How Software Companies (SaaS) Make Money

Software as a service, or SaaS, is the model behind most of the tools you already pay for every month: the email platform, the store builder, the design app, the project tracker. From the outside it looks like the best business ever invented. Build something once, charge a small fee every month, and watch the same customers pay you again and again without you doing anything new. That story is half true. The half nobody tells you is that recurring revenue is a slow build, an expensive one, and it leaks constantly through a hole called churn.

The short version#

A SaaS company sells access to software instead of selling the software itself. You do not buy the program and own it forever. You rent it, usually monthly or yearly, and you keep paying for as long as you keep using it.

The money comes from that repetition. One customer who pays $30 a month is worth far more than one customer who pays $30 once, because if they stay for two years that single sale quietly turns into $720. Multiply that by a few thousand customers who mostly stick around, and you have a business with predictable income that arrives whether or not you sold anything new this week.

The catch is that "mostly stick around" is doing enormous work in that sentence. Every month some customers cancel. If they cancel faster than you can replace them, your predictable revenue starts predictably shrinking. So SaaS is not really a game of getting customers. It is a game of getting them and keeping them, and the keeping is the hard part.

Where does the money actually come from?#

You build software that solves a recurring problem
  ↓
A customer signs up and pays a monthly (or yearly) fee
  ↓
They keep using it, so they keep paying, month after month
  ↓
Each month: new signups are added, some existing customers cancel
  ↓
Monthly recurring revenue = (paying customers) x (average fee)
  ↓
Profit = recurring revenue, minus hosting, support, and the cost of finding customers
  ↓
The business grows only if new revenue outruns cancellations

The money comes from recurring revenue that compounds. The reason software is a good vehicle for this is margin: once the product is built, the cost of serving one more customer is tiny. Adding a customer to a physical business means another unit of inventory. Adding a customer to a SaaS product often just means a bit more server usage. That is why healthy software businesses can carry very high gross margins. It is also why they are worth building despite being slow, a point that connects to what makes a business scalable.

How it actually works#

The engine has two sides, and both run constantly.

On one side, you are acquiring customers. That costs money, whether through ads, content, sales, or partnerships. The important thing is that in SaaS you usually pay the full acquisition cost up front but collect the revenue slowly, a little each month. So there is a period where every new customer is temporarily a loss, and you only come out ahead once they have paid you long enough to cover what you spent to get them.

On the other side, you are losing customers. This is churn, and it is the single most important concept in the whole model. Churn is the percentage of customers who cancel in a given month. It sounds small and harmless at 5 percent. It is not. At 5 percent monthly churn, you lose almost half your customer base over a year if you add nobody new. Every SaaS business is a leaky bucket, and churn is the size of the hole.

This is why the metric that governs the business is the relationship between how much a customer is worth over their lifetime and how much it cost to acquire them. Those two numbers, lifetime value and acquisition cost, decide whether the whole thing is a real business or a treadmill. We break each of them down in customer lifetime value explained and in revenue per subscriber explained.

A simple example with numbers#

These figures are hypothetical and exist to show the mechanism, not to promise a result. Say you run a small SaaS tool that charges $30 a month.

Price per customer:            $30 / month
Customers at start of month:   1,000
Monthly recurring revenue:     $30,000

Now the two forces:
  New customers added:         +80  (+$2,400)
  Customers who cancel (churn):-50  (-$1,500)   [5% of 1,000]
  ----------------------------------------------------
  Net customer change:         +30
  Customers next month:        1,030
  Recurring revenue next month:$30,900

Growth of $900 in a month. Slow, but it repeats and it compounds. Now watch what churn does to the same business.

• If churn rises from 5 percent to 8 percent, you lose 80 customers a month instead of 50. With 80 new signups, you now break even and grow nothing. Same product, same marketing, flat business.
• If you can cut churn to 3 percent, you lose only 30, so the same 80 signups grow you by 50 a month instead of 30. Reducing cancellations did more for growth than finding new customers.
• The average customer who pays $30 and stays 20 months is worth $600. If it cost you $200 in ads to acquire them, that works. If it cost you $650, every customer is a loss no matter how the dashboard looks.

That last point is the whole game. This is why SaaS founders obsess over churn and retention far more than over signups. Keeping customers is cheaper than replacing them, and it is what turns a leaky bucket into a growing one. The same logic makes membership sites work, since they run on nearly identical math.

What you need#

• A recurring problem worth software. The best SaaS solves something a customer faces over and over, so the tool earns its place in their monthly budget. One-time problems do not sustain recurring fees.
• The ability to build and maintain software, yourself or with a team. This is the real barrier, and it separates SaaS from most models on this site.
• A way to acquire customers repeatably, because you must keep filling the bucket as it drains.
• Ongoing capacity for support and updates. Renting software means customers expect it to keep working and keep improving. Neglect it and churn climbs.

What it costs#

Required: development time or developer cost to build the product, hosting and infrastructure that scale with usage, and a customer acquisition budget. Unlike a course or an ebook, a SaaS product is never truly finished, so maintenance is a permanent line item.

Optional: paid tools for support, billing, analytics, and onboarding. Useful, but easy to over-buy before you have enough customers to justify them.

Nice to have: a polished brand and design. It helps retention, but it matters far less than the product actually solving the problem reliably.

How long it takes#

SaaS is one of the slower models to pay off, and honest builders will tell you so. Recurring revenue starts small because it stacks one subscription at a time, and you often spend months building before the first customer pays anything at all. The upside is that once the base is built and churn is under control, the revenue is unusually durable. This is the opposite of a quick flip. For where it sits among other models, see how digital products make money.

What beginners usually get wrong#

• Believing recurring revenue is automatic. It recurs only while customers stay. Ignore churn and the "passive" income quietly evaporates.
• Counting signups instead of retention. A flood of trials that all cancel is a vanity metric, not a business.
• Underpricing. Very low prices attract customers who churn fast and cost the same to support as good ones. Price for the value delivered, not for what feels safe.
• Forgetting the product is never done. SaaS carries permanent maintenance and support costs that a one-time product does not.
• Spending more to acquire a customer than the customer is worth. The most common way software businesses lose money while looking busy.

How I would start#

1. Find a specific, recurring problem for a specific type of user, ideally one I understand from experience.
2. Build the smallest version that genuinely solves it, and charge for it early rather than perfecting it in private.
3. Track churn from the very first paying customers, because it is the number that decides everything.
4. Talk to the people who cancel and fix the reasons they leave before spending heavily to find more.
5. Only pour money into acquisition once I know a customer is worth more than it costs to get one.

What I would not do#

I would not treat "monthly recurring revenue" as a synonym for "money I keep," since hosting, support, and acquisition all take their cut. I would not chase signups while ignoring the back door where customers walk out. And I would not believe any pitch that frames SaaS as a fast or hands-off way to make money. It is a genuinely strong model, but it rewards patience, retention, and a product that actually works, not hype.