unlimited work for
a fixed fee is not a recurring-revenue model; it is a
recurring-margin problem.
For retailers, the answer may be neither software nor a
membership. A shop selling water filters could invite
customers to subscribe to replacement cartridges at
sensible intervals, with easy skips and a price advantage.
A coffee roaster could offer a predictable delivery plan.
A machinery supplier could offer an inspection and
parts-replenishment agreement. In each case the customer
already has a reason to return. The entrepreneur’s job
is to make that return easier.
Find the problem before designing the plan
Before setting a monthly price, ask three uncomfortable
questions. Does the customer face this problem
repeatedly? Will the customer notice
the service’s absence? Can the company
deliver it profitably at scale? If the answer
to any of these is no, the subscription may be an
accounting fantasy.
A successful offer usually promises one of four things:
continuing access, less work, lower risk or a more reliable
outcome. A language-learning subscription promises ongoing
practice. A cybersecurity-monitoring service promises
vigilance. A preventive-maintenance contract promises
fewer surprises. A replenishment service promises that
essential supplies will arrive before they run out.
These are more persuasive reasons to subscribe than
an arbitrary discount.
Make the package specific. Define included services,
response times, quantities, access rights, usage limits,
renewal terms, exclusions and cancellation procedures.
Price the job around the economics of fulfilment rather
than copying a competitor’s cheapest plan. A service
business should model how many accounts one employee
can responsibly support. A software provider should
model infrastructure, support and payment-processing
costs. A product replenishment firm must model shipping,
returns and inventory risk.
Then test it with a small cohort of customers who can
decline without losing their existing relationship.
Selling ten honest renewals is more informative than
collecting a hundred trial sign-ups that never pay
a second invoice.
Three prices: the customer sees one; the accountant sees another
Monthly billing looks accessible, annual billing brings
cash forward, and usage-based pricing aligns the bill
with consumption. Each produces a different economic
experience for customer and supplier.
| Pricing method | Customer advantage | Business advantage | Hidden risk |
|---|---|---|---|
| Flat monthly fee | Easy to understand and budget | Predictable scheduled billings | Heavy users can overwhelm margins |
| Annual prepayment | Potential discount, fewer payments | Cash arrives earlier | Future service and possible refund obligations remain |
| Per user or seat | Price follows team size | Expansion opportunities | Seat cuts can cause contractions |
| Usage based | Pay in proportion to use | Revenue grows with activity | Revenue and costs become volatile |
| Base fee plus usage | Core access plus flexible capacity | Minimum income with upside | Surprise bills can damage trust |
| Tiered service plan | Choice by needs and budget | Upsells when value increases | Confusing packages reduce conversion |
A useful pricing ladder has an entry plan that reliably
solves a narrow problem, a middle plan covering the most
common ongoing needs, and a higher tier where added
features or service genuinely cost more or deliver more
value. Avoid inventing arbitrary restrictions just to
push buyers upward. Pricing should help customers select
an appropriate solution, not punish them for understanding
the fine print.
The mathematics of keeping customers
Now the uncomfortable part. Recurring revenue looks
spectacular in a pitch deck because a monthly amount can
be multiplied by twelve. Real customers do not renew
forever, however. Some become dissatisfied. Some close
their businesses. Others switch providers, reduce spending
or simply forget why they subscribed.
Suppose a company has 100 customers paying $50 a month.
Its starting monthly recurring revenue (MRR)
is $5,000, and the simple annualised run-rate
is $60,000. That does not mean it has earned $60,000,
signed twelve-month contracts or guaranteed twelve months
of payments. It means $5,000 multiplied by twelve—before
customer changes.
With a monthly customer churn rate of 5%, a starting
cohort of 100 customers shrinks to approximately 54
after twelve monthly retention cycles if no one is added
and the rate remains constant. At just 2% monthly churn,
approximately 78 remain. The difference is not academic:
keeping clients is often the difference between a thriving
business and a company that must continuously spend to
replace its own disappearing revenue.
The retention effect / illustrative calculation
100 × (1 − 0.05)12
≈ 54 customers
At 5% constant monthly customer churn, roughly 54%
of the original customers remain after twelve monthly
cycles. At 2% churn, about 78% remain. These are
simplified cohort calculations, not forecasts of
actual customer behaviour.
Churn also has two meanings. Customer churn
measures the number of relationships lost.
Revenue churn measures the associated
lost billings. Losing one client who pays $5,000 a month
can be more damaging than losing ten who each pay $50.
Conversely, one loyal client upgrading substantially
can offset several departures in revenue terms without
repairing the underlying customer experience.
Net revenue retention: can the existing base grow by itself?
Take a cohort of customers paying $100,000 per month
at the beginning of a period. During the period,
cancellations remove $7,000, downgrades remove $3,000,
and existing customers upgrade by $15,000. The cohort
now contributes $105,000. Its net revenue retention
(NRR) is 105%: the original customer base grew
5% in revenue terms without counting any new customers.
New customers are excluded from this calculation.
Net revenue retention / example
($100,000 − $7,000 − $3,000 + $15,000)
÷ $100,000 = 105%
NRR above 100% signals that expansion exceeded
losses within the starting cohort. It does not
automatically mean customers are happy or profits
are healthy. Definitions can vary, so companies
should disclose their methodology.
A company can report NRR above 100% while losing many
small clients, if a few large customers spend dramatically
more. That may be a terrific enterprise strategy—or an
alarming concentration risk. Read the retention number
alongside customer counts, gross revenue retention,
average account value and customer concentration.
The other number founders routinely underestimate
Customer acquisition cost, or CAC,
is what the company spends to win a new paying customer.
It should include appropriate sales and marketing costs,
not merely advertising clicks. A business that spends
$14,000 to acquire 100 customers has a simple CAC of
$140. The crucial question is how long those customers
must remain to pay back that investment.
Assume each customer pays $40 per month and the recurring
service has a 70% gross margin. That produces $28 of
monthly gross profit per customer before other operating
expenses. At $140 CAC, the simple gross-margin
payback is five months ($140 ÷ $28).
This is attractive only if customers tend to stay well
beyond that period and the business has enough capital
to fund the gap.
Under a constant 5% monthly churn assumption, a crude
undiscounted customer-lifetime estimate is 1 ÷ 0.05,
or 20 months. Multiplying 20 by $28 suggests $560 of
lifetime gross profit before acquisition costs, corporate
overhead, financing and taxes. It is a back-of-the-envelope
model, not a promise of profit: churn rates change by
cohort, fulfilment costs fluctuate, and some customers
require disproportionately expensive service.
A tiny subscription business—two very different outcomes
Illustrative inputs: $40 monthly fee; 70% gross
margin; $140 acquisition cost. Monthly gross profit
per active customer: $28. Simple CAC payback:
five months.
If the customer cancels after three paid months, gross
profit of $84 has not repaid acquisition cost. If the
customer stays for twenty months, gross profit totals
$560 before overhead, financing and other costs.
The difference is retention, not clever invoicing.
Businesses should also monitor failed payments, refunds,
cancellation reasons and cohort profitability. A rise
in sign-ups can conceal deteriorating renewal quality.
It is possible to grow MRR while destroying cash through
escalating advertisements, discounts and customer support.
Growth is valuable only when the additional relationships
ultimately generate attractive returns.
Why cash in the bank is not always revenue earned
Recurring-revenue marketing often celebrates annual
prepayments as a financial superpower. Customers pay
early; the supplier receives working capital before
delivering most of the service. It can be valuable,
especially for young companies funding growth. But the
accounting treatment prevents a seductive mistake.
Suppose a customer pays $1,200 in advance for twelve
months of a stand-ready service that is provided evenly
throughout the year. The company may collect all $1,200
on day one. Under the applicable revenue-recognition
framework, assuming the contract qualifies and the service
is satisfied evenly, the company generally recognises
approximately $100 of revenue each month.
At the start, the unearned portion is a
contract liability (deferred revenue),
because the company still owes the customer future service.
| Event | Cash impact | Revenue recognised | Contract liability* |
|---|---|---|---|
| Annual payment received | +$1,200 | $0 | $1,200 |
| After month one | $0 additional | $100 | $1,100 |
| After month six | $0 additional | $600 cumulative | $600 |
| After month twelve | $0 additional | $1,200 cumulative | $0 |
*Simplified example: assumes one evenly delivered service,
no tax, refund, financing-component or other complicating
factors. Actual accounting depends on contract terms
and the applicable reporting standard.
This distinction between billings, cash receipts,
recognised revenue and recurring-revenue run-rate
is essential. They answer four different questions:
what did the company invoice, what did it collect,
what did it earn, and what would its current active
recurring base imply if annualised?
IFRS 15 and the corresponding US customer-contract revenue
framework focus on performance obligations and when control
of promised goods or services transfers to the customer.
Advance cash may strengthen liquidity without making
the unperformed portion of a service profitable or earned.
A company that spends prepaid customer cash recklessly
can face a liquidity crisis when renewal season
arrives—even if last year’s bank balance looked spectacular.
The subscription trap: when recurring revenue becomes recurring disappointment
The subscription economy has produced a new customer
emotion: paying for things one no longer remembers
choosing. Forgotten entertainment plans, little-used
software tools, overpriced maintenance packages and
memberships whose privileges have quietly eroded are
all symptoms of the same failure. The company optimises
for continuation of billing rather than continuation
of value.
This approach can briefly flatter revenue figures.
Then cancellations arrive, payment disputes increase,
trust deteriorates and customer-acquisition costs rise.
Making cancellation deliberately difficult may also
create consumer-protection and reputational risks;
the exact legal duties vary by jurisdiction and product.
Another danger is the unlimited-service
promise. A business might offer “unlimited support”
for a modest monthly fee, discover that a small group
of customers consumes almost all employee time, and
find that increased subscriptions require disproportionately
more labour. Revenue is recurring; losses are recurring too.
Then there is concentration risk. If 40% of a firm’s
contracted revenue comes from one enterprise customer,
the apparent stability may disappear with one procurement
decision. A diversified portfolio of smaller accounts
is not automatically better—servicing them may cost
more—but concentration deserves explicit management
rather than being hidden inside an impressive ARR total.
Finally, technology can change how much value the service
offers. A software subscription may face competition
from cheaper alternatives or automation. A consumables
business may confront lower-cost generic replacements.
A subscription service reliant on a single payment
processor, app marketplace or distribution partner may
find its economics changed by fees or platform rules.
Recurrence is a feature of a relationship, not a moat
around the business.
Recurring revenue is not passive income.
It is a promise that has to be kept again and again.
A better way to think about subscription economics
The 90-day route from one-off sales to lasting income
A small business does not need venture capital,
a software development team or thousands of followers
to test recurring revenue. It does need customers,
a recurring problem and the discipline to measure
whether the offer works.
-
Days 1–15: audit the customer journey.
Review the last 50–100 customers or as many as are
available. What do they buy again? What problems
appear after delivery? Which services do they
request informally? Interview customers about
frustrations rather than asking whether they
“want a subscription.” -
Days 16–30: design one continuing offer.
Select one clear recurring problem. Specify outcomes,
service levels, exclusions, response times, pricing
and a straightforward cancellation policy.
Calculate the direct costs of providing it. -
Days 31–45: invite a small pilot cohort.
Approach suitable existing customers. Explain exactly
why recurring service saves time, money or risk.
Avoid coercing buyers into a plan they do not need. -
Days 46–60: deliver and observe.
Track active customers, usage, hours spent, product
fulfilment, failures, support requests and first
renewals. Ask why customers would stay or leave.
Correct onboarding problems quickly. -
Days 61–75: inspect unit economics.
Calculate gross margin, acquisition costs, payback,
retention and cash obligations. Distinguish the cost
of recruiting a customer from the ongoing cost
of serving one. -
Days 76–90: expand only if the evidence holds.
Improve pricing and service capacity. Document routine
tasks. Automate reminders and collections with customer
consent, not through confusing terms. Scale the offer
that customers repeatedly find worth paying for.
The pilot can be astonishingly simple. A maintenance
firm might offer a scheduled quarterly inspection with
monthly billing. A tutoring business might provide
ongoing practice and feedback rather than isolated
lessons. A specialist consultant might issue a recurring
monthly market brief plus a defined advisory call.
The content, delivery standards and economics matter
more than whether the business has an elaborate app.
The dashboard every owner should examine
The temptation is to watch only one number: MRR.
A more intelligent dashboard reads the business as a
living system. It shows how existing customers change,
what it costs to win new ones and how much cash is
required to serve everyone.
| Measure | What it tells you | Question to ask |
|---|---|---|
| MRR / ARR |
Annualised recurring run-rate using a clearly stated methodology |
Is it growing because of new customers, upgrades or price rises? |
| Customer churn | Share of customers lost in a defined period |
Which cohorts and cancellation reasons drive departures? |
| Gross revenue retention | Existing revenue kept before expansions | How much revenue is eroding underneath new sales? |
| Net revenue retention |
Existing cohort revenue after contractions and expansions |
Can the base grow without new customers? |
| CAC payback |
Time for gross profit to recover acquisition spending |
Can the business finance the gap? |
| Gross margin | Revenue remaining after direct delivery costs |
Do heavy users or rising costs undermine profitability? |
| Collections / failed payments | How much invoiced cash actually arrives | Are billing problems causing avoidable churn? |
| Deferred revenue / obligations | Payments or commitments linked to future delivery | How much service does the company still owe? |
A healthy recurring business does not need every
metric to improve every month. It needs transparent
definitions, consistent reporting and a compelling
explanation when the numbers diverge. If ARR rises
but gross margins fall, investigate. If renewals
are strong but cash collections weaken, investigate.
If customers stay only because cancellation is difficult,
investigate before regulators or unhappy buyers
do it for you.
The best subscription may not look like a subscription
The most interesting opportunity is often to combine
a one-time sale with continuing usefulness.
A manufacturer can sell equipment and then maintenance.
A publisher can sell books and offer optional research
updates. A professional adviser can complete an initial
project and later provide periodic monitoring. A retailer
can sell a durable product and supply legitimate
replacements on a sensible schedule.
These hybrid models often suit established businesses
better than an abrupt attempt to put everything behind
a monthly paywall. Customers still retain the freedom
to buy once. Those who need continued help can pay for
it. The relationship broadens without destroying
the original proposition.
There is no reason to insist that every company become
a subscription business. Some customers prefer one-time
purchases, some services are genuinely episodic and some
sectors are too volatile for fixed commitments.
A contractor who chooses profitable project work over
loss-making retainers is not missing the future.
The goal is not the greatest possible proportion
of recurring revenue. It is the greatest sustainable
value created for customers, translated into durable
profit and cash flow.
Tomorrow’s income begins with today’s trust
In the language of finance, recurring revenue improves
visibility. In the language of marketing, it increases
lifetime value. In accounting, it creates a stream of
billings, performance obligations, costs and—if managed
well—profits. In ordinary human terms, it means that
a customer decides a second time, then a third, that
a business is worth keeping.
That is why the second payment matters more than the
first. The first sale may reflect excellent advertising,
persuasive salesmanship or a moment of urgency.
The second begins to reveal whether the experience
lived up to the promise. The tenth says even more.
And when customers stay for years because the product
reliably solves a problem, the company no longer has
to rebuild its entire future every morning.
The most valuable recurring revenue is not money
that keeps arriving because customers forgot to cancel.
It is money that keeps arriving because customers
cannot imagine doing without the value behind it.